Hide Notes 

Slide 1: pResented by the society of ActuARies, the cAsuAlty ActuARiAl society And the cAnAdiAn institute of ActuARies Risk Management: The Current Financial Crisis, Lessons Learned and Future Implications
Slide 2: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications introduction the current financial crisis presents a case study of a “financial tsunami” (as former federal Reserve chairman Alan Greenspan recently called it) on what can go wrong. its ramifications are far-reaching and the lessons learned will be embedded in risk management practices for years to come. As one of the premier enterprise risk professions in practice today, the actuarial profession is sharing its substantial insight into what went wrong and the implications for the future. on behalf of the society of Actuaries, the casualty Actuarial society and the canadian institute of Actuaries, we are pleased to provide a series of essays on Risk Management: The Current Financial Crisis, Lessons Learned and Future Implications. this e-book is the result of a call for essays on the subject coordinated by the following groups: • • • • The Joint Risk Management Section of the Society of Actuaries, Casualty Actuarial Society and Canadian institute of Actuaries The Investment Section of the Society of Actuaries International Network of Actuarial Risk Managers Enterprise Risk Management Institute International included in this are the opinions of a number of authors written in response to our call for essays. An essay is, essentially, a short non-fiction form of writing expressing the often subjective opinion of the author. the thoughts and insights shared herein are not necessarily those of the society of Actuaries, the casualty Actuarial society, the canadian institute of Actuaries, or the authors’ employers. We hope these essays will provide thought-provoking discussion and commentary in the months and years to come. sincerely, Don Mango, FCAS, ASA, CERA, MAAA Chair THE JoINT RISk MANAgEMENT SECTIoN of THE SoCIETy of ACTuARIES, CASuAlTy ACTuARIAl SoCIETy, And cAnAdiAn institute of ActuARies Marc altschull, FSA, CFA Chair the investMent section, society of ActuARies Dave Ingram, FSA, CERA, FRM, PRM, MAAA Founder INTERNATIoNAl NETwoRk of ACTuARIAl RISk MANAgERS Wayne Fisher, FCAS, ASA, CERA, FCIA, MAAA Executive Director ENTERpRISE RISk MANAgEMENT INSTITuTE INTERNATIoNAl 2
Slide 3: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications the essays ExpLanaTIons, CausEs and CuREs these essays address the philosophies and origins of how the crisis began, with anecdotes about what has happened and insight into encouraging thought leadership going forward. • • • • • • The Causes and Cures of the financial Crisis by Ira Robbin The upside of a Downturn by Krzysztof Ostaszewski Bubbles, Cycles and Insurers’ ERM – what Just Happened? by Paul Kneuer our Titanic Crisis: An Economic Rescue plan by Bertram Horowitz It is Time to Decide what kind of Crises we want to Have in the future by Ioannis Chatzivasiloglou and Charalampos Fytros The financial Crises: A Ripple Effect of Incentivised Disorder by Paul Conlin pRudEnT EnTERpRIsE RIsk ManagEMEnT sTRaTEgIEs some discussion of the crisis has reflected on the perceived failure of enterprise risk management. the consensus of these authors is that the crisis is not a result of a failure of the enterprise risk management process, per se, but rather a failure to implement enterprise risk management processes at all. these authors point out that the key to prudent enterprise risk management is in the enactment of a corporate culture that aligns desired performance with incentives and the matching of the authority to make decisions with accountability for the decisions made. currently, most incentive compensation rewards returns, but without reflecting the risk undertaken to achieve the returns. Culture • • • • • • • your Mother Should know by David Ingram Mixed Risk Management Strategies – Diversification That you Can Count on by Wendy Yu Derivatives, AIg and the future of Enterprise Risk Management by Michael G. Wacek The Democratization of Risk Management by Michael C. Schmitz and Susan J. Forray Should you Have a Chief Skeptical officer? by Max Rudolph lessons from the financial Crisis for Directors and CEos of Insurance Companies© by Jean-Pierre Berliet Against the grain: The wisdom of Countercyclical Capital by Anson J. Glacy, Jr. aligning Incentives • • • Reaffirming your Company’s Commitment to ERM in light of the financial Crisis by Prakash Shimpi A Tale of Improperly placed Incentives by Sam Gutterman Risk Management: The Current financial Crisis, lessons learned and future Implications by Neil Bodoff Managing/Matching authority and accountability • • • • Recent Crisis: Roots and lessons by Stephen Mitchell and John F. McGarry There Is No free lunch by Daniel C.F. Hui Risk Management – Buyer Beware! by Dennis Barry Risk Management and the financial Crisis: why weren’t we protected? by Mike Batty Managing Liquidity • Modeling and Managing liquidity Risk by Gary G. Venter 3
Slide 4: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications soCIETaL ThEMEs these essays center on the behavior of society and provide interesting and thought-provoking commentary on the interconnectivity of individuals’ behavior with the resultant behavior and roles of the markets, regulation and the inherent responsibilities to society. Given the existing limited liability laws, the role of regulation, and imperfect (and at times) “wild” interconnected and correlated markets, how and where does society put the costs and responsibilities of such a crisis? Limited Liability Laws • An Ideal Crisis by Shane Whelan Markets: Regulation Of, efficiency Of, Interconnectivity Of • • • • Transparency and liability Valuation by Philip E. Heckman Creating an Exchange for Insurance Contracts by Oakley E. (Lee) Van Slyke The game of ‘pass the Risk’: Then and Now by Joy A. Schwartzman from liquidity Crisis to Correlation Crisis, and the Need for ‘Quanls’ in ERM by Stéphane Loisel Behavioral/Personal economics • • • Essay on the financial Crisis by Andrew Winkler The Human Touch underlying the Current financial Crisis by Vivek Gupta If It looks Too good to Be True ... by Steve Malerich EFFECTIvE RIsk ModELIng These essays cover the reliance and overreliance on models. Models that tend to break down under ‘crisis mode’ and society’s reactions are discussed. efficient models must reflect correlation and the domino effect that can occur when bad things happen. • • Credit Crisis lessons for Modelers by Parr Schoolman what Is a Robust level of Risk Capital? by Larry Rubin and Xiaokai (Victor) Shi InsIghTs FRoM ThE InsuRanCE IndusTRy the actuarial profession has served as the enterprise risk management backbone of the insurance industry over the centuries. the authors of these essays share their insight on what has worked and what has not in the insurance industry and suggest how these learnings could be applied to other industry sectors in terms of analytical tools, regulation, contractual obligations and prudent risk trading schemes. • • • • Seven Simple lessons from Two Actuaries on the Mortgage Crisis by Randy Roth and John Pierce An Industry in Question, a profession with Answers by James Ramenda The financial Crisis: An Actuary’s View by Louise Francis Actuaries would Have Made a Difference by W. James MacGinnitie 4
Slide 5: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications explanations, causes, cures: the causes and cures of the financial crisis by iRA Robbin the upside of a downturn By kRzySzTof oSTASzEwSkI Bubbles, Cycles and Insurers’ ERM – what Just Happened? By pAul kNEuER our titanic crisis: An economic Rescue plan By BERTRAM HoRowITz It is Time to Decide what kind of Crises we want to Have in the future By IoANNIS CHATzIVASIloglou AND CHARAlAMpoS fyTRoS the financial crises: A Ripple effect of incentivised disorder by pAul conlin 5
Slide 6: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Causes and Cures of the Financial Crisis by Ira Robbin Why are we in a financial crisis and how do we get out of it? fidence in balance sheet valuations because too many assets are overstated, too many liabilities are understated, and The “why” can be simply explained: there is little congovernment bankers have instilled enough confidence in the system that people have, by and large, not felt the need tresses. to withdraw their funds and hide their savings under matCentral banks have made efforts to ensure liquidity, too much information is hidden. The crisis has spread due to a systematic failure of the regulatory system. Over the last 20 years regulations that fostered market stability were propagate without any real oversight. eliminated, and new financial instruments were allowed to The history of markets is one of booms and busts. The and they have applied doses of monetary stimulus. They have reduced interest rates and pumped money into the system. However, these stimuli have not yet proved effective at reversing the downturn. Why not? The problem is twofold. On the one hand, even with money readily available borrowers, and more and more borrowers are becoming questionable each day. On the other hand, overextended consumers are not clamoring to borrow money. They are frightened as their 401ks plummet and the equity in their homes shrinks toward zero. The financial crisis has sparked volatility of the cycles is magnified by leverage and tempered by transparency. The development of new financial instruments set the stage for this crisis because they were effective nitude of risk in the system. A telling symptom of the crisis is amount of risk their firms were exposed to: they did not know they were placing large bets in the financial casino. at pumping up the amount of leverage and masking the magthat leaders of many institutions claim to be surprised at the at low rates, bankers are hesitant to lend to questionable a general recession in the larger economy. Until demand finance expansion of plant and equipment. To summarize, monetary stimuli alone are insufficient to revive demand. How do we get out of this crisis? If our diagnosis of recovers, firms in many sectors have little need to borrow to been made and to understand which have worked and which have not. First consider the bailouts. These have It is important to disentangle the initiatives that have been proffered to a select group of financial institutions of the world financial system. The cost of bailouts has been “why” is correct, and if our assessment of measures undertaken to date is accurate, then it becomes clear that a solution to our economic woes must be focused on two ma- whose collapse was feared to imperil the overall workings enormous and threatens to grow even larger. The arbitrary way bailouts have been implemented in the financial sector presages a possible expansion of bailouts to many sectors jor objectives. First, all reasonable measures must be taken to stabilize and restore demand. Fiscal stimulus ought to be applied vigorously to do this. The federal government should send money to state and local governments in order to keep police, firefighters, schoolteachers and librarians in their jobs. It should increase the size of the armed forces. It of the economy, with political pull and not financial efficacy being the ultimate determinant of who gets bailed out of arbitrary bailouts of financial firms has not been effecsome firms from imminent collapse, it has not pulled the economy out of the larger crisis. and who does not. Despite tremendous cost, the program tive. While it has forestalled immediate crises and saved should provide seed money to finance an accelerated schedule of highway and bridge construction, port improvements auto manufacturers and other industrial firms that employ and alternative energy investments. It should loan money to large numbers of people. Unemployment insurance should have made to stop runs on the banks. By extending insurance What have been successful are efforts central bankers be extended even further. Anything that has a multiplier effect that will foster demand and keep unemployment down should be considered. for bank deposits before a general panic could commence, 6
Slide 7: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Causes and Cures of the Financial Crisis by Ira Robbin conservative, revaluation of all balance sheets and to impose strong capital requirements on financial institutions. The second major objective is force an accurate, if not into a New Company that writes new business and an Old Company that is liquidated. The liquidation and guaranty fund approach provides To do this will likely cause many large firms to fail. But that are sure of the net worth of others in the market. is what is needed. Credit will begin to flow once all players The federal government should stop bailing out financial a way out of the crisis. The government should seize weak financial companies and liquidate them. It should act as obligations, paying them off at 50 percent or some other set a partial guarantor of some of their financial instrument rate. The choice of which instruments should be partially honored needs to be thought through. Overall, instead of investing in AIG, lending money to AIG, and paying off its credit default swaps at 100 percent, the government should firms. That is throwing good money after bad. It should definitely not be taking an equity stake in them. This confuses the market about the net worth of the firms: are they implicitly backed by the government? It also undermines the value of other financial firms that do not have government backing. Part of the process of ensuring adequate valuations is put padlocks on its doors, liquidate it, pay off regular insurance contracts according to existing state guaranty fund rules, and guarantee to make good on 50 percent of its financial insurance obligations. This could be coordinated with foreign governments so policyholders and counterparpainful but not fatal fallout. to impose stringent regulations and capital requirements on whole classes of new financial instruments. Any recent financial mechanism that appears to mask risk or increase ties the world over would be treated to the same degree of For another example, Fannie Mae and Freddie Mac leverage should be subject to such treatment. In effect, all the leverage and hidden risk needs to be unwound, before we can reach the floor and start the way back up on a sound and sustainable basis. The philosophy inherent in the regulation of property should be split into old and new companies. The old ones should be liquidated, and the new ones should be forced to operate under stringent lending rules. The same approaches can be used all the way through the nested chains of tranches It will be very costly, but, in the end, it will cost far less than trying to revive a select few of the comatose and pay off 100 percent of their ill-considered financial obligations. and derivative instruments that wind through the economy. and casualty insurance companies provides an interesting paradigm for how a wholesale revaluation could be accomplished without mortally wounding the whole economy. When an insurance company has inadequate capital, it is subject to seizure by state regulators even though it is techfrom writing any more business and then proceeds to liquidate it. This stops the company from trying to raise cash by writing a boatload of underpriced business. Meanwhile the claimants are not left with worthless paper; instead they nically not bankrupt. The state authority stops the company of the same. Instead of bailing out weak financial firms, we should be liquidating them. All doubtful assets need to be written-downs; the sooner the better. We need accurate effort to clean up our accounting system. But it needs to stop being an investor propping up those that should be in In conclusion, what is being called for here is not more and transparent accounting. Government can help in this are partly compensated by guaranty funds. These funds are variant of this idea is when the existing company is split partially replenished by recoveries from the liquidation. A the morgue. It needs to conservatively regulate all financial instruments. It should foster liquidity and stoke demand. That is what needs to be done to get out of this crisis. Ira Robbin is senior vice-president and chief pricing actuary for the United States and London at Endurance Worldwide Insurance in New York, N.Y. He can be reached at irobbin@enhinsurance.com. 7
Slide 8: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Upside of a Downturn by Krzysztof Ostaszewski The financial crisis that began in 2007 and accelerated greatly in 2008 has posed a unique challenge for the regulators of financial intermediaries. The speed and severity financial and political system in the United States, and subsequently, worldwide. authorities, who, ostensibly, do not regulate insurance firms. At the same time, an unregulated investment bank, Lehman Brothers, was allowed to fail. I should add, however, that the phrase “allowed to fail” seems quite inappropriate, as save Lehman Brothers. of the events that transpired have been quite a shock to the there was no federal or state authority whose job it was to The federal government’s response to the financial and And, it has brought to the forefront the following key quesinstitution? This crisis has posed a unique challenge for regulators. tion: What should regulators do with an impaired financial This crisis is the first crisis of the new financial ser- economic downturn has consisted of several phases: in early 2008. 1. Economic stimulus checks sent to American taxpayers 2. The initial “bailout” proposal: to buy currently nonperforming mortgage-based assets from banks and financial institutions, touted very loudly by many in vices industry, an industry created by the Gramm-LeachBliley Financial Services Modernization Act of 1999, which opened up competition among banks, securities compa- nies and insurance companies. Yet while the industry has the investment industry as a “great trade,” on which the government would eventually make trillions, by spending approximately $700 billion now. become integrated, its regulation has remained fractioned. the Federal Reserve and FDIC, while insurance companies face state insurance regulators; exchanges are supervised by as hedge funds remained largely unregulated. In response to a systemic crisis, the U.S. government has worked on When banks are impaired, their problems are addressed by 3. Capital injection for banks in the form of purchase of partial ownership in banks by the federal government. 4. Assuming full control in some failed banks, nota- SEC and CFTC; and independent investment banks as well bly AIG Insurance, in order to transfer their full or financial institution. crafting an integrated response, and has crossed some of the boundaries between separate parts of the industry. But those boundaries had been breached by the industry itself quite some time ago. partial ownership to some other, better performing As I am writing these words, it appears quite clear that the first step was ineffective in stemming this financial storm. Also, the secretary of the treasury has just Troubled Assets Relief Program). Only steps 3 and 4 are announced abandonment of the second step (known as being implemented. They are similar, yet subtly different. They both involve a transfer of ownership of an impaired financial institution to the regulators, but step 4 causes consequences. As we know from the careful reading of the ture of a firm has no effect on the value of the firm, unless it that ownership to be transferred fully. This has significant Modigliani-Miller Theorem, a change in the capital strucaffects the productive capacity of the firm, its tax expens- firms under their supervision becoming impaired. The FDIC has consistently pursued the strategy of taking over arranging a purchase by another bank, without any interbanks found to be too weak to continue functioning and ruption in services provided to customers. The Federal Reserve has offered unprecedented liquidity, expanding it to all financial institutions and even non-financial firms. State insurance regulators remained mostly uninvolved, bewas addressed in a very unorthodox fashion by federal cause the only major impairment of an insurance firm, AIG, Different regulators have had different responses to es, its bankruptcy cost, or the agency cost of that capital 8
Slide 9: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Upside of a Downturn by Krzysztof Ostaszewski structure. The government is attempting to lower the cost of bankruptcy by lowering its probability. But other, maybe be considered: unintended, consequences of government actions ought to 1. As the permanent income hypothesis tells us, a one- the economy: • A financial intermediary performs two key functions in Uses funds obtained from clients to purchase capital time grant of money from the government, provided in step 1, was unlikely to cause any permanent change in the behavior of all economic decision makers. Given decision makers acted wisely by ignoring stimulus by not sending the checks in the first place. derivative securities. Cash flows of an intermediary’s assets are used to make payments on liabilities issued to customers. Customers’ deposits or insurance poli- assets. This activity is, effectively, equivalent to writing the likely long-term damage of the crisis, economic checks. The government could have been even wiser 2. The question of why financial firms are so uninterested would venture the hypothesis that a firm known to the • cies are, effectively, derivative securities created out of the firm’s assets. This activity, often misrepresented mathematically and virtually never taught this way in business schools. Processes payments for customers. This function used to be simple and mostly banking-like. It has become more complicated with the advent of private transacswaps of all types, that are done outside of the regulated banking system. tions that also amount to payment processing, especially as spread business, is very risky, very complex in selling their troubled assets is quite fascinating. I market as selling their troubled assets is automatically state of the markets, that is a very uncomfortable position in which to be. a target of negative rumors, and, in the current fragile 3. Providing funds for banks without any conditions on restructuring and improving their profitability creates rather wicked incentives. If the bank is reasonably stable loses capital, and may need to be taken over by regulators. But failure is the firm’s own problem. Failure at the second function means that the firm’s customers are unable to The first function is speculative. If a firm fails at it, it after capital injection, but not yet strong, the best course improve its own profitability with the rival’s profits. The result is that a better-managed company is acquired by a company managed badly, and good of action is to purchase a better performing rival, and pay their bills, resulting in a systemic economic crisis. We live in a world in which those two functions are automatically combined. Yet it is the first function, the risky one, that creates the most profits, especially with an assumption without which economic activity stops. managers are let go. Bad managers have their jobs is really on the brink. In this situation the smartest strategy for the managers is to pay themselves large bonuses before the inevitable end happens. And the government conveniently provided the funds. of additional leverage. Failure impairs the second function, We do not need to sacrifice the entire government bud- saved by taxpayers. Things are even worse if the bank get, and the country’s economy, to save badly managed that their customers can pay their bills. If a financial inter- financial intermediaries. We merely need to make certain mediary fails, its managers who did not know how to man- resolution of a situation of an impaired financial intermediary: the takeover by the regulators. Or is this really such an unpleasant resolution? This brings us to the fourth, least pleasant, it seems, age the derivative securities portfolio they created lose their jobs. If the regulators assure continuity of payment processtirely desirable and a positive outcome. When a firm fails in ing functions, such failure can and should be viewed as en- 9
Slide 10: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Upside of a Downturn by Krzysztof Ostaszewski the free market system, its employees and its resources can be utilized productively by other firms, but its managers proved themselves to be incapable. Keeping them in place means that their management policies will continue and the irresponsible behavior of politicians who envisioned grant- Granted, this current crisis has its roots in an utterly ing credit to everyone and pushed hard for it. We cannot remove or punish those politicians through economic mechaallow the return of competence in the financial industry, if we are to have one. nisms. The whole country is punished instead. But we must impairment they brought about will become even bigger. The upside of the downturn is that these powerful, influential, connected and important, yet incompetent people can be removed from their positions. To quote a great insight of Ayn Rand: “There is no substitute for competence.” Not even $700 billion of taxpayers’ money will do. Krzysztof Ostaszewski, Ph.D., FSA, CFA, MAAA, is professor of mathematics and actuarial program director at Illinois State University in Normal, Ill. 10
Slide 11: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Bubbles, Cycles and Insurers’ eRM – What Just Happened? by Paul Kneuer Interest rates rose from 2004 through 2007, so adjustable After home prices topped out in 2006, speculators and some buyers who had overstretched found it impossible to refinance. This caused mortgage insurance claims. Collateralized debt obligations (CDOs) rapidly lost particular, but also to other failed firms. Falling stock and bond prices also caused insurers unrealized capital losses. At the same time, Lehman’s and others’ failures and mortgage rates increased and many borrowers fell behind. falling prices for CDOs caused losses on credit default swaps. The losses and collateral calls impaired AIG’s capirequiring federal help. value. Under mark-to-market accounting, many insurers must immediately reflect the lower values. Some CDOs prices fell. Others were backed by credit default swaps. were backed by guaranty insurers, who took large losses as CDO pools had not been recorded on the balance tal, leaving them little time or flexibility to raise more, and The write-downs at U.S. public companies are now ap$1 trillion is the worldwide annual P&C insurance premium, or the reinsurance industry’s total assets. $1 trillion averages to about a month’s income per U.S. worker. This is equivalent to each of us delaying retirement by a month or two. • proximately $1 trillion. In context: • • sheets of the sponsoring banks, as they thought the credit risk had been passed to others. However, the banks often retained liquidity commitments, and when these were drawn down, they were required to consolidate the CDOs onto their balance sheets. This increased their assets and reported leverage, and further stressed their capital levels. Higher leverage contributed to a fatal bank run at Bear Stearns on March 17. FNMA (Fannie Mae) and FHLMC (Freddie Mac) were chartered by Congress to deepen the The 1990 S&L crisis cost over $150 billion and represented 3 percent of GDP. In today’s $14 trillion economy, $1 trillion is relatively bigger, at about 7 percent. A $600 billion problem in mortgages alone would be comparable market in residential mortgages. They also acted as hedge to the S&L’s and is about 4 percent. 1990 followed the 1987 crash, Latin debt problems of the ’80s and two oil shocks. We are better prepared now than we were then. What Have We Learned? Insurers may want to study others’ missteps. A common observation is that institutions are moving toward a “supermarket” approach. Bear Stearns, Lehman and Merrill funds with large positions in CDOs, which wiped out their on September 7. Many insurers held significant positions and took large capital losses. capital bases. They were put under federal conservatorship affecting all major banks and many insurers, especially ones who sold annuities with guaranteed benefits. No institution trusted another’s balance sheet, and so bank investments By the third quarter, credit problems were seen as Lynch didn’t have both commercial and investment banking, while Citigroup did and survived; Bank of America was the rescuer of Merrill; and Goldman Sachs and Morgan observation may be wrong. became increasingly expensive. This made it impossible to survive other bank runs and caused the failure or forced Mutual (September 26), Fortis (September 28), Wachovia (September 29) and many others. These failures cost com- Stanley reorganized as commercial banks. The common Diversification is a strength in times of stress, but sales of Lehman Brothers (September 15), Washington another lesson is that commercial banks and insurers are more tightly regulated than investment banks. The insurers who had direct losses (AIG, Swiss Re and XL) sustained them in operations they saw as diversifications. Their core mercial paper investors large losses, and dried up that market and the money market funds, which invested heavily in it. Insurers have large exposure to Lehman and Fannie Mae in 11
Slide 12: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Bubbles, Cycles and Insurers’ eRM – What Just Happened? by Paul Kneuer businesses kept strength and value. While successful diversification reduces risk by reducing correlation, it is hard to know what areas are correlated when. In times of great stress, new correlations appear. Diversification brings in isn’t diversification. new operational and execution risks. Vertical integration The banks that failed tended to have higher leverage as falling prices, not rising. Bubbles are a concern to any challenge to ERM. trading firm, including banks and insurers, and a particular The key attributes of ERM are that firms should: • • • • Identify risks in all areas at once, Consider how risks can affect more than one area, Use consistent measures so that different risks and different operations can be compared and Use these measures in decision making. Bubbles end much faster than they inflate, after a trading ratios, and thus, less flexibility. While banks have higher leverage than insurers (as high as 30- or 40-to-1, measured as assets to equity), leverage still matters for P&C insurers. AIG was over 13-to-1 in January; 5-to-1 is a more typical ratio. Insurers also take leveraged positions on the amount of coverage they provide. Line limits and aggregate accumulations are two ways to look at this. Investment banks break directs the market’s attention away from trends and prices. ERM innovations did not manage this bubble well. back to fundamental value, which changes much less than Price and volatility data before and during a bubble do not had CDO positions that were multiples of their capital. AIG did the same with credit swaps ($450 billion in limits on $70 billion of capital, over 6x). Leverage (in assets or premiums) can quickly turn from a financial advantage to a survival threat, even if it has been built up gradually. stochastic estimates of impairments. Objective measures of leverage are at least as important as Small premiums can bring big risks. Insurers need to reflect levels after the “pop,” so economic capital models the way up. The Basel and Solvency II standards rely too much on firms’ internal assessments. are inaccurate. Mark-to-market rules accelerate bubbles on assessments are no better. Rating agencies and modelers are But inflated views of values are widely held, so external see and understand the potential downside on all of the risks that they take, even in small operations. AIG never cost them their company. Bubbles and eRM We just saw classic, but particularly severe, bubbles in diligent and objective, but they are also susceptible to bubbles. CDOs were highly rated before their collapse. Ratings and models evolve as their authors learn from events, so they do not mean the same things at different dates. Cycle Management The insurance market has similar cycles driven by delays in recognizing results. Cycles cause a bias in how individual insurers, their reinsurers and the market measure their costs, exposures and risk. had as much as 1 percent of its revenue from swaps, yet it home prices and debt securities. During a bubble, the buyers, sellers, investors and lenders of an asset class all develop actions. Other recent bubbles involved technology stocks an elevated view of values, based on observing each others’ in the 1990s and commercial real estate in the 1980s. The ties (most recently oil) have also had bubbles, as have had include the potential phase of the underwriting cycle (that is, its level). The phase of the cycle can only be known after retentions and catastrophes. The charge for cycle risk will be it turns, so the phase is unknown, just like trends, renewal An economic capital model for insurance risks should debts of several emerging markets and various commodiseveral insurance products. In insurance, however, the costs and risks are born by sellers not buyers, so bubbles are seen 12 the derivative of a risk measure with respect to the unknown
Slide 13: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Bubbles, Cycles and Insurers’ eRM – What Just Happened? by Paul Kneuer phase variable. (It will be sinusoidal because a function with f’ = -f is in the sine family.) The charge reflects: • • • (bigger bias = more risk), Amplitude of the bias that is expected for a product Time since the last trough (longer = more risk) and Perceived height of the current cycle (the better you think things are, the worse they can really be.). In a simpler approach, without capital modeling, insur- times over, even though a rational buyer might look through In derivative products, that fall can be magnified many to the underlying mortgage and see a higher value. Since financial institutions trade across boundaries, and trades can be multiples of firms’ capital, their results are closely linked. The same losses have actually been reported over again, as “spirals” of mortgage insurance claims, CDO write-downs, guarantees of CDOs, falls in the price of bonds, preferred stock and debt, credit default swaps and bankruptcies. A lot of the current problems, perhaps the by mark-to-market accounting. dominant share, and the speed of the decline, were caused While having collateral from your trading partners is ers should use line or position limits to control leverage, grow less than simulated risk models suggest is prudent, underwriting performance, as well as the net. Regulatory and accounting Challenges With commitments now of $150 billion to AIG, the outgoing administration, many legislators and some trade groups have called for federal regulation of insurance and derivatives. We expect increased regulation and solvency oversight from all aspects of government. expect good results to “slip” back and manage their gross always good, getting extra collateral later adds instability to the entire system. This is required in the EU for banks under Basel and for insurers under Solvency II, and is inintended hazard should be discussed by policymakers. cluded in the NAIC reinsurance security proposal. This un- a result, when a firm sells an underperforming asset, that single point has an immediate effect on others’ statements. In a time of distress, price information is limited. As Paul Kneuer is senior vice president and chief reinsurance strategist at Holborn Corporation in New York, N.Y. He can be reached at paulk@holborn.com. 13
Slide 14: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Our Titanic Crisis: an economic Rescue Plan by Bertram Horowitz, Inc. Our current financial crisis has wreaked havoc on the credit evaluating a homeowner’s eligibility for a refinance, markets, the stock market and the entire economy. We must remember that the root cause of this crisis is the collapse of the residential housing market, which has been further exacerbated by the continuing downward housing spiral. Most of the $700 billion in TARP rescue funds have been deployed as capital infusions into banks and financial institutions in an attempt to address our economic crisis at the the economy. A comprehensive solution must directly address the residential housing market catastrophe. institutional and trading level. This alone will not rescue vide mortgage insurance to the banks for primary residence modification or new mortgage. The FHMIC would pro- mortgage loans that meet modified traditional mortgage underwriting standards. Traditional mortgage underwriting standards include: no more than 80 percent mortgage devoted to housing expenses; full income verification; and good credit history. loan to appraised value; no more than 28 percent of income be required to pledge additional assets such that the pledged least, 5 percent greater than the amount to be refinanced. In return, the homeowner’s monthly payment is reduced to as low as the original monthly payment. Upon meeting ing institution would receive FHMIC mortgage insurance, FHMIC would require that the distressed homeowner establish “Federal Home Mortgage Insurance Company” A solution to the financial crisis would be the creation of a Federal Home Mortgage Insurance Company (FHMIC). Just as the FDIC was created to encourage and safeguard assets plus the current assessed value of the home is, at individual bank deposits, the FHMIC would retain the private banking mechanism for home mortgages and insure bank mortgages to homeowners. Banks and lending institutions would then have an incentive to refinance FHMIC mortgage insurance underwriting criteria, the lendwhich guarantees against credit default risk. Mortgage insurance premium is passed through to the homeowner and is paid in the future (e.g., when the home is sold). FHMIC would be the underlying insurance mechanism employed to address the residential housing crisis. and issue mortgages since the FHMIC would insure mortgages against default. FHMIC would be a nonprofit insurer designed to serve as a catalyst to stimulate and supnancially distressed homeowners refinance or modify their port the residential housing market, which would: help fimortgages and stay in their homes; provide incentive for gradual rise in prices in the housing market. The FHMIC sential to the rescue of the economy. Dire Case: Home Underwater, skyrocketed Monthly Payments Most difficult to address is the case of a homeowner whose house is “underwater” (i.e., worth less than the outstanding mortgage) and where the homeowner’s monthly payment has skyrocketed. FHMIC would facilitate efforts for such the general principle of traditional underwriting standards. Consider the following representative cases: homeowners to keep their home while largely retaining lending institutions to grant new loans; and facilitate the could be structured to be self-supporting and would be es- FHMIC guarantees against Mortgage Credit Default Risk The root of the housing crisis is that mortgages were granted without regard to traditional mortgage underwriting standards, and this resulted in the current subprime mortgage crisis. It is imperative that lenders return to more traditional mortgage underwriting criteria when 1. 2005—homeowner purchases house in Stockton, Calif., for $500,000 and receives an interest-only mortgage of $500,000 at an initial “teaser” ARM interest rate of 4 percent resulting in a monthly mortgage payment of $1,667. 14
Slide 15: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Our Titanic Crisis: an economic Rescue Plan by Bertram Horowitz, Inc. 2. 2008—interest has adjusted upward to 8 percent, resulting in a monthly mortgage payment of $3,333 and the current appraised value of house has fallen to $300,000. The house is now $200,000 “underwater.” the $300,000 appraised value plus the $225,000 pledged assets equal $525,000 and this is, at least, 5 percent greater than the $500,000 outstanding mortgage balance, the modified FHMIC loan-to-value criterion has been met. 7. Assuming the homeowner meets all FHMIC under writing criteria, the FHMIC grants mortgage insurance to the lending institution, and the homeowner is offered eowner can now afford. 3. Homeowner faces loss of home with the original lending institution left holding a non-performing mortgage in default. There is great temptation for the homeowner sure and/or bankruptcy. As this sort of dire situation to simply walk away from his house and face foreclohas played itself out over millions of homes, it is magnified upward through the banking system, our credit and stock markets, and through the entire economy. FHMIC Facilitates Solution… Instead of suffering the burden of home loss (#3), FHMIC provides an insurance mechanism to progress toward a solution: 4. Lending institutions and service providers are required to contact homeowner and advise eligibility to refinance or modify his current mortgage with the same or another lending institution. Qualifying criteria and a availability of counseling services from the FHMIC regarding refinance options. 5. Homeowner applies for refinance or modification. In order to meet the loan-to-value requirements, the homnon-borrowed funds totaling, at least, $225,000 eowner must pledge, as collateral, additional assets in (special legislation would be enacted to also permit assets from retirement plans and anticipated inheritance to be held in trust until the value of the home increases) current appraised value of the home. new application form would be included as well as a new $500,000 mortgage loan under terms the hom8. The homeowners accept the new terms and the original $500,000 loan is satisfied. The homeowners’ monthly payment is returned to $1,667 computed at the original 4 percent interest rate. Assuming the prevailing interest payment of $417 (and interest on lost interest) is added rate is 5 percent, the imputed difference in the monthly to the outstanding principal. When the loan is satis- fied, the homeowner is required to pay the bank’s morton the mortgage. At any point throughout the loan, the homeowner could pay off a portion of the loan, gage insurance premium for obtaining FHMIC insurance which would lower their final mortgage insurance prices rise, pledged collateral may be released back to the homeowner as the loan-to-assessed value ratio decreases. premium as well as the outstanding principal. As housing a Comprehensive Plan: Fairness and Benefits to all Parties This plan is intended to balance equities between all parties. It is worthwhile to consider the benefits to various parties from each of their perspectives: Distressed homeowner—retains ownership of home and in addition to the $300,000 in collateral from the 6. Homeowner’s application is reviewed to determine whether it qualifies for refinance or modification of current mortgage using modified traditional under stays in community; monthly payments returned to original low monthly payment; retains ownership of pledged assets for benefits received at time of loan satisfaction. that may be released back to homeowner at later date; pays writing standard criteria as required by FHMIC. Since 15
Slide 16: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Our Titanic Crisis: an economic Rescue Plan by Bertram Horowitz, Inc. market at rock-bottom foreclosure prices—decline in home Secure homeowner—fewer homes dumped onto the Why This Plan? Unlike typical insurance, a unique characteristic of FHMIC insurance is that it actually acts to prevent losses. We are only at the beginning of a looming foreclosure crisis. The financial markets have, in effect, responded to a projection of future foreclosures on a massive scale. The proposed values reversed because of decreased supply; recognizes “for free.” that distressed homeowner neighbor is not being bailed out Prospective homeowner—the existence of FHMIC mortgage insurance encourages lending institutions to grant low interest mortgages to new home purchasers. FHMIC intervenes and significantly reduces the otherwise anticipated foreclosure catastrophe. Addressing individual residential homeowners’ distressed mortgages becomes banks, the financial institutions and finally through our MIC mortgage insurance; investment is largely protected is recommended so the lending institution assumes some risk); roughly 5 percent rate of return on largely government-backed mortgage assets. Taxpayer—FHMIC is a sound, non-profit, self-sup- Lending institution—incentive to lend because of FH- the engine that propels the rescue solution upward to the entire economy here and, potentially, abroad. The continhomeowners would remain in their homes with affordable is effectively mitigated; housing values would rise and by FHMIC mortgage insurance (a 10 percent deductible ued downward economic spiral would be reversed: more loans; financial markets would recognize that the problem the market value of mortgage-backed securities would increase; lending institutions would unlock credit; balance sheets would be cleaned up; and employment would rise. The contagion spread by the subprime crisis would be reversed as the solution spread through the economy. FHMIC would provide a catalyst for banks and lending porting insurer with strict underwriting standards; FHMIC could be initially capitalized with some of the TARP funds MIC as housing analog to FDIC for bank deposit insurance; overall economic health of country enhanced. designated to address potential foreclosures; perceives FH- surance premium is commensurate with the risk: the greater As with all properly priced insurance policies, the in- the risk of default, the higher the mortgage insurance premium. In addition to mortgage insurance premium, FHMIC could marshal additional assets from bailout funds as well as other parties who may be responsible for the subprime institutions to rescue sinking homeowners and avert catawatching our economic Titanic sink and a large segment of strophic consequences. We have a choice. We can settle for its homeowner passengers perish while we concentrate on saving financial institutions…OR… we can intervene in a meaningful way and provide incentive to the lending institutions to throw the sinking homeowners a lifeline and save both the financial institutions and the individ- crisis including appraisers, credit rating agencies, fraudubrokers. lent applicants, predatory lenders and predatory mortgage This plan would not prevent all foreclosures as many homeowners would still be unable to meet even modified underwriting mortgage standards, and FHMIC would deem such homeowners’ mortgages to be “uninsurable.” Lending ual residential homeowners. In the process, we will have stabilized and improved the entire economy through “trickle-up economics.” institutions would still be encouraged to directly renegotiate mortgage workout terms with homeowners who not qualify for FHMIC mortgage insurance. This essay offers an opinion regarding a proposed solution to our current financial crisis and is intended only as a general discussion. Bertram Horowitz, Inc. Actuarial and Risk Consultants, would be pleased to provide additional details. 16
Slide 17: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Our Titanic Crisis: an economic Rescue Plan by Bertram Horowitz, Inc. Bertram Horowitz, president of Bertram Horowitz, Inc., Actuarial and Risk Consultants, has 30 years of experience tive solutions to insurance and regulatory financial crises. in insurance regulation and has been involved in innovaAs a financial actuary with a regulatory specialty, Bertram of the economic consequences of government policy and the development of innovative solutions. Mr. Horowitz is the former special deputy superintendent and financial actuary for the New York State Insurance Department. Bertram Horowitz, Inc., Actuarial and Risk Consultants, was founded in 1986 and specializes in assisting governments with the administration of their laws and policies. Horowitz’s career has been largely devoted to appropriate implementation of government regulations, consideration Bertram Horowitz is at Bertram Horowitz, Inc., in New York, N.Y. He can be reached at bert@bhiactuaries.com. 17
Slide 18: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications It is Time to Decide What kind of Crises We Want to Have in the Future by Ioannis Chatzivasiloglou and Charalampos Fytros Crises are devastating. They leave behind a high amount of entropy (unemployment, poverty, lack of safety). But this is half the truth. The other half says that crises are to Fund transferability (at the one bottom of the pyramid) is defined as the ability of funds to be transferred from one specific subsidiary of a group to another subsidiary of the some extent good, because they provide a natural crash test to discover current practices’ risks and limitations, and in response give rise to more efficient regulation, to financial society, as the old and corrupted die and new ideas emerge and give us the opportunity to reflect upon the future we want to have. innovations and to the wakening of the vitalizing powers of same group which is in financial difficulty. The higher the ability to transfer funds, the higher the probability a potenof fund transferability implicitly defines the frequency for a micro-crisis to emerge. tial crisis inside a group never emerges. Therefore, the level have collapsed, it is time to reflect and decide what kind of having crises in the future, unfortunately, is unavailable). Now that the basic structures of the financial system benefits at the level of a group are recognized, the overall (i.e., at the group level) capital available to fund potential subsidiaries of the group. The lower the perceived correlation between the subsidiaries, the higher the diversificamicro-crises is lower than the sum of the capital of the On the other bottom of the pyramid, if diversification crises we really want to have in the future (the option of not In order for our decision to be effective, it should be inforstructure of the current crisis. mative. This is why we should elaborate more on the basic The three factors that this paper distinguishes to sup- tion benefits recognized by the regulator and the lower the total capital of the group—which essentially means lower of diversification benefits implicitly defines the severity of micro-crises. risk capital to absorb potential losses. Therefore, the level port the genesis and the proliferation of crises can be depicted as the three angles of a pyramid: of within-group diversification benefits (severity) make up the bottom line of the pyramid, i.e., the micro-level. Both fund transferability (frequency) and recognition RebAlAncinG on top of the pyramid. Why is such a term used? Consider an institutional or even an individual investor: consistent with his/her investment policy statement, s/he follows a strategic allocation is not heavily distorted by the drifting possible investor—the government. Does this exceptional method of rebalancing so that the portfolio’s long-term values of the underlying assets. Now consider the largest Let’s make some definitions: company, or a group of companies (either banking group, insurance group or a conglomerate). A micro-crisis is restricted to the group and is local in nature. Let a micro-crisis be a crisis that is faced by a certain investor differ so radically? Governments spend, leverage, lio is society, and this portfolio should at minimum be balRecoGnition of diveRsificAtion benefits Within conGloMeRAtes Rebalancing is the term used to denote the macro-level, fund tRAnsfeRAbility sell and lately invest in distressed securities. Their portfoanced. So, rebalancing seems quite essential not only for case too. Letting big-company financial establishments be- nomic factors and consequently real economy. And let a macro-crisis be a crisis that affects macroeco- the average institutional or individual investor but in this come bigger is like letting your originally strategic alloca- 18
Slide 19: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications It is Time to Decide What kind of Crises We Want to Have in the Future by Ioannis Chatzivasiloglou and Charalampos Fytros tion drift away while undertaking a series of concentrated exposures that might affect the value of your whole portfolio—in this case, society itself. Financial institutions have find two antidiametrically placed massive bobs: “Fund transferability” and “Recognition of diversification benefits within conglomerates.” The higher their masses (when, for the potential to leverage profits or losses by using a void but powerful substance: it is called credit. Proper rebalancing, i.e., imposing a ceiling on the amount of credit a financial institution might use or invest upon, contributes to a longopt for such a long-term balanced strategy). example, lower fund transferability and higher diversification group benefits coexist), the higher inertia contributes more intense the crisis. “Rebalancing” determines friction to a highly uncontrolled swinging back and forth, i.e., the in the pendulum movement: the lower the ratio of credit to term financially balanced society (given that governments We talked about a pyramid. This pyramid is not static: real output (imposed by a consistent rebalancing strategy), resistance. the lower the size of financial risk: swinging finds a natural We may have a choice upon future crises: Micro Level it constantly moves like a pendulum. The nail is at the point of “Rebalancing”; at the bottom of the pyramid, you may Macro Level financial Risk fund transferability Regulator’s Recognition of Real diversification benefits within conglomerates high crisis no Rebalancing low Macro-level: uncontrolled Micro-level: low/High frequency, high/low severity overall: varying effects that do not exclude extremities no Rebalancing high/low high/low Macro-level: uncontrolled Micro-level: low/High frequency, high/low severity Overall: varying effects that do not exclude extremities no Rebalancing high low Macro-level: uncontrolled Micro-level: low frequency, low severity Overall: varying effects that do not exclude extremities. Managed Rebalancing high/low high/low Macro-level: controlled Micro-level: low/High frequency, high/low severity Overall: Managed Rebalancing high low Macro-level: controlled Micro-level: low frequency, low severity Overall: preferred Ioannis Chatzivasiloglou, FHAS, ASA, MAAA, is an actuary at the Greek Private Insurance Supervisory Committee (PISC). He is responsible for the actuarial supervision of the insurance companies operating in Greece. He can be reached at i.chatzivasiloglou@psic.gr or hadjivassiloglou@yahoo.com. Charalampos Fytros, FHAS, CFA, is a consulting actuary for Prudential Co. Ltd., an independent employee benefits consulting firm located in Athens, Greece. He can be reached at harrisfy@yahoo.gr or hfytros@prudential.gr. 19
Slide 20: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Financial Crises: a Ripple effect of Incentivised Disorder by Paul Conlin The current financial crisis is the result of the interplay of changing circumstances, since the 1970s, on Main Street, Wall Street and internationally. oping economies (first Japan, then Korea/Singapore/Hong Kong and finally China/India) adopted protectionist/capi- talist strategies similar to what the United States adopted in the 1800s. But unlike the American consumer, the cona savings- to a consumption-based lifestyle. Even in nowdeveloped Japan, consumers save a substantial portion of their income. So the large surpluses in the world continue to recycle back into the United States, including disproporsumer of these export-led economies never switched from cally. There’s the classic scene from “It’s a Wonderful Life” where the depositors of a local S&L demand their savings says, correctly, that he can’t give them their money back because he doesn’t have it—it’s invested in the homes of back from the bank manager, played by Jimmy Stewart. He On Main Street, savings and lending used to occur lo- their neighbors. In the world portrayed by the movie, lending standards were monitored by local lending officers in the communities—lending never got too loose, since the S&Ls would have to eat the losses; or too tight, since no loans would happen. And when blow-ups did occur, as in tionately (thanks to the implied, and after Sept. 8, 2008, the official, government backing of Fannie Mae and Freddie Mac) into American housing. The world of the local S&Ls, the partner-owned invest- Texas in the 1980s oil glut, they tended to be contained lo- ment bank and Bretton Woods isn’t coming back. So what second law of thermodynamics, which says that existence sage of time, since any work done to address the current it fixes. cally. But the S&L crisis of the 1980s destroyed the local savings institutions and ushered in the era of interstate fithe worldwide credit markets. nancial conglomerates that borrow funds, short term, from On Wall Street, the investment banks were partnerships. does that leave us to do next? Part of me is persuaded by the gets progressively more and more disordered with the pascrisis expends more energy and causes more disorder than On the other hand, we need to at least try something. When the Wall Street firms underwrote a deal or traded se- curities, it was the personal wealth of the partners on the line, so solvency was highly incented. During the 1970s, the partnership ownership structure constrained growth of the Wall Street firms and caused capital shortages. Donaldson Lufkin Jenrette filed for public ownership in 1969, the others to follow. With shareholders putting a premium firms increased leverage and risk. They had no margin of 2007. The lending of money by $500-per-year Chinese factory workers to Americans to buy $250,000 houses strikes me as the best place to start. A home mortgage should not be able to be securitized. If this makes it more difficult for Americans to borrow for homes, so be it—the true economic costs (and risks) of such loans must be reflected in mortgage rates, and the ripple effect must be felt in home prices. There is an fer risk to a reinsurer, but always remains on the hook if the reinsurer defaults. A mortgage loan must be a permanent and the stock market boom of 1982–2000 caused most of on earnings growth and return on equity, the Wall Street error for the wave of credit defaults, which were to begin in Finally, internationally, the lapse of the Bretton Woods insurance precedent for this—a primary insurer can trans- arrangement between the lender and borrower—if this is not acceptable to either, no problem: no deal. Legislation and regulation codifying such a (seemingly, by comparison to current rules) draconian regime must be enacted while the consequences of the alternative are fresh. monetary system in 1973 and free float of currencies allowed the United States, the highest currency on the food chain, to sustain massive budget and trade deficits. Devel- Paul M. Conlin, FSA, is an actuary at Aetna in Hartford, Conn. He can be reached at conlinp@aetna.com. 20
Slide 21: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications prudent enterprise Risk Management strategy: culture your Mother Should know by dAvid inGRAM Mixed Risk Management Strategies – Diversification That you Can Count on by Wendy yu derivatives, AiG and the future of enterprise Risk Management By MICHAEl g. wACEk The Democratization of Risk Management By MICHAEl C. SCHMITz AND SuSAN J. foRRAy Should you Have a Chief Skeptical officer? by MAx Rudolph lessons from the financial crisis for directors and ceos of insurance companies© By JEAN-pIERRE BERlIET Against the Grain: the Wisdom of countercyclical capital By ANSoN J. glACy, JR. 21
Slide 22: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Your Mother should know by David Ingram Something as massive as the current financial crisis is much number of consequences. First of all, it could well stop the idea of high speed trading of very, very complex risks. If these risks are too complex to evaluate fully during the intended holding period, then perhaps it would be better for all if the trading just did not happen so very quickly. In the very different risk analysis requirements for trading books of risks vs. their banking book of risks. The banking (credit too large to have one or two or even three simple drivers. Below is a discussion of three drivers that are often not at the top of lists about the origins of the crisis. And in all mistakes. three cases, my mother would have cautioned against those When I was 16, I had some fine arguments with my case of the recent subprime-related issues, banks often had mother about the girls that I was dating. My mother did not That was absolutely silly, I argued. I was years and years want me dating any girls that she did not want me to marry. away from getting married. That was a concern for another time. My mother knew that in those days, “shotgun marriages” were common, a sudden unexpected change that mostly) risks required intense due diligence or underwriting. The trading book only had to be run through models, where upon internal analysis. the assignment of assumptions was not required to be based My mother would often caution me against some triggered a long-term commitment. Even without getting a I started dating when I was 16. shotgun involved, five years later I got married to a girl that Well, there are two different approaches to risk that activity by saying, “What if everyone did that?” She did course for everyone. not allow any actions that were not sustainable as a general Well, an implicit assumption in the way that many firms in the risk-taking business use. One approach is to risks at will. The other approach is to assume that any risks the firms with the risk-trading approach would have listened assume that they can and will always be able to trade away will be held by the firm to maturity. If the risk managers of to their mothers, they would have treated those traded risks as if they might one day hold those risks until maturity. In most cases, the risk traders can easily offload their risks at will. Using that approach, they can exploit little bits of risk practitioners use financial models is that their planned activity is marginal to the market. If you ask the manager of a large mutual fund about that assumption, they will generally laugh out loud. They are well aware that their trades must they will build up a position over a period of time based upon the normal flow of trading in a security. That is a very the subprime mortgage market was a drastic shift in activity that was clearly not marginal. When the volume of subbe made carefully to avoid moving the market price. Often micro-example of non-marginality. What happened with insight to trade ahead of market drops. But when the news reveals a sudden unexpected adverse turn, the trading away option often disappears. In fact, using the trading option will often result in locking in more severe losses than what might eventually occur. And in the most extreme situations, trading just freezes up and there is not even the option to get out with an excessive loss. prime mortgages rose tenfold, there were two major changes that occurred. First, the subprime mortgages were no longer going to a marginally more creditworthy subset of the folks who would technically fit into the subprime class; they were going to anyone in that class. Any prior experience factors that were observed of the highly select subprime folks would not apply to the average subprime folks. So what was true on the margin is not true in general. entity that handles risks should be assessing what would So the conclusion here is that, at some level, every happen if they ended up owning the risk that they thought they would only have temporarily. This would have a estate market that was driven by the non-marginal amount The second marginal issue is the change in the real 22
Slide 23: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Your Mother should know by David Ingram of new subprime buyers who came into the market. On the way up, this expansion in the number of folks who could buy houses helped to drive the late stages of the price runup because of that increased demand. That increase in price fed into the confidence of the market participants who were feeding money into the market. back on the subprime-related activity a battle between the financial/risk area and a major revenue source. The extreme version of this issue is what is being reported in the press about the accounting for the financial products unit at AIG, where the business unit head excluded a key audit person CDS business—where decisions were made that ultimately auditor. from their discussions of how they would account for their led to a finding of material weakness by the independent Risk takers need to have a reliable source of independent information about the risks of their businesses that is outside of a political fray. It happens again and again that business managers portray the risk assessment as a political decision. But a simple look at incentives would reveal that only one player—the business managers themselves—has the incentive to push a particular point of view. A simple grid can be established that looks at four possibilities: 1) that the negative risk assessment is true and the firm acts to reduce potential losses; 2) that the risk assessment is false and the firm reduces activity to reduce losses; 3) that the assessment is false and the firm acts; or 4) that the assessment is false and the firm does not act. Under most compensation programs, the business manager will be incented to continue business regardless of the risk. They are incented AGAINST risk management. Usually, the risk manager incentives do not change materially under any of the four scenarios. Top management needs to be aware of this incentive mismatch when listening to the arguments. principle,” which almost always indicates that it is the money. My mother would have known. Often you hear the phrase “it’s not the money it’s the analysis can produce incorrect results. They should follow look into their statistics more carefully. Risk managers should always be aware that marginal my mother’s caution “What if everybody did that?” and When my friends and I gathered in my backyard to play various games, my mother always kept her ears open. impose rules or rulings that would in current business terms be called self-dealing. She was always pointing out that I was often trying to wrong in the U.S. housing market at least six months 2007, HSBC reported large additions to mortgage loan loss There were signs that something was dangerously prior to the August 2007 market freeze. Back in February reserves for its U.S. business. The first public reports of a stoppage in the run-up of real estate prices came out in the largest losses did not stop their activity until the day that the market froze. How could they be so blind? Some of this was driven by the folks who themselves spring of 2007. But several of the firms that experienced the were employed full-time doing that business. To them, this activity was the sole source of their income. They had to keep dancing. They had to hold the opinion that the bad news was a temporary blip and that things would soon turn around. In fact they had very strong incentives to portray the situation that way and to cast doubt on anyone who claimed otherwise. That would make the decision to pull David N. Ingram, FSA, CERA, MAAA, FRM, PRM, is senior vice president at Willis Re in New York, N.Y. He can be reached at david.ingram@willis.com. 23
Slide 24: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Mixed Risk Management strategies – Diversification That You Can Count On by Wendy Yu After all this time, I am still shocked that certain investment banks (that are experts at distributing credit risk) held so much credit risk at the worst possible time. For insurers to avoid similar situations, the obvious solution is to clearly identify their roles as risk distributors or risk insurers. HowStrategies that are dynamic with respect to time can potentially eliminate the latter problem. Consider the following as an example. Stable markets • • • Transfer long-term risks Use market price of risk in pricing Lower capital requirement ever, this is not always a realistic option in a world with sticky prices and volatile market shares. A more realistic by a diverse set of risk management strategies. solution is to develop a diverse portfolio of products backed I recently witnessed a number of heated discussions Volatile markets • • • Hold long-term risks pricing Use long-term historical data and actuarial prudence in Higher capital requirement However, strategies that are dynamic with respect to between actuaries on whether market-implied volatility (the price of risk according to the options market) should be used in pricing long-term products. More generally, in our increasingly mark-to-market world, what does it mean to use the market price of risk in pricing long-term products? For a distributor of long-term market risk, it makes time are difficult in practice, since they require more capital in volatile markets. Again, the fact that product pricing is sticky poses a significant challenge. Strategies that are diversified with respect to product sense to use market-implied volatility, because the distribu- tor generally pays the market price of risk when it transfers facing risk distributors is that product pricing is sticky, the risk to the capital markets or reinsurers. The problem whereas the market price of risk can be quite volatile. This disconnection can become particularly severe during market downturns. An insurer of long-term market risk, on the other hand, are less demanding in terms of capital management and product pricing. Consider the following as an example. Accumulation product • • • • Transfer long-term risks needs to take a long-term perspective. At the peak of an economic cycle, the risk appetites of market participants are high, driving down the market price of risk. Does it peak of an economic cycle, when downward shocks to the Use market price of risk in pricing Lower capital requirement Hold long-term risks Income product • • make sense to charge a low price for long-term risk at the financial markets are more likely? Conversely, does it make Use long-term historical data and actuarial prudence Higher capital requirement The key is to maintain a diverse portfolio of products sense to charge a high price for long-term risk at the bottom of an economic cycle, when upward market movements long-term historical data and actuarial prudence is likely to are more likely? For a risk insurer, an approach based on do better. The problem facing risk insurers is maintaining market share in benign market conditions when the market price of risk is low. so that the capital requirement is relatively stable across time. For each product, the pricing methodology is consistent across time. In benign market conditions, the price of the accumulation product is likely to be competitive. In volatile market conditions, the price of the income product is likely to be competitive. 24
Slide 25: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Mixed Risk Management strategies – Diversification That You Can Count On by Wendy Yu strategy is that the additional knowledge acquired in imple- Another important benefit of a mixed risk management the best asset class ceases to be the best when everyone invests in it, best practice risk management ceases to be best practice when everyone practices it. menting multiple strategies would help insurers refine each strategy. This benefit should not be underestimated. Just as Wendy Yu, FSA, is a financial services consultant at Oliver Wyman in Toronto. She can be reached at wendy.yu@ oliverwyman.com. 25
Slide 26: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Derivatives, aIg and the Future of enterprise Risk Management by Michael G. Wacek As one highly rated financial firm after another blows up, what is the right conclusion to draw about enterprise risk 09 demonstrate its criticality, or does it bring the whole concept into disrepute? management (ERM)? Does the financial crisis of 2008balance sheet leverage, many companies have looked for ways to minimize the size of the assets and liabilities they reported on the balance sheet. Derivatives are attractive because they can often be structured to replicate traditional asset transactions but with a much lighter balance sheet in some cases, they can facilitate a business strategy that would not be executed using traditional assets. AIG’s short portfolio of credit default swaps (CDS), sion in Congressional testimony that he had “looked to the equity” and was “shocked” that such self-interest had not such a view was naïve and that more regulation was (and remains) necessary. Alan Greenspan has been criticized for his admis- impact. Clearly, not all derivative transactions are bad, but self-interest of lending institutions to protect shareholders’ motivated better risk management. His critics claim that reportedly totaling $450 billion of “notional” limits at the a CDS on a corporate or asset-backed bond is effectively the same as buying the bond and shorting a risk-free gov- time of the government rescue, is a case in point. Writing know how to optimize their own enterprise risk and will voluntarily seek to do so. Would more prescriptive regulation really improve ERM effectiveness, or would it merely Greenspan believed that well-managed companies ernment bond to harvest the risky bond’s credit spread. entries on the balance sheet. The short CDS position is However, the two approaches result in radically different booked at market value as an “other liability.” The market value for a CDS is roughly the present value of the market credit spread on the reference bond in dollar terms over value of a seven-year $10 million notional amount CDS on further encourage companies to manage the measures outERM is to optimize a company’s upside aims and downappetite. If Greenspan’s critics are right, and we must look to rating agencies and regulators to define the risks and how siders have decided are important? The real imperative of side risks within a set of constraints defining its own risk the remaining life of the swap. For example, the market Fannie Mae subordinated debt in late February 2007 was (0.19 percent) per annum. A writer of that CDS on that date $110,000 “other liability.” In contrast, assuming the Fannie to measure them, then ERM will likely never amount to about $110,000, reflecting a credit spread of 19 basis points would have booked an asset of $110,000 and a matching Mae bond was trading at par value, an institution execut- more than a game of minimizing the reportable magnitude of externally specified risk measures. Unfortunately, that seems to be how it has been practiced until now by many risk management prowess. companies, including some highly touted for their superior The surge in recent years in the use of derivatives ing the other equivalent credit-spread-harvesting strategy would have booked a “fixed income” asset of $10 million and a matching $10 million “obligation to return borrowed securities” liability. In February 2007, the balance sheet impact of the CDS-based strategy was only about 1 percent of the equivalent strategy that used Fannie Mae and riskfree government bonds! to obscure balance sheet risks is a manifestation of that approach. Investors and other users of financial statements, including regulators, have developed an intuition about the likely volatility of balance sheet assets and liabilities. They know that, everything else being equal, companies with highly leveraged balance sheets (i.e., high ratios of assets and liabilities to equity) are riskier than those with less leverage. Knowing that the market frowns on excessive trillion, and its GAAP shareholders’ equity was about As of Dec. 31, 2006, AIG’s assets totaled about $1 $100 billion. Would its management have been willing to execute the bond equivalent of its CDS business plan, i.e., 26
Slide 27: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Derivatives, aIg and the Future of enterprise Risk Management by Michael G. Wacek borrow $450 billion to purchase a portfolio of corporate, asset-backed and foreign government bonds? I doubt it. AIG’s investors and creditors would have allowed it. How- default, are not included in AIGFP’s VaR calculation.” Even if management had been willing, it is unlikely that ever, the favorable balance sheet optics provided by CDS (AIG 2007 Form 10-K, p. 124.) That is like a property inaffecting policies exposed in Florida…except for hurri- surer monitoring the potential cost of claims from all perils canes! It suggests that AIG management did not understand modeling of the CDS portfolio. enabled AIG management to pursue an extremely leveraged perhaps without itself fully understanding it. business strategy without attracting much attention and The danger posed by short CDS arises from their its business well enough to properly supervise the risk This episode and others like it potentially create cred- enormous leverage. As a result of a widening of credit value of the Fannie Mae CDS cited earlier skyrocketed nearly eightfold from $110,000 in February 2007 to about $800,000 by June 30, 2008, obliterating the initial $110,000 bonds fared even worse. ibility problems for ERM. Because risk modeling is a centerpiece of ERM, when poorly supervised, but apparently sophisticated modeling exacerbates a business disaster rather than helps to avoid it, it is viewed by some as a general failure of both risk modeling and ERM. spreads on relatively safe Fannie Mae debt, the market asset and producing a loss of 627 percent! CDS on riskier As risky as they are, the issue is not the CDS them- To prevent the baby from being thrown out with the bathwater, it is essential that the current financial crisis failures, and to distinguish between those companies who pretended. The results should be widely promulgated. be studied closely to identify the real ERM successes and truly managed their enterprise risks and those who merely Ultimately, the market will decide on the importance selves, but rather how they are modeled and how well modeling results are understood and used by management. A self-disciplined company with an effective ERM program does not merely take its risk management cues from how its actual risks inherent in its business plan and balance sheet. risks look from the outside. It seeks to model and limit the AIG reported a very low “capital markets trading” of ERM. Over time, the market will punish companies practicing the window-dressing version of risk management and reward those whose ERM proves effective. In the current crisis, while critics complain about government “bailouts” of failing companies, the fact is that investors in those companies have suffered enormous losses. In the seek to learn much more about the quality of ERM within the companies in which they invest. value-at-risk (VaR) as of December 2007 for the financial products unit, which wrote the CDS portfolio. While that and foreign exchange risks, the company admitted that, “Credit-related factors, such as credit spreads or credit VaR calculation reflected interest rate, equity, commodity future, once burned, twice shy, investors will undoubtedly Michael G. Wacek, FCAS, MAAA, is president & chief executive officer at Odyssey America Reinsurance Corp. in Stamford, Conn. He can be reached at mwacek@odysseyre.com. 27
Slide 28: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Democratization of Risk Management By Michael C. Schmitz and Susan J. Forray This year’s financial crisis was described by Alan GreenRisk management efforts can be jeopardized when span as a once-in-a-century event. Similarly, the floods in the Midwest of this past summer were described as 500-year floods. Both characterizations, and both events, serve to remind us of the difficulty in predicting such calamities and in gauging their ultimate scope and effects. Just think: How many once-in-a-century events have occurred this decade? managers blindly rely on models without regard to the the assumptions. One cannot forget that the risk illustrated amount of data underlying them and the reasonableness of by the model will only be the risk that has been present historically, or judgmentally added into the model by its developer. assumption underlying the majority of their work—that historical patterns have predictive value for the future. Yet, paradoxically, it’s also understood that assumptions are For the most part, actuaries accept as necessary a key just when they are needed the most. For example, one Models are often subject to their greatest limitations an excess layer than to price a primary layer—where his- might be more inclined to build a model to assist in pricing not precise, that models sometimes fail, and that the only ence is factored in, the actual number is sure to be different. torical data by itself may be considered sufficient. However, the volatility inherent in any excess layer may not lend itself easily to modeling, and could easily be over- or underestimated. Similarly, mortgage credit risk models that leveraged understatement of risk in the mezzanine and higher layers of mortgage-backed securities. certainty attached to any point estimate is that, once experiThe current financial crisis reminds us of risk’s su- premacy, while leaving us with several key lessons that can help all of us better manage risk in the future. 1. Models are not Perfect Most of us know it already: Models are imperfect approximations of reality. They are valuable, but incomplete, abstractions and only as good as their underlying assump- have too little tail risk built into them would result in a the projections in any model we build, as well as to com- As actuaries, our role is to assess the reasonableness of municate the uncertainty inherent in the model to others. The use of models should be accompanied by specialized risks they are intended to capture. 2. Troubles With Bubbles The housing bubble, at its root, was not difficult to observe. measures of affordability1, even over 75 years or more,2 unprecedented was occurring. Risky mortgage products products pushed the bubble further. By comparing house prices relative to income and other it became clear that these prices were surging. Something never boded well in this context and indeed these same But even when it’s possible to identify economic bubprofessional judgment concerning those models and the tions. Actuaries are well-positioned to build more robust models and to assist in understanding the assumptions and limitations of those models. In some cases, models may be incomplete partly because they rely on a relatively shallow pool of data. Mortgage credit risk models based on limited data from gested only limited risk. However, with more robust data years with consistently rising home prices might have sugover a broader range of economic conditions, a good deal more risk would be evident. There are strategies to build more uncertainty into models, but they are based at least partly on judgment calls. bles, it’s very hard, almost impossible, to predict when they 1 2 “What Happens When Credit Risks Come Home to Roost?” Michael C. Schmitz and Kyle S. Mrotek, Insight Magazine, Autumn 2006. Irrational Exuberance, Robert J. Shiller, Princeton University Press, 2005 (p. 13). 28
Slide 29: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Democratization of Risk Management By Michael C. Schmitz and Susan J. Forray will deflate, and how that deflation will manifest itself. The 4. saving For a Rainy Day Maintaining cash reserves to address unexpected emergencies is, of course, a basic tenet of commercial or personal finance and this lesson also applies to capital management. further economic fundamentals get out of balance—that is, the bigger the bubbles get—the more unpredictable, severe and widespread the ultimate fallout becomes. That makes this lesson one that’s simple to understand, In the same way that an individual might build a cash cush- yet extremely difficult to implement. Once a bubble has and prudence to kick in, even in the face of exuberance— especially in the face of exuberance. ion, they should also plan for the instability of our cycli- been recognized, that’s the time for mechanisms of caution cal economy. But instead, there is a tendency for the good their balance sheets and take on more debt in order to maximize the advantage of participating in the market rise. Then, as markets fall again, the bad times get very bad; times to get very good, as people and companies leverage 3. The Importance of staying Hydrated Global liquidity flowing into mortgage assets to chase yield during a low-interest-rate environment was instrumental in fueling the bubble. This cheap funding to mortgage distribu- de-leveraging can cause a downward turn in the market to accelerate as market participants rapidly sell off declining tors facilitated the expansion into risky mortgage products. quidity of mortgage assets went out the window when many decided that borrowers could not be counted on to pay their Once lost, liquidity can be difficult to regain. The li- assets to avoid being caught short by their debt. For explans during the boom years when their stocks were at their teeth of the crisis with their stock prices significantly down from their highs. ample, many companies participated in stock repurchase heights only to find themselves trying to raise money in the mortgages and when house prices began to decline. More broadly, the capacity of some entities backstopping them was called into question as well. As a result, a downward more collateral, then subsequently downgraded, and in turn downgrades. government. If the federal government is the risk manager Also caught up in the financial downturn is the federal spiral ensued as these entities were first required to post required to post more collateral again in response to the At the same time, the market for mortgage assets was of last resort, it might be expected to manage symmetrically so as to strive for price stability. It might be expected to lean into booms by raising interest rates and reducing the federal debt. However, fiscal deficits grew leading up to the crisis and some say central bankers have opted for a preemptive asymmetric monetary policy3, in which rates were only slowly increased during upswings. quickly lowered during downswings in the economy, but All of these factors created a more challenging position, drying up, causing difficulties in converting these assets collateral. As the stressed entities were forced to raise cash sitely sold into a declining market, exacerbating the spiral. orate with blazing speed and only returns slowly. to cash, and making it increasingly difficult to raise that to cover their shortfalls in capital and liquidity, some requiLiquidity (and trust) is plentiful until it’s not—it can evap- because what’s generally needed in a financial calamity is loose monetary policy and fiscal stimulus. Indeed, we’ve seen radical intervention of late, and the scale of the crisis has thus far assisted with a relatively cheap governmental 3 The Origin of Financial Crises: Central Banks, Credit Bubbles, and the Efficient Market Fallacy, George Cooper, Vintage Books, 2008. 29
Slide 30: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Democratization of Risk Management By Michael C. Schmitz and Susan J. Forray borrowing cost through the flight of investors to treasury securities. However, one hopes that the seeds for the next crisis are not being sown by debt-financing the country’s paid to counter-cyclical capital regimes in light of the current crisis. • Competitors, who now experience difficulty raising money because of the endemic problems in capital confidence in the financial markets markets, including liquidity drying up, and waning stimulus and bailout packages. Much attention is being • The public, forced to bail out troubled entities in the value of their own investments perceived as too big to fail, and facing large reductions Risk is contagious, which means every single employ- 5. Risky Business Is everyone’s Business Finally, it’s important to recognize that unsafe or excessively risky practices can become everyone’s problem. In this case, it’s not just the buyers and sellers who are affected: • The neighbors of homeowners and real estate speculators who borrowed beyond their means now see their own property value falling as these homeowners are foreclosed upon, their homes possibly vacant and boarded up • Subsidiaries and affiliates, whose product units, while entirely solvent in their own right, were brought down had taken excessively risky mortgage positions by the financial obligations of parent companies who ee and citizen has a vested interest in taking on the role of chief risk officer. The cost of bailouts falls on all of us. Thus these bailouts come to the financial industry with the quid pro quo of greater regulation and supervision. However, it is critical that such regulatory initiatives are carefully crafted to improve safeguards to systemic risks and transindustry can take a leadership role in designing and distribparency without stifling innovation. Indeed, the insurance uting new products that allow individuals and companies alike to manage risk in a more effective way.4 Actuaries have much to offer in this endeavor. Michael C. Schmitz, FCAS, MAAA, is a consulting actuary at Milliman Inc. in Brookfield, Wis. He can be reached at mike.schmitz@milliman.com. Susan J. Forray, FCAS, MAAA, is a consulting actuary at Milliman Inc. in Brookfield, Wis. She can be reached a susan. forray@milliman.com. 4 For example, see Robert J. Shiller’s suggestions of home equity and livelihood insurance and continuous workout mortgages as part of his prescription for the democratization of finance in The Subprime Solution, Princeton University Press, 2008. 30
Slide 31: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications should You Have a Chief skeptical Officer? by Max J. Rudolph What a roller coaster ride the past year has been! Occaa rating agency opinion of an asset, with no due diligence performed despite the obvious conflict where the issuer paid for the opinion. sionally we had respites as the financial markets gained scarier. Volatility in the system has increased, not only altitude. This made the next day’s drop even deeper and between months and days but from morning to afternoon. Are we at the bottom now? Who knows? But unless we are switching back to a barter system, we should look for think the world economy is completely falling apart and we learnings that could help prevent future problems. This will rebound. pline? Analysts were considered lacking if an investment opportunity made no sense to them. Putting some Powersupposed expert. Large egos ruled. Point slides together and giving a presentation created a Investors, government, lenders and borrowers were all Where was the due diligence? Where was the disci- also provide a competitive advantage during an inevitable The current financial cycle has been particularly harsh. at fault. Where did the skeptics go? Where was someone asking the pointed questions? Why didn’t chief risk officers identify and mitigate this situation? Why have they been so quiet? Some did identify the growing problem. Those who tried to slow down the “good times” were neutered or fall guy. Risk managers at financial firms have not proven effective. They have either not had the authority to address their concerns or have gotten caught up in the excitement of the “new paradigm.” Common sense has sometimes seemed in short supply, but this is not entirely fair. As with any crisis, hindsight will be 20/20. The economy has many moving to unintended consequences. ignored. Options included quitting, being fired or being the Risk and return are key components to creating an parts. Actions by government and management often lead Many of the reasons for the current crisis are not new. optimal position, and there needs to be a healthy balance models is useful as much for what is discovered from extreme scenarios as from the average results. while managing against goals and constraints. Building History truly does repeat itself. “It’s different this time!” and “You don’t understand the new economics” are mantras that have been repeatedly proven false. Some people truly recognize when they create a scheme to rip people off, while No one wants to slow down the bus as it rolls downhill. others buy into the excitement and sustain the momentum. Many call this the sub-prime crisis, but risky mortgages lay audience, this helps everyone to better understand the risk/return relationship. Risk management, especially when applied holistically to the enterprise, combines the best of provide an immense amount of information, but can mislead without proper context. Scenario planning, where When a modeler communicates complex results to a quantitative and qualitative methods. Quantitative models were simply symptomatic of the underlying excesses building throughout the financial system. Some Wall Street firms take on high amounts of leverage. Many firms borrowed ther investors nor bankers knowing the total extent. Private used low interest rates driven by government policies to more than $30 for each $1 of their own capital, with neiequity firms, investment banks and hedge funds were ring- specific concerns and assumptions are investigated, can a story to accompany the recommendations. Qualitative methods, built from common sense and an effective risk provide knowledge to the strategic planning process and culture, lead to superior results. But these can’t work unless the culture encourages challenges to assumptions, models and strategic thinking. Better decisions can be made. There is no free lunch. leaders, but were joined by many other willing participants. Sometimes investments were entered into based entirely on 31
Slide 32: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications should You Have a Chief skeptical Officer? by Max J. Rudolph incentives ignore risks taken, it is only a matter of time until the process implodes. People will naturally gravitate incentive pay. A company with a review team available and to practices that enhance their pay. That’s why it is called encouraged to challenge assumptions for both new and old When a firm’s culture is driven by growth and manager ear and be aware of all corporate initiatives. This devil’s advocate should be part of the C-suite, and have ownership of the strategic planning process. While internal audit The leader of such a team should have the CEO’s has a role to play, the need here is for a broader role that in addition to checking processes. This leader should be products provides a competitive advantage for that firm. Alternative perspectives are healthy. Their task is to ask framework. The review team could consist of a combination tough questions and maintain a consistent company-wide of internal staff with broad exposure to a company’s products and existing balance sheet, external consultants with knowledge of best practices and academics. Broad financial skills developed through the credentialing process make not be viewed as a cost, but rather as an enabler. Better uct impacts the firm’s existing balance sheet on a marginal basis, not just its standalone effect. actuaries natural members of such a team. This team should decisions are made by those who think about how a prod- challenges the risk culture and develops best practices prepared to state strong opinions so that improved deciaccess to this person, but the primary focus should be on sions can be made. It is important for the board to have full educating and advising the CEO. This person acts as the a new idea to the CEO, he should be able to ask, “Have you run this past the chief skeptical officer and does she concur firm’s “chief skeptical officer.” When a business line brings with this proposal?” The CSO (could also be referred to as the common sense officer) might not always be popular, but the improved decisions made will allow the CEO to more confidently execute the company’s strategic plans. Max J. Rudolph, FSA, CERA, CFA, MAAA, is the owner of Rudolph Financial Consulting LLC in Omaha, Neb. He can be reached at max.rudolph@rudolphfinancialconsulting.com. 32
Slide 33: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Lessons from the Financial Crisis for Directors and CeOs of Insurance Companies© by Jean-Pierre Berliet The present crisis in global financial markets has created Managing strategic Risks It is not enough for insurance companies to understand and manage the financial risks of their business that can cause insolvency. They need also to manage external “strategic” risks to their business. Strategic risks result from events and strategies or reduce their growth prospects and damage that can undermine the viability of their business models their market value. Strategic risks include changes in com- an impression that enterprise risk management (ERM) has failed broadly to protect the safety and soundness of the insurance companies. financial system as well as that of many institutions, including It is ironic, however, that the crisis is often attributed to failures of risk management in leading commercial banks, investment banks and credit guarantors that were once governance and risk management. viewed as pace setters in the use of “best practices” in risk Results speak for themselves: Bear Stearns and petitive dynamics, regulations, taxation, technology and other innovations that disrupt market equilibrium. They also include events and changes in other industries that can performance of insurance companies. Lehman Brothers have disappeared; Fannie Mae and impact adversely the going concern viability and financial Until the present crisis, many insurers did not think Freddie Mac have been nationalized; in the insurance industry, XL and The Hartford have had to raise significant amounts of equity to restore their capital strength; AIG has been partially nationalized while other leading compa- much about their dependence on the efficient functioning and soundness of the banking system. Now they do. Al- of credit and other financial markets or the overall safety though the sub-prime mortgage crisis and resulting credit market meltdown can be viewed simply as market risk events, they should be seen as the combined, unexpected nies such as MetLife and Prudential are rumored to have approached the U.S. Department of the Treasury about the possibility of receiving aid under the financial rescue plan that is being implemented. Meanwhile Wells Fargo, Bank of America and Berkshire Hathaway have been able to complete strategic acquisitions or investments that will serve them well in the future. Looking at outcomes, it prepared. They have done comparatively well. is clear that some companies were stronger and better So what went wrong? What lessons can directors and but theoretically predictable result of design weaknesses in institutional and regulatory arrangements and changes in financial technology. cial system resulted from: • • • From this vantage point, the near collapse of the finanPro-cyclical effects of capital regulations under fair value accounting standards, Explosive growth of outstanding derivative contracts, especially credit default swaps and The redistribution of housing finance risks (especially sub-prime) across financial institutions on a global basis, facilitated by securitization. Together, these factors combined to create a time bomb. CEOs of insurance companies learn from the crisis? What can they do to help their companies become more resilient? have been disciplined about: • • • Companies that appear to have withstood turmoil best Managing strategic risks, Holding sufficient capital and Aligning interests of shareholders and managers. Their discipline demonstrates that they have been tak- ing risk governance and risk management seriously. That it exploded is no market risk event, but rather a failure of risk management. 33
Slide 34: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Lessons from the Financial Crisis for Directors and CeOs of Insurance Companies© by Jean-Pierre Berliet not abdicated their responsibilities to rating agencies and conducted appropriate due diligence, toxic securities would not have found their way to their balance sheets. Simi- The explosion could have been anticipated. Had CROs decision making, building resilient portfolios of insurance businesses and protecting them from strategic risks is a challenge. In their oversight roles, directors and CEOs can help company executives by re-examining the appropriateness of traditions, conventions and modes of thought that influence risk assumption decisions. They should demand that company management: • Conduct periodic defensibility analyses of their consideration of weaknesses in institutional arrangements of the financial system. Such strategy review companies’ business models and strategy, including larly, fundamental changes in the characteristics of mortgage products and the creditworthiness of the customer base should have been examined closely. Such examina- tion would have diminished the attractiveness of CDOs as investments, have reduced their spread throughout the financial system and have prevented or reduced the losses of capital that caused confidence to collapse and market liquidity to vanish. Insurers, however, did not understand that risks to must also focus on the identification and monitoring of emerging trends with adverse effects on competitive advantage and pricing flexibility (loss of business to competitors, emergence of new risk transfer technologies or product innovations, regulatory developments, rapidly. • the financial system were elements of their strategic risk. Strategic risk elements embedded in the financial system are difficult to mitigate. They create dependencies among businesses that undermine diversification benefits achieved through underwriting of a multiplicity of risks and exposures. They have a tendency to hit all activities at the same time. In this area, prudence is the source of wisdom. Com- etc.) that can reduce company valuations sharply and Reassess periodically the company’s strategy for controlling performance volatility and achieving a in risk assumption, diversification (e.g., across lines, industries, regions or countries), ceded reinsurance or structural risk sharing and financing vehicles such as captives or side-cars. panies that have had the discipline not to underwrite exposures that they did not understand, or invest in financial instruments or asset classes that they could not as- balance between risk and return through specialization sess to their satisfaction (e.g., tranches of securitization backed by sub-prime mortgages), have withstood the crisis comparatively well. Some of these companies are benefiting from the weakness of their less thoughtful and less • Assess the possibility for disruption of business plans flexibility in capital deployment. caused by events that reduce capital availability or Develop appropriate responses through adjustment in disciplined competitors. For example, Warren Buffett’s decision to create a financial guaranty insurer recently and timed to capitalize on opportunities created by the weakmarket value of many companies. to resume investing in U.S. companies appears perfectly ness of established competitors and the steep fall in the Methodologies for identifying, measuring and manag- • capabilities, redeployment of capacity across lines of and conditions, ancillary services provided, lobbying of lawmakers and regulators and participation in industry associations. activity, change in limits offered, exclusions, terms ing strategic risks are in their infancy. Since there are no established conceptual frameworks to guide analysis and • Hold executives accountable for discipline in under writing and investment decisions. 34
Slide 35: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Lessons from the Financial Crisis for Directors and CeOs of Insurance Companies© by Jean-Pierre Berliet lated, many insurance companies have not developed a urgently. Because the insurance industry has been highly regu- deep strategic risk assessment capability. They need one a company should hold more capital, as an insurance policy against many risks to its survival that are ignored under a solvency framework. They conclude that, under a going concern assumption, Holding sufficient Capital The issue of how much capital an insurance company should hold beyond requirements set by regulators or rating agencies is contentious. Many insurance executives capital productively on behalf of shareholders would it impossible for many banks and insurance companies to shore up their capital positions. It prompted federal authorities to rescue AIG, Fannie Mae and Freddie Mac. The “going concern” view appears to have been vindicated. The recent meltdown of debt markets, however, made hold the view that a company with a reputation for using be able to raise additional capital rapidly and efficiently, as needed to execute its business strategy. According to this view, a company would be able to hold just as much “solvency” capital as it needs to protect itself over a onein-force policies plus one year of new business. In this ensure that their companies hold an amount of capital that is appropriate in relation to risks assumed and to their business plan. Determining just how much capital to hold is fraught with difficulties, however, because changes in Directors and CEOs have a fiduciary obligation to year horizon from risks associated with the run off of framework, the capital need is calculated to enable a capital held have complex impacts about which reasonreduces solvency concerns and the strength of a company’s of return on capital that is being earned; and conversely. able people can disagree. For example, increasing capital ratings while also reducing financial leverage and the rate Since directors and CEOs also have an obligation to act company to pay off all its liabilities, at a specified conunder the assumption that assets and liabilities are sold into fidence level, at the end of the one-year period of stress, the market at then prevailing “good prices.” If more capital were needed than is held, the company would raise it in the capital market. Executives with a “going concern” perspective do not prudently, they need to review the processes and analyses used to make capital strategy decisions, including: • assumed under a going concern assumption, with Economic capital projections, in relation to risks agree. They observe first that solvency capital requirements increase with the length of the planning horizon. Then, they correctly point out that, during a crisis, prices at which consideration of strategic risks and potential systemic shocks, to ensure company survival through a collapse raised or becomes exceedingly onerous • • • • and financial analysts of financial markets during which capital cannot be Management of relationships with leading investors assets and liabilities can be sold will not be “good times” Asset prices are likely to be lower, perhaps substantially, prices upon which the “solvency” approach is predicated. while liability prices will be higher. As a result, these executives believe that the “solvency” approach, such as the Solvency II framework adopted by European regulators, understates both the need for and the cost of capital. become too onerous or unavailable in the capital market. In addition, they remember that, during crises, capital can Development of reinsurance capacity, as a source of “off balance sheet” capital agencies and regulators Management of relationships with leading rating Development of “contingent” capital capacity. 35
Slide 36: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Lessons from the Financial Crisis for Directors and CeOs of Insurance Companies© by Jean-Pierre Berliet is very important to success. Directors and CEOs cannot The integration of risk, capital and business strategy its performance. This is dangerous. Mitigating this form of moral hazard is difficult because its effects are not readily apparent. let actuaries and finance professionals dictate how this is to happen, because they and the risk models they use have tions, directors and CEOs need to remember that models are bound to “miss the point” because they cannot reflect business systems to which they apply. been shown to have important blind spots. In their deliberacannot reflect credibly the impact of strategic risks. Models surprises that occur outside the boundaries of the closed ers have even greater opportunities to “game” incentive plans: they can increase reported business volume and profit in the current period by slightly underpricing or increasing risks assumed. This approach to “making the numbers” is particularly tempting in lines of coverage in which In insurance companies (and banks), business manag- losses can take many years to emerge and develop; it is also particularly dangerous because losses from mispriced poliloss experience, can be devastating. Similarly, investment crease portfolio performance, while involving risks that can result in significant capital losses later. cies, especially in lines with high severity/low frequency officers can invest in assets that offer higher yields to in- aligning Interests of shareholders and Managers Separation of ownership and control creates conflicts of interests between managers and owners. To mitigate this situation, companies expend much effort to develop and implement incentive compensation systems that align the interests of managers and shareholders. The present crisis demonstrates clearly that such arrangements are imper- fect: large incentive payments were made to many people There has been a public outcry. insurance companies need to work with management to: • Based on these observations, Directors and CEOs of Link incentive compensation payments to the ultimate (especially when fair value accounting standards cause immediate recognition of profits on contracts). in companies that have performed poorly or even failed. But there is nothing really new in misalignments of outcome of business written rather than to current profits incentives, or weaknesses in incentive designs that produce harmful results: they exist in every company to some degree. In a typical situation, managers are concerned about mini- • Establish and empower an internal control and audit with business strategies and plans. mizing financial and career consequences of not achieving their objectives. If the situation requires it, managers will to achieve their targets. They will seek and capitalize on opportunities to convert unreported intangible assets, such as market share, product or service quality, product leaderexploit every opportunity to change their operating plans • function to verify that managers’ actions are aligned Verify the integrity of underwriting and investment and processes. decisions, in relation to explicitly approved guidelines The present crisis has demonstrated how unbundling ship, plant productivity or customer service responsiveness expenses. Financial results will look good, and they will of risk assumption businesses can increase moral hazard by redistributing risks, gains and potential losses across origitors/risk bearers. nators, arrangers of securitization transactions and invesReconstruction of incentive programs and establish- into current profits by postponing and reducing related be praised for accomplishing their objectives. Actions that they took, however, accelerated uncertain future income to the present period while undermining the company’s competitive capabilities and reducing the sustainability of ment of appropriate oversight and enforcement mechanisms are needed to reduce moral hazard and restore confidence 36
Slide 37: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Lessons from the Financial Crisis for Directors and CeOs of Insurance Companies© by Jean-Pierre Berliet in the financial system, including insurance companies. Conclusion In the aftermath of the present crisis, directors and CEOs of insurance companies should demand that management enhance the effectiveness of ERM frameworks and processes. which directors and CEOs work with management to: • • • deployment and risk management processes, companies as “going concerns” and and managers. Greater progress will be accomplished by companies in Add a strategic risk management component to capital Increase capital held to support the value of their Reshape incentives to align interests of shareholders cians and citizens will be watching. Regulators, rating agencies, investors, clients, politi- Note: See “Increasing the Usefulness of ERM to Insurance Companies,” by Jean-Pierre Berliet, in the newsletter of the Joint Risk Management Section of Society of Actuaries, Casualty Actuarial Society and Canadian Institute of Actuaries, August 2008 and at: http://www.soa.org/library/ rmn-2008-iss13.pdf newsletters/risk-management-newsletter/2008/august/ Jean-Pierre Berliet is the founder of Berliet Associates, LLC, a New York area-based advisory firm on strategy and risk management. He can be reached at jpberliet@att.net. 37
Slide 38: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications against the grain: The Wisdom of Countercyclical Capital by Anson J. Glacy, Jr. Over the past two decades, capital regimes for financial gimes (like Basel II and Solvency II, the European direcnotion of value at risk (VaR) to make those establishments. tive on insurers’ capital adequacy effective in 2012) use the intermediaries have evolved from simple, static estimates toward dynamic, risk-based methodologies that reflect the real nature, extent and mix of risks to which an organization is exposed. This evolution is undoubtedly a good thing, helpand, hopefully, stay solvent through extreme conditions. ing companies maintain capital in proper proportion to risk However, these new ways of calculating capital re- large, multivariate, normally distributed model compoassociations but usually are stationary across the event space. The institution sets a benchmark probability of ruin, VaR-based capital implementations typically involve nents. Correlation factors are applied to reflect risk-factor quirements are relatively untested. We may be witnessing economy is almost surely entering a serious recession. One the results of such a test today. As of this writing, the global of the reasons for this appears to be inadequate capital provisioning. The failure of capital regimes to protect financial entities (and, by extension, their customers) has had unfortunate effects on the markets and economies of the world. based on a desired level of financial strength, aligned with rating agency standards for a target rating. Capital is then set based on quantile statements about ruin: a given level solvency, for example. of capital assures a 99.5 percent probability of continued Equity market volatilities (like the CBOE Volatility Index®, or VIX®) often used in VaR reckonings exhibit today’s capital regimes measure risk based on prevailing One contributory theory for why this happened is that a well-known inverse relationship to the general levels measures change as market conditions change—in of the stock market. Capital estimations based on these precisely the wrong direction. Capital amounts based on these volatilities will tend to shrink as markets advance. intermediation business, promotes the application of leverThis improves the return-on-economic-capital profile of an age and motivates management to sell more business. One could also say it adds fuel to the fire. In effect, VaR tells us, big bets.” The problem is that “right now” is not the appropriate time horizon for measurements of risk. In fact, some have criticized Basel II from the begin“Right now, things look pretty good, so go ahead and make market conditions. Bull markets with low volatility levels and upward trajectories are taken to indicate conditions of markets with high volatility and falling prices are taken to requirements rise. low risk. In these conditions, capital requirements fall. Bear indicate conditions of high risk. In these conditions, capital Taken as a whole, this approach to capital is pro- cyclical: it intensifies economic swings, enabling companies to take greater risks when times are good and restrict- ing their options when times are bad. This is being seen asset depreciation are simultaneously told to post draconian tal’s intended effect. today as companies facing severe balance sheet losses and levels of capital—achieving precisely the opposite of capi- ning for enshrining pro-cyclical capital estimation methods. They point out that it establishes VaR-like capital levels by incorporating market-implied volatilities, requiring too little capital during economic upswings and too much berant times, false asset bubbles are created that end in tears. during recessions. Instead of restraining lending during exuIn the Solvency II framework, the minimum required Value at Risk Puts the economy at Risk An economic approach to the establishment of capital is based on measuring the calls upon capital resources that could happen under extreme events, the proverbial “tail” of the distribution. Today’s evolving risk-based capital re- capital is prescribed to be at the 99.5 percentile (i.e., one failure every 200 years). Assuming normality… 38
Slide 39: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications against the grain: The Wisdom of Countercyclical Capital by Anson J. Glacy, Jr. Required Capital = Ф-1(0.995) = 2.58σ …where Ф is the normal distribution function and σ estimates variability of company economic outcomes. The multiplier factor of 2.58 means that required capital exness risk. At the 99.95th percentile (one failure every 2,000 pands and contracts at more than twice the rate of busiyears) the capital multiplier is 3.29. Thus, small reductions free up outsized amounts of capital. cess capital on their balance sheets and have the resources When times are good, companies can afford to hold ex- to build it up. In effect, they should be saving for a rainy day. When the rain comes, companies need to spend down capital to protect themselves from ruin. During the middle part of this decade, risk spreads contracted to levels that made the intermediation business difficult. It seemed that all for alpha became a consuming obsession for companies as they sought the slivers of a basis point necessary to keep financial assets became “priced to perfection.” The search in estimations of σ under the Solvency II framework can Reversing the Thrusters: Implementing Countercyclical Capital In fact, an opposite approach—one that is countercyclical— might serve us better. Countercyclical capital measures risk based on a more expansive time horizon than does VaR. their business models afloat. In the United States, leverage came to the rescue as the Federal Reserve graciously assisted in enhancing the risk/reward profile of an intermediation business through aggressive “bubble” management efforts. For example, if financial intermediaries were required to evaluate loss potential by considering performance over the entire economic cycle, they could establish capital levels that counteract cyclical forces. This approach is consistent with a recognition that financial risk actually arises at the inception of a loan or insurance policy and lasts for its entire lifetime. So, instead of a market-implied measure of σ one could either reflect real-world risk dynamics observed recent observations. the over-reach during these times of tight risk spreads, making a whole raft of marginal intermediation projects would have been banked that could be of good use right A countercyclical capital regime would have restrained uneconomic and therefore undone. At the same time capital now in arresting the de-leveraging spiral. Put simply, the would not be nearly as bad. Few observers of today’s ecois a bad thing. good times would not be quite as good—but the bad times nomic turmoil would argue that a dose of such moderation over a full-cycle historical period or a rolling average of Anson J. Glacy, Jr., ASA, CERA, MAAA, is a senior consultant at Milliman Inc., in Chicago, Ill. He can be reached at jay.glacy@milliman.com. 39
Slide 40: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications prudent enterprise Risk Management strategy: Aligning incentives Reaffirming your company’s commitment to eRM in light of the financial crisis By pRAkASH SHIMpI A tale of improperly placed incentives by sAM GutteRMAn Risk Management: the current financial crisis, lessons learned and future implications by neil bodoff 40
Slide 41: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Reaffirming Your Company’s Commitment to eRM in Light of the Financial Crisis by Prakash Shimpi The current financial crisis underscores the need for companies to take a sobering look at their approach to risk management. Among the many lessons to be learned, one is immediately clear: The subprime debacle represents a failure in risk management, rather than a failure of risk management. While we are still in the midst of the crisis and there Finance executives in the cross-industry survey reported that improving their own companies’ risk management was a priority, even ahead of short- and long-term access to capital. In fact, only 4 percent of respondents feared the current financial meltdown would have a severe impact on their companies’ financial prospects. However, 72 percent of responmanagement practices and ability to meet their strategic plans. dents expressed concern about their own companies’ risk These survey findings indicate a renewed resolve on may be other shoes still left to drop, some general views are already emerging. There are many reasons why we are in a crisis, but inadequate risk management practices feature high as a contributory factor. Clearly, improvements need to be made, and we see three aspects of enterprise risk the part of financial executives to invest in more effective risk identification, measurement and management procedures. Moreover, 42 percent of the respondents also policies on the part of boards of directors as well as increased employee-level involvement. predicted greater involvement in risk management management (ERM) implementation that should be strengthened. First, far from being a compliance exercise, risk management is a strategic imperative and should be treated as such. Second, financial managers should urgent- ly reassess the adequacy of their current risk management capabilities in order to do so. Finally, the greatest engagement of employees, as well as the board and senior executives responsible for risk management. Finance executives’ Viewpoints Towers Perrin conducted two surveys in 2008 that provide cussed here. The first study was a cross-industry survey of of September 22, just as the first U.S. Treasury bailout plan a fact base for the conclusions and recommendations dis125 top U.S. finance executives1 conducted during the week shortcoming is cultural; management should improve the 62 percent of the cross-industry survey respondents pointed When asked to lay blame for the current financial crisis, to poor or lax risk management at financial institutions as increased complexity of financial instruments (59 percent), the single greatest contributor. Other major causes included financial market speculators (57 percent), predatory lending tices in the financial services sector (44 percent). practices (50 percent) and incentive compensation pracAs executives take a closer look at their own risk man- agement practices, one problem they are likely to find is insurance industry survey found that only a small fraction implemented ERM into their culture. incomplete, slow or uneven application of ERM. Our of companies around the globe can claim to have fully Within the insurance industry, embedding ERM into was heading for legislative approval. The second study was a global survey of over 350 top finance executives in the insurance industry,2 the fifth in a series of biennial insurMay and June 2008, as the crisis was brewing. ance industry ERM surveys, which was conducted during business processes is proving to be a challenging mis- 1 2 Senior Finance Executives on the Current Financial Turmoil. A report prepared by CFO Research Services in collaboration with Towers Perrin. November 2008. http://www.towersperrin.com/tp/getwebcachedoc?webc=USA/2008/200811/TP_Financial_ Crisis_Survey_Report.pdf. Embedding ERM — A Tough Nut to Crack. Towers Perrin. October 2008. http://www.towersperrin.com/tp/showdctmdoc. jsp?country=global&url=Master_Brand_2/USA/News/Spotlights/2008/Oct/2008_10_28_Spotlight_Embedding_ERM.htm. 41
Slide 42: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Reaffirming Your Company’s Commitment to eRM in Light of the Financial Crisis by Prakash Shimpi for making risk-based decisions, has become increasingly two years. However, more than half (55 percent) of survey sion. For example, economic capital (EC), a robust metric 1. Treat eRM as a strategic Imperative If ERM is to be truly integrated with how firms are manment of the firm’s board and senior executives. aged, then implementation must begin with active engageReinforce the role of the chief risk officer. This is the important to regulators and rating agencies over the last respondents believe that substantial work is needed before they can use EC to guide risk-based decision making, and 60 percent noted that considerable strides must be made ment. Only 10 percent of the firms responding said they before they can link EC metrics to performance managehave appropriate EC capability fully in place. More than 40 percent said they remain focused on getting the basics right bedding ERM, significant numbers of respondents indicated changes in risk strategy or appetite (36 percent), asset strategies in their EC calculations. But in spite of the slow pace of emtheir ERM programs have already resulted in key business (35 percent) and product pricing (31 percent). single most important action that a company can take to recognize ERM as a strategic imperative. Many companies have appointed a senior executive (chief risk officer, CRO) to oversee risk management. The current financial crisis not sufficient. If, as we believe and our surveys indicate, ERM is viewed as critical to the survival and profitability of a firm, then the CRO’s responsibility must be commensurate. Studies have shown that problems arise when risk has shown us that merely making such an appointment is management does not have a seat at the management table, or when risk management’s warnings are ignored, or when risk management is performed unevenly. No doubt, authors find evidence of all three. the importance of managing their entire risk landscape, not just those risks that are familiar or easy to quantify. One particular problem area is operational risk. According to the survey, only 7 percent of insurers believe they have appro- More and more companies are beginning to recognize and academics writing the history of the current crisis will The current validation of the risk management function priate operational risk capability in place, while 37 percent admit significant work is still required. Yet despite these admissions, operational risk ranks only seventh among set limits to govern day-to-day risk taking, over 70 percent have limits for market, credit and insurance risk, but just 26 percent have limits for operational risk. could result in a dramatic improvement in corporate prestige. Just as a CFO has a specific set of responsibilities, we may soon see a convergence of responsibilities that are survey participant priorities. Of those companies that have aligned with the CRO. Indeed, these new responsibilities may require the establishment of new professional standards and levels of experience for future CROs. As stakeholders CROs may see their professional and fiduciary obligations come to realize the importance of risk management, increase. And, as regulators and the financial industry seek ways to prevent past mistakes, risk managers will likely play an increasingly important public policy role. Increase board engagement on risk. We expect that strengthening eRM Implementation Based on these surveys and discussions with finance managers in a variety of industries, the commitment to ERM remains strong, and there is increased urgency to strengthen ERM implementation. Although there are many ways to do this, we have identified three areas of focus and recommend specific actions within each area that require immediate attention. boards should and will demand better metrics and informaboard’s level of questioning dig deeper and be less satisfied tion about risk management performance. Not only will the by traditional compliance or audit reports, the questioning 42
Slide 43: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Reaffirming Your Company’s Commitment to eRM in Light of the Financial Crisis by Prakash Shimpi will place a premium on verifiable evidence of employee involvement. We anticipate a significant increase in the expect that their scope of oversight will be broad. number of board-level risk oversight committees, and we Align incentives to reflect risk. Although this has been fungible within the different parts of a conglomerate finanflow of capital and cash between legal entities within an restrictions would have prevented AIG from dealing with its problems. Some type of fungibility testing has been clear from AIG’s collapse is that capital and cash are not Fungibility should be stress-tested. One lesson made cial institution. Legal and regulatory restrictions limit the enterprise. Even if the needed funds were available, these a topic of discussion for some time, the current crisis has demonstrated that compensation practices can be at odds sation programs will undergo a transformation as companies with managing risk appropriately. We believe that compenattempt to rid themselves of inducements to exceed stated risk tolerances. We expect the scrutiny of incentive compensation programs, historically left to policymakers and directors and fellow managers, who are loathe to share the fate of companies that have failed in the wake of this crisis. 2. Improve Your eRM Capabilities Companies need a variety of skills, methodologies, tools and processes to manage risk appropriately. Each of these is probably worth reassessing in the current environment to suggested within the Solvency II framework, and its potential the limits of capital and cash flow between legal entities within the same organization is vital. value to risk management is now evident. Understanding investor groups, will come increasingly from boards of 3. Understand and Manage Your Risk Culture At the end of the day, good risk management results from people doing the right thing. It is not sufficient for ERM nor should it be put on the shoulders of employees without proper guidance. to impact only a few people at the top of the organization, identify and overcome any significant shortcomings. If one gregate risk exposure across the firm, then two issues need urgent attention. of the aims is to add up all the bits to develop a view of ag- has been said about setting the right “tone at the top” for Establish clear guidance on accountability. Much ERM. Companies still have a long way to go to do that in a way that is clear and engaging to employees. A starting values as well as its risk strategy and objectives. Ultimately, though, it is management’s own actions in holding people accountable in a way that reinforces the alignment of that will make a difference. point may be to articulate a company’s mission, vision and there is a fundamental disconnect between the way institutions Recognize operational risk as material. In our experience, view operational risk and the way operational risk manageoccur because the term operational risk conjures up images of day-to-day processing errors. These minor operations ment should be implemented. To a large extent this may interests of employees, management and other stakeholders Assess your risk culture regularly. In order to make a issues are often only a small part of operational risk, which is driven in large part by catastrophic failures in management (e.g., inappropriate sales practices or unauthorized activities). Data shows that a significant number of corporate bankruptcies and insolvencies during the past 20 years have been caused by operational failure. Indeed the current management at so many levels. financial crisis can be viewed as a failure of operational risk difference in employee engagement, management needs to determine whether management’s impression of the company’s risk culture is borne out by rank-and-file opin- ion. Employee risk awareness and engagement should be assessed regularly to identify gaps between management ate measures undertaken to bridge the divide. expectations and employee understanding, with appropri- 43
Slide 44: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Reaffirming Your Company’s Commitment to eRM in Light of the Financial Crisis by Prakash Shimpi recommended actions were put forth with the view that had These three aspects of risk management and supporting might not be in the midst of such a severe financial crisis. It is the actions we take now that can help us prepare to navigate the complex and inherently risky world of the future. such practices been more firmly established, perhaps we Prakash Shimpi, FSA, CFA, MAAA, CERA, is global practice leader, ERM at Towers Perrin in New York, N.Y. He can be reached at prakash.shimpi@towersperrin.com. 44
Slide 45: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications a Tale of Improperly Placed Incentives by Sam Gutterman Arguably, many of the fundamental contributing factors much if any investment, and (2) purchasers who did not have the current financial capacity to repay their mortgage loans, taken out for houses whose values had been bid up in a housing value bubble. Some of the latter mortgages, issued to those with limited financial resources or weak loan experience (often referred to as sub-prime mortgages), were to some the underlying source of financial chaos. However, to the 2008 mortgage crisis involved misdirected incenthe housing finance process. The incentives discussed here tives and misinformation available to key participants in include those of the mortgage holders, mortgage intermeand other highly paid staff of those involved. Unless simi- diaries, mortgage providers, securitizers, raters and CEOs lar incentives are recognized, future public policy decisions will again fail to avoid this system-wide risk. Since I am most familiar with the U.S. situation, my comments will be limited accordingly. It has been U.S. public policy to facilitate and help en- I believe that they were one source of the underlying problem, with an overall credit problem covering a much larger percentage of the population. Mortgage products that were designed and actively able home ownership for as many people as possible. This marketed to these individuals enabled and encouraged commendable policy objective has been promoted through mortgage purchases. These included mortgage contract full or partial mortgage loan warranties provided by pseudopublic entities [such as the Federal National Mortgage Association (Fannie Mae), and the Federal Home Loan Mortgage Corporation (Freddie Mac)], public entities (including federal agencies or departments such as the Federal Housing Administration and the Department of Veterans Affairs), private mortgage insurers, and through the securitization of mortgages. These insurance or financial which mortgage loans might be marketed to a wider mar- features such as teaser loans (with extra-low interest rates for an initial period), interest only loans, loans equal to or even in excess of current house value and adjustable rate loans. Some of these became quite popular because they got around financial regulations, such as banks’ capital adequacy formulas. Some have blamed loose underwriting standards, or even lack of standards, on broader access to credit. Too vehicles have provided financial support to lenders through ket than might otherwise have been achieved by individual lenders. Unfortunately, at the same time, these have led be considered to be misguided subsidies to the housing market. in part to unintended consequences, what in retrospect could many people with inadequate financial resources were encouraged to take out excessive mortgages for their houses when housing values were at their peak. Loans were issued on the basis of no financial documentation (no doc loans), in some cases as a result of fraud. These factors in turn helped fuel the housing bubble. limited financial documentation (low doc) mortgages, and value inflation would continue at a rapid double-digit rate of growth was widely held and encouraged speculative Especially since 2004, the view that recent housing offer these loans to these markets? Although it is all too easy to attribute their actions to simple greed (and, of course, there was a little of this) in taking advantage of an enthusiastic market; in part, a long period of low inflation and economic stability reduced investors’ perceptions of risk. From the view of the lending institutions, a relatively What motivated mortgage intermediaries and lenders to excesses from two types of mortgage purchasers: (1) investors who purchased multiple residences under the assumption that buying and selling these properties at able by an expectation of flipping them quickly, without very little investment or initial cost could be very profitincurring much if any personal financial risk or even to put large up-front profit could be obtained accompanied by what seemed to be limited or no cost or risk. This 45
Slide 46: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications a Tale of Improperly Placed Incentives by Sam Gutterman incentive to increase market share seemingly without adding risk promoted bad business practices. Although in many respects similar tendencies existed in prior housing finance cycles, they were exacerbated here by the existence of vehicles that appeared to pass risk to others—so under- continuation of the rapid increase in housing values. Those who sold the mortgages thought that, since they didn’t have in the risk chain also continued to believe that housing to bear any downside risks and as long as the next level value trends would continue, the greater volume generated enhance overall income. Those who bought the mortgages were banking on the writers took inadequate precautions to ensure that only appropriate mortgage loans were taken on. It was further models based on recent price histories. exacerbated by the extensive use of Wall Street computer The availability of pass-through vehicles, be they through more creative debt and derivative products would Will such products and underwriting ever be seen through securitization or special purpose vehicles that did not need to be consolidated, either of which could avoid again? Well, it is too early to look ahead to the next time the housing finance cycle reaches this stage, but similar variants, possibly with different names, are likely occur. capital charge requirements, seemed to provide profit withInvestors desperately seeking any additional yield eventuout risk—rarely possible in a competitive environment. Or similar trends will arise in other areas (e.g., credit card loans and mortality). But it is likely that if short-term thinksame type of situation will arise. Recommendations Future systemic risks need to be better identified and assessed. One example is to ensure that more stakeholders ing and current trend extrapolation occur in other areas, the ally led to a bonanza in mortgage-backed securities that not only helped create a glut of new homes, but ultimately led to an inability to sell homes in foreclosures in some areas that further led to the downward housing price spiral. existed at the entity level, in retrospect there were reasons Although seemingly sound risk management practice bear some of the cost or keep some of the risk; otherwise why inadequate charges were being made. Inadequate risk assessment and transparency, accompanied by systemic led to huge prospective costs foisted on the entire system. financial risk that proved ultimately to be unavoidable, In part this was due to a lack of individual risk-bearing, a feeling of not being responsible, that ultimately created create risk where none existed). Should (or rather can) this be eliminated in the future? huge moral hazard (in this case provided incentives to moral hazard will become significant (simply passing the risk along, providing through fixed fees the incentive to write business at a loss). pants and stakeholders in the system, education that reflects both short-term and long-term incentives and views. This Better financial education is needed for various partici- need not only exist for potential mortgagees, but also for executives and compensation consultants. Effective corpoinvolve actuaries. A populist solution inevitably points a finger to CEO and executive compensation—indeed, overemphasis on short-term features of such compensation may rate governance will listen to risk management teams that lighted: (1) an overemphasis on short-term thinking and Two further contributing factors need to be high- (2) the typical human tendency to assume that current trends will continue. These are closely interconnected, although the former can be also viewed as equivalent to the application of very large discount rates. These factors have been at the root source of most housing bubbles, as they also have been a factor for most underwriting cycles in insurance. have contributed to the damages caused; more long-term formulas. performance incentives should be featured in compensation Policymakers need to better assess the unintended consequences of their actions. Financial service regula- 46
Slide 47: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications a Tale of Improperly Placed Incentives by Sam Gutterman tors need to increase (risk-based) capital requirements for pass-through vehicles or require that an underwriter off-loading 100 percent of the risk for the entity who performs the underwriting and product design offering. keep a minimum percentage of each risk written to avoid bilities, using more robust stress-testing whose results are not ignored as being impossible. They should also avoid an overemphasis on recent experience when dealing with potentially cyclical phenomena. Modelers should pay more attention to outlier possi- Sam Gutterman, FSA, FCAS, CERA, HonFIA, FCA, MAAA, is the director & consulting actuary at PricewaterhouseCoopers LLP in Chicago, Ill. He can be reached at sam.gutterman@us.pwc.com. 47
Slide 48: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Risk Management: The Current Financial Crisis, Lessons Learned and Future Implications by Neil Bodoff Various ingredients contributed to the current financial In addition, how a firm measures profit usually crisis, and we can learn many lessons from the crisis. Yet the current circumstances, we ought to pay special attention to those central issues that underpin the current crisis, past crises and future crises. I believe that the central issue that underlies the current rather than myopically focus on the particular minutiae of affects how the firm compensates its key employees; in turn, a firm’s compensation system directly affects employee incentives for risk taking. Therefore, a firm that does not adjust for risk when measuring profit for internal compensation purposes creates powerful incentives for excessive risk taking. On the other hand, when a firm measures profit for compensation purposes on a risk-adjusted basis, there is less of an incentive for employees of the firm to take exces- crisis and many others is how one measures profit. Specifically, the current crisis derives from the lack of “risk adjustment” when reporting profit; a key reform crucial to mitigating future crises is to ensure that we always measure profit on a “risk-adjusted basis.” both of the following contexts: shareholders and others. calculations. Measuring profit on a risk-adjusted basis is critical in sive risk, because increased risk does not automatically lead to the potential for increased compensation. Shareholders, creditors, rating agencies, regulators and taxpayers all have does not generate structural incentives for future excessive risk taking. The proposal to measure profit on a risk-adjusted basis an interest in ensuring that a firm’s compensation system 1. When measuring profit for external communication to 2. When measuring profit for internal compensation There are two critical reasons for measuring profit on a would likely require major changes to accounting standards and financial statements, certainly no simple task. Moreover, it would also likely require substantial changes in the compensation systems of many firms, another difficult underlying intellectual framework, psychology and risk-adjusted basis: ability. challenge. Most broadly, it would require changes in the embedded practice of current methods of measuring profit. 1. It provides more meaningful information about profit 2. It reduces the incentive to take excessive risk in order to increase profits. Currently, a firm reports its profits to external audiences basis would be quite a challenge, but this reform has the Ultimately, the task of measuring profit on a risk-adjusted potential to preemptively undermine the perverse incenlikelihood and severity of future financial crises. without any risk adjustment; thus the firm’s financial statements do not reflect the level of risk the firm undertakes in order to achieve these profits. As a result, many groups that rely on financial statements, such as investors, lenders, when a firm’s steady, long-term profits suddenly invert to tives for excessive risk taking and, as a result, reduce the credit rating analysts and regulators, can be blindsided sudden, catastrophic losses; reporting risk-adjusted profit, however, highlights the extent to which a firm’s profits derive from taking risk. Neil R. Bodoff, FCAS, MAAA, is senior vice president at Willis Re Inc., in New York, N.Y. He can be reached at neil.bodoff@ willis.com. 48
Slide 49: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications prudent enterprise Risk Management strategy: Managing/Matching Authority and Accountability Recent crisis: Roots and lessons By STEpHEN MITCHEll AND JoHN f. MCgARRy there is no free lunch by dAniel c.f. hui Risk Management – Buyer Beware! by dennis bARRy Risk Management and the financial Crisis: why weren’t we protected? By MIkE BATTy 49
Slide 50: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Recent Crisis: Roots and Lessons by Stephen Mitchell and John F. McGarry Risk management lessons abound from the current crisis. Understanding the lessons requires a clear assessment of the roots of the crisis. A simplified assessment of the underoffers valuable learnings. Roots basis, by three L’s: • • • Leverage The roots of the crisis are understood, on a simplified Lack of accountability Liquidity Lack of accountability is most evident in the fundamental pinnings of the crisis sheds light on pivotal mistakes and The impacts of these changes in the fundamentals of the mortgage market were further magnified by expan- interest-only payment patterns, less money down) and creditworthiness of applicants. Leverage was a second key factor. Institutions guaran- sion of mortgage availability with regard to terms (e.g., teed mortgages against default. Based on historical experience, only small amounts of money, typically measured in terms of basis points, were required by the companies acceptlarge sums of money based on comparatively little income. ing this risk. This left institutions exposed to guaranteeing There were two primary problems with this approach. First, many of the institutions failed to evaluate the potential risk and hold reserves or capital against any material differences from historical default rates and patterns. Second, insuring institutions often failed to recognize the changing dynamics noted above and the conditions they of a downturn in property values. changes in the mortgage market in recent historical periods. mortgages from local institutions that would conduct the evaluation and hold the risk. Loan officers of these institu- Historically, people requiring mortgages sought their tions were accountable for the outcomes of their decisions— defaults were a direct reflection on their judgment. In recent periods, mortgage originators abound. In this model, the created for significantly increased default risk in the event When the first two L’s came together, the third L, originators create the mortgage and then send it to other institutions for holding. Originators are compensated on the end outcome relative to default. volume of mortgages created, but are not impacted by the Historically, people sought mortgages to buy a primary liquidity, became a problem. When property values even- tually evidenced their inevitable fall, defaults occurred at historically high rates. This led those institutions that had ization, to have large and immediate cash needs. Those with cash held onto it, while those without it sold what they could at high cost. A crisis of liquidity emerged. Lessons Separating authority and accountability seldom results in responsible outcomes. Whenever decision-making is dienough “skin in the game” to create incentives necessary vorced from accountability of outcomes, there is often not for parties to expect rational behavior, or at least not to could—pressuring values—and borrowed what little they assumed the risk, potentially without appropriate capital- residence—the clear intent being long-term occupancy. In recent times, based on a protracted period of increasing property values, people sought mortgages to fund investment properties. “Flipping houses” has become a popular income and investment strategy, as well as a national past gages create an updated set of “walk away” rules. While time. As a leveraged investment financing vehicle, mortpeople do not vacate their primary residences in periods of property value declines, investors will abandon their interest (“turn in the keys”) when the investment does not produce value. Walking away is rational economic behavior given the altered nature of the interest of the property and the use of the mortgage. expect historically experienced behavior. When fundamental changes in the mortgage market changed the interest of originators and the “walk away” rules of mortgage holders, the change in accountability had profound results. 50
Slide 51: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Recent Crisis: Roots and Lessons by Stephen Mitchell and John F. McGarry in many areas. We should continue to be on guard for The lesson of accountability has continued applicability resilient under a variety of circumstances and adaptable to change. Focusing on fundamentals never goes out of style. The elements of the financial markets that may not evidence should also look at our own organizations and institutions to service, management and professional ranks have the interest in the outcomes necessary to drive rational behavior. a crisp relation between authority and accountability. We be sure that the individuals we rely on for decisions in sales, evaluation of defaults was guided by history. An evaluation of the changes sweeping the origination of the mortgages and the profile of their holders should have revealed a dif- ferent picture. Institutions that should have been expert at managing financial risk failed to fairly assess and provide provision for adversity. Looking through assessments provided by others and that nothing lasts forever. Markets can begin to behave as if a Secondly, it is a refreshed truism for us from the crisis trend, such as increasing housing values, will go on forever. trends move in one direction is surely doomed to failure. Any system that is geared to function only when values or As we step into the future, we need to be on guard for constantly challenging our understanding of the dynam- ics around us is critical. In a world of increasing volumes of collateralized securities and derivatives, it is hugely important to look beyond the wrappers to the base assets its parts. any system of business or financial strategies or behav- iors that is based on the fundamental tenet that a certain trend will continue indefinitely. The best strategies are both and behaviors. The sum may indeed be different than Stephen J. Mitchell, FSA, is vice president, benefits, ops policy at Unum in Portland, Maine. He can be reached at sjmitchell@unum.com. John F. McGarry, FSA, is senior vice president at Unum in Portland, Maine. He can be reached at JMcgarry@unum.com. 51
Slide 52: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications There is no Free Lunch by Daniel C.F. Hui Investing for insurance companies in the United States and brokers misrepresented or even forged the income level Canada is a balancing act. There are numerous restrictions on allowable investments. Portfolio yield is very important, because insurance companies are relying on the investment returns to supplement the operating income. Since the burst of the Internet bubble in 2001, the interest rates fell and stayed low for a number of years. Finding higher yielding of NINJA (no-income-no-job-and-asset) borrowers. Addi(ARM) and option-ARM attracted many borrowers with tionally, the lower teaser rate of adjustable rate mortgage limited resources to invest in a home. As mortgage rates showing signs of stress. The value of MBS, ABS and reset while interest rates got higher, these groups were CDO with subprime exposure was dropping like a stone. ly, and the flow of securitization deals slowed to a trickle. fixed income instruments is a tough act as highly liquid instruments tend to provide lower yields than longermuch liquidity will drag down portfolio yield. duration assets in a normal interest rate environment. Too Mortgage-backed securities (MBS), asset-backed Eventually, all securitized products were impacted adverseUnderwriting is very important, and actuaries understand securities (ABS) and their derivatives like collateralized debt obligations (CDO) or CDO-squared provided a welcome respite for institutions seeking higher interest income in the derivatives were marketed as AAA-rated instruments with that. There are many players in the mortgage securitization market. The investors, who have a direct interest in the performance of the underlying loans and therefore how process. Andrew Davidson of the Andrew Davidson Com- low interest rate environment. Some of these securities and higher coupons than comparable treasuries. Some firms bought truckloads of these instruments because higher yields were hard to resist, and AAA-rated securities are considered safe investments, aren’t they? these loans were underwritten, are far removed from the pany suggested that there are six degrees of separation. the loans are underwritten. Investors loaded up on AAA- Investors are left without a clear view or control on how rated MBS, ABS and CDO on the belief that the ratings truly indicated the stated level of credit and that the first loss investors had applied due diligences to the investment process. Trust was misplaced. The quality of some of the AAA-rated CDO or morphed into a credit crisis that eventually engulfed a lot of financial institutions around the world. If a AAA were really a AAA, this crisis would probably have been more This crisis first started with subprime loans and quickly contained and localized. As the crisis developed, evidence emerged from the shadows that revealed the questionable quality of some AAA-rated MBS or ABS. Subprime loans were the weakest link. As the credit boom came to an end CDO-squared tranches was questionable as well since BBB ratings could be transformed into AAA using the CDO technology. Indeed, 62 percent of the MBS/ABS tranches that were rated BBB or below were turned into AAA-rated CDO tranches according to an IMF study. This seemed to be a win-win situation for everybody. Now we must ask: Are these securities really AAA-rated? and the economy slowed down further, the Alt-A, prime and credit card loans were also affected. The subprime crisis is only the symptom of a much larger underlying problem. Both consumers and corporations were overU.S. phenomenon. leveraged. In time, we would find that this is not a strictly The foreclosure rate of subprime loans started to in- than in other countries. A glance at the balance sheets of Securitization is more popular in the United States U.S.-based insurers would reveal that this group originates very few, if any, loans now. The Canadian insurers across of their portfolio in private placements and commercial the border, on the other hand, hold a higher percentage mortgage loans than their U.S. counterparts. Further, crease toward the end of 2006. Borrowers with minimal resources were given loans in the credit boom. Mortgage 52
Slide 53: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications There is no Free Lunch by Daniel Hui there are Japanese insurers making residential mortgage loans and commercial loans directly. In addition to buying protection, one of the benefits that policyholders could the investment prowess of insurance companies. Individual investors generally could not access asset classes like assumptions that can be observed from the market. Solvency is the paramount concern of insurers. Conservatism is necessary to ensure solvency. Pricing and valuation funccarried out. possibly enjoy by buying insurance products is to access tions are generally staffed separately to ensure that this is The subprime crisis is only the tip of the iceberg. As private placements, residential and commercial mortgage loans and commercial real estate. This was the competitive advantage that insurance companies once enjoyed. This is not true any more. Some of these asset classes are securitized in the United States today. Insurers are facing stiff home prices continue to sink lower, more homeowners are saddled with negative equity and are opting to walk away from their investments. The trend of increasing foreclosure card default is moving higher too. Consumers are overrates has since spread to Alt-A and even prime loans. Credit leveraged, and firms fare no better. The assets that firms thought they once had have either shrunk or are not there any more. There is not much left on the left-hand side of the competition from other financial institutions. Securitization same time the opaque securitization scheme has made them themselves, they can access the creditworthiness of the investors in MBS, ABS and especially CDO would have no not. There is no gatekeeper guaranteeing the loan quality. has promoted the liquidity in these asset classes but at the vulnerable. When insurance companies originate loans borrowers and have control over underwriting. Now, idea about whether the borrowers could repay the loan or Lack of appropriate modeling capabilities and technol- balance sheet. The corporate default rate is getting higher. CDOs derived from MBS, ABS and corporate securities swaps (CDS) on MBS, ABS or corporates are losing money contain the crisis. are losing values. Therefore, firms that sold credit default fast. This cross-product contagion makes it difficult to The future of the current originate-and-distribute ogy presents additional difficulties for investors. MBS with with multiple sector exposures is even more challenging. Theory and model development has to catch up with the credit default is very complex to model accurately. CDO model is still up in the air. It is clear that securitization has promoted liquidity and enhanced mortgage financing, so together. There are proposals to shed more light on the any case, the lessons for investors are clear. There must be a gatekeeper to assure the quality of underlying loan it is difficult to imagine getting out of securitization all opaque structures and make them more transparent. In trades happening in the marketplace. There is no recogderivatives. Sell-side firms have invested a lot in this already. Unfortunately, buy-side firms are lagging far behind. nized and tested model for valuing these securities and still after the credit crunch. Valuing these in an illiquid market is very challenging. Interestingly, actuaries have been valuing illiquid insurance liabilities for decades. It The market for MBS, ABS and CDO came to a stand- portfolios and take responsibility of the underwriting in loan origination. Models, technologies and risk management tools must be strengthened. Investors will need to invest in the capability of analyzing structured financial securities. After all, there is no free lunch. is recognized that assumptions are very important in the valuation process. There are two natural sets of assumptions —a realistic set for pricing and a conservative set for The author wishes to thank James Chang, FCAS, and Paula Kwiatkowska, FSA, FCIA, FIAJ, for reviewing and providing feedback to this article. valuation. This is because there is not a single set of Daniel C.F. Hui, MBA, FSA, FCIA, CFA, is senior quantitative analyst at American International Group (AIG) in New York, N.Y. He can be reached at daniel.hui@aig.com. 53
Slide 54: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Risk Management: Buyer Beware! by Dennis Barry “Beware of geeks bearing formulas.” —Warren Buffet, 2008. “If it looks like a loan, acts like a loan, and perprobably a loan.” forms (at least in the beginning) like a loan, it’s —Paraphrased from a truism about waterfowl, origin unknown. There are certain aspects of the world of finance, and If Institution C, and all the other institutions that end up with pieces of the loan originated by Institution A repackage all or part of their respective shares, recharacterize the resulting whatcha-ma-call-its, rename those same whatcha-ma-call-its, and resell them to Institution D, the and possibly more. same amount of capital is still required to support that loan, You get the idea. No matter how the original loan is more particularly of the world of loans, that seem obvious if only to the non-professional. For example, any time you lend money, to anyone, there is some risk that you might the same amount of capital is still required to support that capital may be different, but the total amount of capital aswhere in the chain, one of the parties guarantees the performance of all or part of the original loan, no matter how probably more, is required to support the loan. loan, and possibly more. The provider(s) of the necessary sliced, diced, packaged, wrapped and maketed by anybody, not get it back. With the possible exception of full faith and credit debts of the United States, every loan carries some risk with it regardless of the borrower, and this risk varies sociated with the original loan cannot be less. And if, some- depending on the creditworthiness of the person or institu- tion that is supposed to repay the debt. Bearing that risk requires some sort of capital, both as a legal requirement and as evidence of good sense in lending, but recently we seem to have forgotten the basic relationship between a it should work: loan and the capital associated with it. So, let’s review how If Institution A makes a loan, a certain amount of capital the guarantee is structured, the same amount of capital, and Somewhere in the credit default swap assembly line, that little piece of fact was lost, ignored or redefined out of existence. The same is true for the mortgage-backed securities issued by Freddie Mac and Fannie Mae, although in that case perhaps the existence of such a strong connection to the federal government made security seem stronger than it security (being backed by mortgages), is only as good as the underlying assets—garbage in, garbage out, as the is required to support that loan. The amount of capital should be related to the likelihood of repayment of the loan. really was. But, at the end of the day, a mortgage-backed regardless of how that offload is structured, the same If Institution A offloads that loan to Institution B, amount of capital is still required to support that loan, and but the amount cannot be less than it was before the loan was transferred. If Institution B breaks the loan up into little pieces saying goes. The same can be said for securitized credit card loans, and so on. And if a counterparty that has guaranteed turns out to not have the wherewithal to make its promise come true, the whole chain comes apart, as we’ve seen. The lesson of the current debacle is, or certainly ought performance of the original loan at the end of this chain possibly more. The provider of the capital may change, and packages the pieces with pieces of other loans it has acquired and then sells the package to Institution C, the same amount of capital is still required to support the original loan, and possibly more. Again, the capital provider(s) may change, but the overall amount cannot be less. to be, that the risk inherent in a financial transaction cannot be made to go away by the mere act of repackaging that cess to write rotten loans and then have the risk disappear via the vehicle of someone renaming those loans leads to transaction and renaming it. For the front end of the pro- ultimate financial ruin, as evidenced by what’s going on 54
Slide 55: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Risk Management: Buyer Beware! by Dennis Barry now. The sow’s ear is still a sow’s ear, despite all attempts to make it otherwise. If the current regulatory structure, coming from whatever regulator(s), allowed some or all of the capital needed to support a loan to disappear just that is a major regulatory failure. And if current audit procedures and debt-rating processes allowed the same thing, those are failures as well. Whatever regulation comes out of this mess should because the original loan has been repackaged and renamed, stupidity (aka irrational exuberance) without becoming a That’s not a practical regulatory solution for anyone. surrogate, hands-on manager for all regulated institutions. Regulation that requires conservation of capital, and reserves if appropriate, for financial transactions would be a strong first step in preventing a recurrence of what’s going on today. No restructuring or recharacterization of a loan, or other transaction, downstream can eliminate the need issues, they need to be sorted out as well, but adequate capital for financial transactions must be the bottom line. A second step might be to require the institutions for the capital originally required. If there are jurisdictional focus on this fact. In the scenario above, Institution A should not be able to release the capital attached to that loan until it has regulatory proof that Institution B has that amount of tion C and Institution D and so on all the way through the capital set up for the loan, or more. Similarly for Instituchain. Moreover, if the repackaging of the original loan, in whatever form, adds risk to the overall chain, then the appropriate level of additional capital must be held as well. This sort of regulatory approach already exists in the involved to limit incentive compensation to time frames would not be risk-free, it is clear that one-year incentives unhealthy for all of the institution’s constituents. It would of at least three years. While any arbitrary time frame can reward behavior that, over time, might turn out to be be better to remove temptation, to the extent possible, world of insurance. A direct writer may cede reinsurance reinsurer puts both on its own books and is allowed by without simultaneously removing all incentive for a company to grow profitably. Three years seems like a suitable period. A third step, admittedly hard to precisely define before- but may not release either reserves or capital unless the regulators to do so. If retrocessions are involved, each ceding company must retain whatever portion of the necessary reserves and capital unless the accepting company agrees to take on the risk, and is legally recognized hand, should bring the auditors and rating agencies into the equation as well. It is hard to imagine that AIG deserved a clean audit opinion as of December 2007, and was broke by the end of September 2008. The same can be said for Merrill Lynch, and many, many others. And the same is true for many securities that were highly rated when they were as being permitted to do so. If a reinsurer is not legally permitted to accept a particular piece of reinsurance, other means of security must be provided before the ceding approach makes sense and works well, but it doesn’t make the system foolproof. There are lots of horror stories company can be done with that portion of the risk. This in fact rotten. No doubt the courts will determine culpability in what has happened recently, but for the future responsibility ought to be clear, unambiguous and enforceable. We began this essay with a couple of quotes that seemed involving complex reinsurance pools that failed. Risks sometimes the accepting company isn’t sure of what’s going on upstream. However, that’s a management failure, appropriate to the topic. Let’s end with one that summarizes the current situation, succinctly and accurately: “We have met the enemy and he is us.” —Pogo the Possum (Walt Kelly), 1970. get pooled and offloaded in all kinds of packages, and not a regulatory failure. It’s almost impossible to regulate Let’s hope we can do better in the future, for all our sakes. Dennis R. Barry, FSA, MAAA, is principal at Barry Consulting Services in Little Rock, Ark. He can be reached at fbarry9@ comcast.net. 55
Slide 56: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Risk Management and the Financial Crisis: Why Weren’t We Protected? by Mike Batty The Failure of eRM? Risk management can be a thankless profession. In bull who, as some might say, “take away the punch bowl just when the party starts getting interesting.”1 Upon a turn for markets, risk managers are often viewed as wet blankets coming? Low interest rates and regulations promoting housing, but the downfall of risk management was also a powerful force. Due to the explosion of the originateto-distribute business model (fueled by the growth of securitization), underwriters of suspect home loans were freed from significant responsibility for whether the questionable loans onto highly leveraged investors and home ownership played a role in the growing demand for the worse, people wonder why they weren’t warned earlier. Even when successful risk managers limit losses, their recognition is somehow lacking. Maybe it’s the loss aversion comfort in situations that are merely bad, rather than terrible. loans could ever conceivably be repaid. They passed the then focused on their core competency, making more loans. As the demand for these mortgage-backed securities skyrocketed, underwriters tapped pools of more and more suspect borrowers. baked into our psychology, but it can be difficult to find The ongoing financial meltdown has cast a shadow over the entire economy, and caused some to question whether (ERM) has failed. My answer is, yes, an overarching theme the much-hyped movement of enterprise risk management of the credit crisis is a failure of risk management. Howers themselves, I look to the breakdown of the entire risk incentives to monitor their quality, but why didn’t inves- So the people making the loans lacked the proper ever, rather than placing all the blame on the risk managmanagement system. Instead of revealing a fundamental tors impose market discipline on the lenders to make better loans? For some investors, the simple answer seems to be they didn’t realize how risky these assets were. No risk on the rating agencies’ seal of approval. However, many flaw in ERM that will banish it to the annals of academic management systems were in place, and they instead relied sophisticated investors did utilize complex financial models to measure their risk, and were no more successful insulating themselves from loss. It wasn’t that these models were research, I believe the single most important message from management capabilities. The Making of a Crisis The story of what has gone wrong with the current financial situation is lengthy. In general, the credit crisis this financial situation is the need for vastly improved risk incorrect, per se, but they did provide an unrealistic picture of risk. It’s true these financial securities are opaque, but with correct assumptions of home price declines and the severe constriction of credit that ensued, the models do The models went astray largely because the likelihood provide a mathematically correct description of the crisis. of these market conditions was not given due weight. before deriding the modelers too harshly, let’s try to under- revealed a lack of enterprise-wide risk management, a an outright disregard for well-informed risk managers. failure of risk management techniques and, in some cases, We know the problems began in the housing market. After 2003, when the prices of homes began to deviate Undoubtedly, there was widespread error in judgment, but stand the difficulty of modeling rare events. Home prices have not fallen this sharply since the Great Depression. Given that the financial landscape is so radically different from historical relationships with inflation, income and productivity, a time bomb was created for their collapse. But how did this develop? How did these problems so permeate the global economy? And why didn’t we see it 1 William McChesney Martin, chairman of the Federal Reserve Board (1951–70), describing the role of the central bank. 56
Slide 57: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Risk Management and the Financial Crisis: Why Weren’t We Protected? by Mike Batty today, and we understand much more about how the on 75-year-old events? There is no definite answer, and modelers will continue to struggle with this question, but a clue can be taken from emerging behavioral economic research that indicates people tend to systematically under- economy functions, how much weight should models place beliefs when the actions of others indicate a completely different view. Improving Risk Management While there’s no panacea for risk management, I believe there are steps we can take. One of the likely benefits of this crisis will be greater appreciation for risk management, both by managers and investors. Many CEOs have taken the fall for their firms’ poor performances. in the future, ultimately they are agents of investors, and incentive for risk management should come from this diffuse group. More thoughtfully designed, shareholderapproved pay packages that incorporate risk-adjusted performance measures (likely measured over time) can send important messages to firm managers about the risk level with which they are comfortable. estimate the likelihood of rare events. No doubt this has contributed to the running financial joke that a once-in-acentury event occurs every few years. or faulty assumptions. Though in the minority, a number troubles several years ago. Why were they largely ignored? Still, this is not only a story of a lack of information, While executives have an interest in avoiding large losses of risk managers and economists warned of the impending Many would answer, “greed and arrogance,” without hesitation. Those traits certainly played a role, but I believe of a bubble is very difficult to defeat. Standing up for the there is another, underappreciated cause. The psychology contrarian view requires extreme fortitude in a world of cheap credit and seemingly riskless return. Whether the collapse in home prices was inevitable, merely likely regulate financial markets made mainstream by former The libertarian idea that rational self interest will Federal Reserve Chairman Alan Greenspan is compelling, aimed at preventing the powerful from taking advantage were caused by people systematically acting in opposi- or rather a realization of a rare event is open for debate. John Maynard Keynes tempers its significance, “The marHowever, the famous quip by the legendary economist but even he now admits a fatal flaw. Many regulations are of the weak, but the real problems with the credit crisis tion to their long-term self interest. Predatory lenders did lending businesses no longer exist. Borrowers did lie about Imagine the unenviable positions of bankers, investors and mortgage lenders who suspect the tenuous nature of their many who are engaging in the risky behavior. Some execusituation, but face immediate pressure to compete with so tives likely faced the option of a) holding their ground and holders), or b) going along with the trend of leveraging were not realized. In these situations, it can be rational to ket can stay irrational longer than you can stay solvent.” take advantage of naive borrowers, but by and large these their incomes, but were the years spent living in nicer homes question of why such large scale departures from rationality than they could afford worth the pain of foreclosure? The risking removal immediately (by either their boss or shareinvestments as much as possible, and hoping their fears abandon your principles and take the risks. The decision occasionally occur is fascinating, and far too complex to address here. Suffice it to say the purest form of Homo Economus is a myth, and while we should respect the right of people to make some foolish decisions, regulations are to their effectiveness is the structure of regulations. Tighter to heed the warnings of the risk managers, however believable, is much more difficult (and risky in the short term) than expanding with the bubble as it inflates. In addition, necessary to limit this type of systematic failure. Paramount restrictions on making loans may have prevented this crisis, but will those rules prevent future crises? More likely even the brightest minds can begin to question their own 57
Slide 58: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Risk Management and the Financial Crisis: Why Weren’t We Protected? by Mike Batty than not, future crises will look different, exploit other regulatory loopholes or result from new products that aren’t sufficiently covered under current regulations. Designing an effective system is a difficult task, but a principle-based approach to regulation focusing on the enterprise-wide risk position of a firm, and the role the firm plays within the These ideas are emerging in talks of regulatory reform in of rating agencies. Hastings in 1066 (even more than since the Great Depression). Relevant historical data to model losses with such precision is generally unavailable. Models can and will get better, but mathematical sleight-of-hand can only go so far in overevents always in question, and knowing our inherent biases in assessing them, we may find it beneficial to downplay as is often the goal. A lot has changed since the Battle of economy, could help limit systemic problems in the future. Europe and the United States, and are an increasing focus While more risk management is needed, ignoring its coming a lack of data. With the likelihood of extremely rare the role of tail probability in our analysis, and instead ask questions such as: Are we comfortable with the knowledge that such scenarios might occur? How can we mitigate the risk? How should we react if those situations begin to play but we will get better. out? Again, improving risk management will be difficult, inherent limitations is, in itself, a risk. Mathematical modthe worst loss that might occur in one year out of 1,000, els are quite useful, but consider the scale of determining Mike Batty, FSA, CERA, is actuarial and insurance solutions consultant at Deloitte Consulting LLP in Minneapolis, Minn. He can be reached at mbatty@deloitte.com. 58
Slide 59: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications prudent enterprise Risk Management strategies: Managing liquidity Modeling and Managing liquidity Risk by GARy G. venteR 59
Slide 60: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Modeling and Managing Liquidity Risk by Gary G. Venter Much of the current crisis can be traced to models that failed to adequately reflect risk, both in housing costs and complex financial instruments. Even if historical homerecently, data from other bubbles, from tulip bulbs on, Typical ERM modeling emphasizes the risks to asset and liability values, but the current crisis has made it clear that liquidity risk has the potential to sharply undermine a company’s financial position over and above price risks. well as the existing price and value risks. price data had never recorded changes like those realized could have been used. It was not clear that housing was in a bubble, but bubble scenarios should have been in the models. Those model issues need to be, and are being, addressed, but here the focus is on liquidity risk. Future ERM modeling will have to address liquidity risk as Basic liquidity risk is the chance of not having the funds available to pay liabilities due. But being forced can be magnified by market price disruptions, and these ef- Regardless of the underlying causes, liquidity problems to post collateral could be another type of liquidity risk, speaking, realizing losses because of forced sale of immature assets, and even loss of investment opportunities even if that collateral is technically an asset. More broadly fects should be included in risk models. Such modeling needs “liquidity black holes” as arising from price movements and common trading strategies of short-term investors: “liquidity black holes have the feature that they seem to gather momentum from the endogenous Rather like a tropical storm, they appear to gather more energy as they develop. Part of the explaresponses of the market participants themselves. to postulate a mechanism. Morris and Shin (2004) model due to cash constraints, could be included under the rubric of liquidity risk. With a severe market disruption, liquidity problems can be exacerbated when normally liquid assets model scenarios. become illiquid. These possibilities can all be reflected in Liquidity management has features in common with capital management. Maintaining a stock of liquid assets can provide a liquidity cushion. Also matching cash flows nation for the endogenous feedback mechanism undergo changes when prices change. Market distress can feed on itself. When asset prices fall, are induced to sell. But this selling pressure sets off lies in the idea that the incentives facing traders of assets and liabilities, or at least some portion of them, can help manage liquidity risk. Contingency funding plans are a useful part of liquidity risk management as well, where less liquid assets are to be used as loan collateral. However this strategy may fail to work under a market disruption unless lines of credit are secured in advance, as even collateralized some traders may get close to their loss limits and further downward pressure on asset prices, which induces a further round of selling, and so on. Portfolio insurance based on dynamic hedging rules is perhaps the best known example of such feedback.” loans may become unavailable. Specifying the liquidity of various assets and liabilities in the model formulation and evolving them over time can incorporate liquidity into risk scenarios. For instance, Das and Hanouna (2009) discuss a few measures of liquidity. Property-liability insurers use reinsurance as a cost- use of dynamic hedging, but not recognizing such mecha- Certainly market disruptions predated the widespread nisms can overstate the protection these strategies provide, efficient substitute for capital, but in a disruptive event, and result in more reliance on them and an understatement of the risk of increasing leverage. Dynamic hedging strategies need to be updated to include the possibility that the called-for trades cannot be completed as prescribed. reinsurance prices and availability can change sharply. Management of this risk could include having contingent capital sources in place, as well as including reinsurance terms that can expand coverage, such as additional rein- 60
Slide 61: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Modeling and Managing Liquidity Risk by Gary G. Venter statements. Again dynamic modeling should include the possibility of such liquidity issues arising. Modeling liquidity risk can start with stress tests. work of timeline simulation, where events are simulated What is now important in models is to have scenarios in order of occurrence and time stamped (see Kreps, 2009). and responses take into account the possibility that other players are following the same strategies; liquid assets may be triggered, etc. Models for these possibilities and the become illiquid; off-balance-sheet commitments might interaction of price and liquidity are appearing in pubfor liquidity and corresponding pricing events could be an area of research for some time to come. References Kreps, Rodney. 2009. “Theory and Practice of Timeline Simulation,” Variance 2:2, forthcoming. Das, Sanjiv and Hanouna, Paul. 2009. “Hedging Credit: The current market is one example of a stress scenario. A convergence of adverse asset, liability and credit availability situations can be postulated and the cash flows projected along with the value changes. Probabilistic scenario generation requires asigning probabilities to the stress scenarios and including them in a larger market dislocations, such as Morris and Shin’s, can help simulation. Having a model that predicts occasional incorporate liquidity events in the scenarios. Certainly there is an interaction between price movements and liquidity movements that can be taken into account. Such modeling can quantify the impact of liquidity lished theory, but nailing down reasonable probabilities risk on capital adequacy. Part of the problem is recogniz- ing off-balance-sheet cash needs that can arise in a market disruption, such as collateral requirements, embedded options, refunds due to ratings down-grades, etc. This models that include response strategies to various events. Equity Liquidity Matters,” Moody’s & Stern 5th Annual Credit Risk Conference and Journal of Financial Intermediation, forthcoming. also emphasizes the utility of dynamic ERM models— Dynamic ERM models can also benefit from the frame- Morris, Stephen and Shin, Hyun Song. 2004. “Liquidity Black Holes,” Review of Finance 8:1, pp. 1–18. Gary G. Venter, ASA, FCAS, CERA, is managing director at Guy Carpenter & Company LLC in New York, N.Y. He can be reached at gary.g.venter@guycarp.com. 61
Slide 62: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications societal themes: limited liability laws An ideal crisis by shAne WhelAn 62
Slide 63: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications an Ideal Crisis by Shane Whelan Risk modelling is a risky business, but the burden of risk mainly funding his activities from bank loans; to the buy- model failure is often borne by society in general rather than the firm in particular. This division of the ultimate cost ensures that risk models systemically underestimate the risk, as they are designed to capture only that part of the risk borne by the firm. In short, the risk models that underthat entail a lower return on their increased capital. to-let investor or owner-occupier almost entirely funded by banks; and, to the banks themselves whose liability is underpricing of risk in the property market bursting the banks that were meant to hold it back in the fall of 2008. limited to their capital base. And so we have the systemic estimate risk will drive out the more reliable risk models The underlying dynamic is simple. Consider Firm X lenient bankruptcy laws and limited liability allowed to therefore the inevitably of such episodes. According to this explanation, the world can point its finger at the United States who, first amongst nations in modern times, allowed Ever since debtor prisons were abolished in favor of that puts all its capital, made up of 50 percent equity and 50 percent of borrowings, into a venture that has, say, a 50 percent chance of doubling the investment and a 50 percent chance of losing it all. The expected payoff of the invest- firms, society created the incentive to misprice risk and ment is the sum of the probability of each outcome times its payoff. In this example the expected payoff is simply the return of the original investment (that is, 0.5 times twice the outcome for the firm’s equity holders: their expected payoff unrestricted limited liability to firms from 1811 (beginning in New York state) and, from 1833, began repealing harsh treatment of defaulting debtors. Even today the United States remains to the fore with some of the most lenient bankruptcy laws in the world. This analysis is, however, only part of the explanation for the current system failure, capital plus 0.5 times nil). However, that is not the expected is one and a half times their original investment, (calculated as 0.5 times [four-times the equity holder’s original investment less loan of once their investment] plus 0.5 times nil). venture as it amply rewards their portion of the risk, even though it is not rewarding the overall risk run. and the proposal to repeal the laws is perhaps not the least costly solution: such laws arguably enabled the emergence of modern innovative economies. Modern economies are based on the premise that all The equity holders are clearly incentivized to invest in the the main players look after themselves. The bankruptcy and limited liability laws gave property speculators and banks a put option on society so they could walk away from one material aspect: the risk could be quantified precisely. In practice, payoffs of ventures in the real world cannot The stylized example above is oversimplified in just losses above their capital base, yet enjoy all the gains of such speculation. They acted in what they believed were their own interests. It seems that society—well aware of what was happening—did not effectively look after its own interest and now must pay the price. Society, of course, appoints a financial regulator to look be determined, as Keynes famously remarked, by “strict mathematical expectation.” This observation means that guesses. And it is the firm, and its risk models, that are the odds must be regarded as guesses—at best educated regarded as providing the most educated guesses as it is in their chosen specialty. Incentives to bias risk measurement for those most expert in measuring it can be expected from firms to be mopped up by the rest of society. after its interests in this regard. The aim of regulation is designed to keep the probability of insolvency sufficiently to lead to recurring disasters as risk periodically leaks out The simple model applies to the property developer, low so that the direct and indirect damage caused by insolvency is set equal to the broad social ills of an inefficient overcapitalization of the industry. The expected payoff to the shareholder, when the financial regulator understates 63
Slide 64: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications an Ideal Crisis by Shane Whelan the real probability of default, is increased at the expense of society as our example showed; as the shareholder maintains greater exposure than is reasonable with the risk capital employed. The regulator must ensure the shareultimately be borne by society (so regulation is designed to assets does not exist (that is, the sample kurtosis will tend to infinity as the sample size increases) and that therefore; the VaR, and the extent of the expected loss once the VaR was exceeded, was considerably higher than was previously believed (see Whelan (2003), Chapter 4, for an overview of the literature). The closer we look, the bigger the investment risk appears. Simultaneously, there were some tell-tale signs that the financial services industry was coming to appreciate the magnitude of the risks and had busied itself over the last couple of decades in passing on sphere of an actuary’s influence, investment guarantees holders factor into their decision-making the risks that will “internalize the externalities”). This requires a reasonably accurate model of the behavior of the extreme left tail of outcome distributions and ensures, in the current case, that banks are suitably capitalized. The current dominant methodology for banks assessing investment risk to where it is least appreciated. Within the on pension and life products were withdrawn or reduced, defined benefit schemes were wound up, risk was transferred to members via defined contribution arrangements, exposure (the development of so-called “finite” reinsurance). their capital needs (or, indeed, hedge funds) is based on estimating the value-at-risk (VaR)—that is, the capital that is needed so that the probability of needing more is sufficiently low (say, 1-in-100 or 1-in-1,000 chance). For speculative assets, one typically fits a distribution to possible returns and works out the implied VaR from this. It is well-known that the Normal distribution tends to underestimate the proba- and even reinsurers began setting limits to their ultimate So, according to the assessment above, one might bility of extreme movements (the failure of the hedge fund, of this; see Jorion (2000)). A more common approach of late is to fit a Student t-distribution to historic returns with conclude that actuaries should get higher marks than bankers for their arithmetic. Yes, but society is not primarily concerned with who gets their sums right. Keynes knew, and the limited liability and bankruptcy laws enshrine the view, that getting the sums wrong is often better: “it is probable that the actual average results of Long Term Capital Management, was a dramatic reminder the degrees of freedom selected so that the kurtosis of the Student t-distribution matches that of the sample kurtosis (see, for example, Jorion (2002)). This typically produces a higher VaR, but still appears to understate the true risks run. Berkowitz & O’Brien (2002) studied how risk models investments…have disappointed the hopes that prompted them… If human nature felt no tempta- employed by six large multinational banks performed in the VaR and raised the concern that such occurrences may tion to take a chance, no satisfaction (profit apart) practice. They reported that losses can substantially exceed be correlated across the banks—indicating the possibility to such extreme falls. in constructing a railway, a mine, or a farm, there of cold calculation.” might not be much investment merely as a result The world banking crisis allows us to point the finger at of a systemic risk across the banking sector when it comes There were, of course, many more warnings that banks the bank regulators who got it wrong by failing to enforce capital requirements commeasurable with the risks run. No disapprobation applies to the pension and life assurance regulators who allowed actuaries get their sums right and quietly pass on the risks to individual savers. But which leads to the greater cost to society? To solve the banking and other financial companies were significantly understating the risks in their portfolios. A growing literature was showing that the kurtosis of the return distribution of speculative 64
Slide 65: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications an Ideal Crisis by Shane Whelan crisis, each economy must now redistribute the losses to those that can bear them by some mechanism or other. Jorion, P. 2000. “Risk Management Lessons from Longment 6: 277–300. However, it is difficult to envisage a solution to the greater misallocation of risk in society—there is unlikely to be an individual pensioners who learn too late the true cost of investment risk. The economic system that has developed over the last acknowledged crisis and certainly no bail-out of all of those Term Capital Management.” European Financial ManageJorion, P. 2002. “Fallacies in the Effect of Risk Manageview13 (Dec.): 115–127. ment Systems.” Bank of England, Financial Stability ReKeynes, J. M. 1936. The General Theory of Employment, Royal Economic Society. couple of centuries comes with embedded periodic crises Interest, and Money. The Macmillan Press, London, for the Whelan, S.F. 2003. Actuarial Investigations in Irish Capital Markets. Unpublished Ph.D. thesis. due to its inevitable mispricing of risk. That is our system, and it is the best yet devised. What we can do is choose the type of crisis we get. The current loud global banking crisis, insisting on the simple if unpleasant measures, is altogether pensioners, isolated and ignored in their increasing poverty. Selected References Berkowitz, J. and O’Brien, J. 2002. “How Accurate are Value-at-Risk Models at Commercial Banks?” Journal of Finance 57: 1093–1112. more preferable than the future silent problem of individual Shane F. Whelan, Ph.D., FSA, FSAI, FFA, is a lecturer at the School of Mathematical Sciences, University College Dublin in Dublin, Ireland. He can be reached at Shane.Whelan@ucd.ie. 65
Slide 66: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications societal themes: Markets: Regulation of, efficiency of, interconnectivity of transparency and liability valuation By pHIlIp E. HECkMAN creating an exchange for insurance contracts By oAklEy E. (lEE) VAN SlykE The game of ‘pass the Risk’: Then and Now By Joy A. SCHwARTzMAN from liquidity Crisis to Correlation Crisis, and the Need for ‘Quanls’ in ERM by stéphAne loisel 66
Slide 67: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Transparency and Liability Valuation by Philip E. Heckman The most notable thing about the current crisis in the financial markets is the nature of the instruments that caused the trouble. The subprime mortgage debacle would standards are intended as a means of conveying timely, ac- have been bad enough by itself, but it has been aggravated out of all proportion by marketing of mortgage obligations as CDOs, sliced and diced in backroom chop shops, blessed by the laying on of hands by the rating agencies and sold of these instruments was strictly hands-off. Similar things can be said of the rogue portfolio of credit default swaps The fair value initiative was put forward by FASB and curate and relevant information to the investment markets. IASB in order to advance this goal. Does it have a prospect to-market accounting in the banking community concerns which market values are unknown or erratic. (As an aside, before we talk about liabilities, I suggest that this outcry, arising from the subprime crisis, would have been nipped in the bud and confined to a narrow sector if only CDOs were of succeeding? The current outcry against fair value markmainly the valuing of assets with impaired liquidity, for over the counter to the unsuspecting. In fact the regulation that brought about the near-collapse of AIG. The common thread here is that all these factors conspire to confound to the originators, in a word, to undermine transparency. scrutiny and to frustrate attempts to trace obligations back Supposedly the regulatory vacuum surrounding these traded on exchanges.) Apart from problems with asset volatility, liabilities add a whole new layer of confusion to the fair value puzzle, which so far has attracted little notice. The central concept of fair value is to record values instruments, imposed by law in the case of CDSs, was intended to permit the “free” market to work its magic without interference. Here we are led to draw a distinction between “wild” markets and “free” markets. A wild market purposely impeded for competitive reasons and reduced is unregulated and unscrutinized. Information flows are to trickles from rumor and espionage. No one knows what such a market, there are no safeguards against anticompetitive behavior and no guarantees that the market will clear. The only guarantee is that there will be liquidity crises. When economists claim almost mystical advantages for assets and liabilities which are as close as possible to ket, supposing one existed. IASB and FASB (abetted by the values these instruments would have in an open marsome in the financial economic community) recognize no difference in the valuation bases for assets and liabilities: liabilities are recorded at the current market price payable sequences. For instance, a company can reap profits from by the beneficiary of the contract. This has perverse cona credit downgrade, which leaves it free to write down its anyone else is doing, and pricing is blind and haphazard. In liabilities because it is now less likely that it will be able to honor them. A dramatic instance is provided by Radian Group in first quarter 2008, where a $215 million loss was Street Journal, May 19, 2008, p. C12.) Radian made full turned into a $195 million profit. (David Reilly, The Wall for “free” markets, are they talking about the wild mar- kets described above? Hardly. All the empirical evidence comes from exchanges. Only exchanges produce the quanti- for the virtues of free markets, market efficiency foremost, tative data needed to support such conclusions. Only markets disclosure of this oddity, but it was under no legal or regulawindfalls might be less forthcoming. tory obligation to do so. Other companies enjoying similar This opacity in the accounting for liabilities did not where transactions and valuations are disclosed timely and market. There is no efficiency without transparency. accurately can enjoy the advantages claimed for the free So what does all this have to do with accounting stan- begin with fair value but is as old as the discipline of accounting itself. In accounting for debt, it has always been liability and to amortize on a fixed schedule using the implied the custom to record the proceeds of the loan as the initial interest rate. The change introduced by fair value is to take dards for liability valuation? Although many think of them only as a source of operating costs, financial accounting 67
Slide 68: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Transparency and Liability Valuation by Philip E. Heckman liability and running the changes through income. This is precisely what produced the Radian anomaly and drew attention to the defects of the entire scheme of accounting for liabilities. account of changes in market interest rates by revaluing the regardless of whether the market is functioning properly or not. It seems that an enterprise accounted for as a going well known future cash flows discounted at rates which concern should be able to record assets with reasonably exclude illiquidity penalties. I do not know of a widely accepted analytical approach for filtering out the effects of dust it off; else, now is the time to invent one. illiquidity on market prices. If one exists, now is the time to As I noted above, the financial accounting system is an alternative. Professors Chasteen and Ransom of Oklahoma State University propose a revolutionary schema (Accounting Horizons, July 2007) for rationalizing the In point of fact, the accounting discipline has produced accounting for liabilities. Their approach rests on the realization that the insolvency put (the value of the corporate owners’ immunity from recourse in the event of default), itself. In the case of obligations certain as to amount and timing, they propose recording the liability at the appropri- intended to enhance transparency in the financial markets. The fair value reform was intended to further this goal. In its present state, it does not. “Mark to market” assumes trade in the same market. They would not. Liabilities would the cost of surety (the insolvency put again) to the asset is a benefit directly to the owners and not to the enterprise that, if liabilities were traded, assets and liabilities would trade in a market where the price is determined by adding price. The current formulation of fair value is too simplistic ate risk-free rate. The difference between this value and the actual proceeds (the share of the insolvency put pertaining to the liability) would then be recorded as a direct charge against equity. Future changes in valuation due to changes the effect of changes in credit standing would be charged by half, and the result is not transparency but opacity. It will not serve its purpose until accounting for liabilities is rationalized and means are provided for dealing with market pathologies. in the risk-free rate would be taken through income, while directly against equity. This approach would lead to a balance sheet transparent as to the enterprise’s actual financial obligations and would provide directly useful information to the capital markets. Further, it would conform to the going that the enterprise will, in fact, honor its obligations. listen carefully to discern whether he is advocating When someone extols the virtues of “free markets,” disciplined, efficient, transparent and orderly markets or opacity rules, anything goes, ample scope is provided for deeds done in darkness, and epic train wrecks are inevitable. no marshal in town, the bad guys take over. concern assumption of financial accounting, since it assumes Does the going concern assumption cut both ways? whether he is making a sly appeal for laissez-faire where There has been considerable anguish over the fair value requirement to record assets at current market value We should heed the lesson of Blazing Saddles: when there’s Philip E. Heckman, ACAS, is president of Heckman Actuarial Consultants Ltd., in Park Ridge, Ill. He can be reached at peheck@aol.com. 68
Slide 69: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Creating an exchange for Insurance Contracts by Oakley E. (Lee) Van Slyke Observations As risk professionals, we know that managing leverage is the heart of risk management. Leveraged businesses naturally tend to use excessive leverage. We deal with regulations that exist in order to dampen the tendency among insurance companies to become overextended during the underwriting “soft market.” We observed that hedge funds, investment the credit boom went on. We observed an attitude of laissez faire rather than a call for regulation. As risk professionals, we know the importance of As risk professionals, we know that wise insurance regulation has often led to the merger of a poorly managed book of business into a well-managed company. Bankruptcy is seldom the best option. As risk professionals, we know ity and service. Most of this cost is borne by customers also that regulations have a cost in terms of productivbecause it can’t be passed on to shareholders. Customers vency protection and better service. The best regulation provides enforceable contracts, criminalizes fraud and minimizes bad information. Insurance contracts are living documents. Many are banks and private equity funds used excessive leverage as also get most of the benefits of regulation, including sol- personal underwriting. During the credit boom, lenders did little or no personal underwriting. Lenders relied on credit were laying off to investors. The rating agencies in turn used technical measures of risk rather than evaluating each package of loans individually. ratings, especially on ratings of the packages of loans they endorsed, renegotiated, cancelled mid-term or subject to likelihood that at least thousands of subprime mortgages audit. In the credit boom, on the other hand, despite the would default, lenders did not designate people or agencies to renegotiate the terms of loans. Foreclosure was the presumed outcome, and it became the only outcome even when foreclosure was not in the financial interest of the lenders. of managing leverage. Investment models that relied on ratings assumed that those ratings were bets on independent events. Actually, all the risk events were linked together. The ratings shared the same defects and were subject to a As risk professionals, we know the importance capital markets. Some governments such as Iceland and resort or investors in banks. As risk professionals, we appreciate the importance of engaging the world’s capital one that can be adopted globally. markets, the more directly the better. The best solution is As Steven Cecchetti, now the chief economist of the No government commands all of the resources of the Switzerland have quite limited resources as lenders of last single point of failure. Investors and rating agencies failed short, ratings failed to assess risk. to recognize the increasing risk of increasing leverage. In As risk professionals, we know that outcomes don’t els are wrong, although some are useful. We appreciate and selecting an estimate that makes sense in light of all of fit simple models. Indeed, our training suggests that all mod- Bank for International Settlements, has pointed out, “The difference between futures and swaps is that futures are This distinction explains why Amaranth’s failure provoked a yawn, while LTCM’s triggered a crisis. It suggests that standardized and exchange-traded through a clearing house. the importance of using a variety of valuation approaches the results. When values must be set using algorithms, we encourage algorithms that blend the estimates of several and calibrating them with actual data derived from many regulators, finance ministries and central bankers should be changes as possible. This should be the standard structure in financial markets.” models. We respect the importance of testing those models years’ experience. We observed in the current crisis the banks placed excessive reliance on a single simple model, often called value at risk (VAR). pushing as many securities on to clearing house-based ex- Crowds,” has pointed out, the presence of a well-respected 69 As James Surowiecki, author of “The Wisdom of
Slide 70: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Creating an exchange for Insurance Contracts by Oakley E. (Lee) Van Slyke company on an exchange does not mean that that exchange is for every company. Companies that raise capital for nonfinancial activities by selling bonds and issuing stock are ever, companies that rely on or issue financial guarantees (such as investment banks and financial guarantors) take on additional risk when they issue stock because a general loss of investor confidence will reduce the fair market value of of investors to hold their stock. their assets at the same time that it reduces the willingness That is, we are cautioned to use exchanges and to suitable candidates for listing on stock exchanges. How- has been “the last transaction,” which has caused all CFOs or seller. Accounting rules also had the effect of keeping a to use the same estimate no matter how unwilling the buyer homogeneous class of contracts “off balance sheet” until, in the span of just a few months, the contracts became illiquid, at which point other accounting rules brought those contracts onto balance sheets. Investors can move as a herd toward ever-higher asset prices even when a few investors attempt to turn against the herd. When a contrarian loses a bet, his loss both increases the wealth of the herd and justifies the herd’s direction. As than you can stay solvent. Lessons Learned Neither securities exchanges nor commodity futures ex- choose them wisely. At present there are only two kinds of exchanges in practice: securities exchanges and commodity futures contracts exchanges. Securities markets cope with price bubbles followed by bankruptcies. Commodities exchanges handle both asset positions and liability positions, but seize up when prices change quickly; many kinds changes can’t be placed on commodities exchanges. Keynes said to contrarians, the market can stay wrong longer rapid price changes, but are characterized by long-term changes were designed to deal with securitized derivative contracts. Securitization can be a good way to access the global capital markets, but only if the problems we’ve observed are successfully addressed. These problems are: 1. Securitization must no longer be an impediment to the normal process of renegotiating contracts. Packages of contracts can be listed on a contracts exchange, but underlying contracts. of contracts which can be expected to have rapid price The invisible hand of the market works in theory only when there is an active exchange between willing buyers and willing sellers. When there are willing buyers and willing to work well in practice, too. The problem is not with the sellers for goods and services, the invisible hand seems theory of the active market. The problem is that from time to time there are reasons that buyers buy against their will or sellers sell against their will, or that buyers are restrained from buying or sellers are restrained from selling. The more the investors believe any one theory or this should not preclude negotiations of changes in the 2. There must be an abundant flow of transactions between willing buyers and willing sellers. Securitizations must be standardized and traded on exchanges. Every position should be carried on a balance sheet. Transparency is important, but “fair market value” and willing sellers. does not help if there is no market that has willing buyers explanation, the more they tend to move as a herd. In the recent bubble, accounting rules and the pressures on CFOs added particularly to the herd-like behavior. This always happens a bit. But in this case the change to “fair market value” accounting caught many CFOs without the training, gent estimates of fair market value. The default valuation experience or data processing capabilities to make intelli- 3. Prices on securities exchanges can change quickly without seizing up but can’t go close to zero without inviting bankruptcy. Prices on commodities futures contracts exchanges can be “long” or “short” but can’t 70
Slide 71: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Creating an exchange for Insurance Contracts by Oakley E. (Lee) Van Slyke change quickly. Both of these problems must be addressed at the same time. Policy Implications The system of using securities exchanges and commodity Neither can transfer the risks of derivative financial confutures contract exchanges is inadequate to the task. tracts to the capital markets. Regardless of the degree of regulation or the financial incentives, this system is insufficient. A new type of exchange is needed that copes with fast price changes—even from “long” to “short”—without inviting bankruptcy. A new type of exchange is needed. This new type margin accounts, position limits or daily price change limits. Those traders must be listed on a securities exchange so that the world’s capital markets are able to invest in them. manipulation, but not to prevent any trader’s insolvency, as ease of entrance and exit must be built into the system. Regulations would be needed to prevent fraud and price uct warranties as well as derivative contracts of all kinds Providers of insurance, financial guarantees and prod- should be either highly regulated or listed on contracts exchanges that ensure transparency and liquidity, permit ease of entry and exit and collectively have the backing of the world’s capital markets. Retail insurance companies are could be listed on liquid contracts exchanges or be highly regulated. highly regulated. Reinsurance companies and syndicates of exchange would enable its traders to trade shares of standardized packages of financial contracts without Oakley E. (Lee) Van Slyke, FCAS, ASA, MAAA, is president of LIC Development LLC in San Clemente, Calif. He can be reached at leevanslyke@licdevelopmentllc.com 71
Slide 72: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The game of ‘Pass the Risk’: Then and now by Joy A. Schwartzman Financial crises are nothing new. What is unique and sobering is the far greater speed with which the current situation has evolved from a weakening of the U.S. housing market into a full-blown, global economic meltdown. Also new this time is how quickly the fallout spread beyond the financial sector into all areas of the economy. Advanced communications and information technologies are creating an era of greater risk with more serious and far-reaching consequences when compared with even the recent past. The increased speed, complexity and interconnectedness of global markets today make them more vulnerable to correlated risks that can combine to magnify liability. These new time and impact factors must be included in any future risk analyses. ages most people to work harder—to be more creative, The desire to succeed and do well financially encourward innovation and enlightened self-interest, and yet restrain it from boiling over into rampant greed? the current financial crisis for two reasons: These questions become more urgent in the wake of 1. The greater speed at which financial data travel around the world today. New communications technologies have turned the investment world into a that feeds on and multiplies its own exuberance or large and loud economic echo chamber—one panic. The instantaneous transmission of both good and bad economic news creates a climate of hair-trigger reactions as traders in different markets buy or sell on the rumor of the day. These knee-jerk transactions are then turned immediout from the initial event. In this way economic ripples quickly become market-swamping tidal waves. 2. All markets, industries and economies are now truly ately into new data that exponentially amplify the fall- innovative and productive. But at what point does enlightened self-interest mutate into greed and reckless disregard? When does innovation cease to be about creating a competitive edge and start to become a tool for gaming the system? Consider some past economic crises: The London connected. The consequences of economic events are The fallout from the LMX spiral was largely restricted no longer confined to one region, industry or nation. to the U.K. reinsurance market. The savings and loan late 1980s; the 1990s U.S. savings and loan scandal; or the Market Excess (LMX ) spiral1 that began to unwind in the more recent bursting of the Internet and housing bubbles. These financial debacles all began life as positive examples of innovation. Each was touted originally as a new and profitable way of conducting business. Each promised to bring benefits to all, but culminated instead in bringing economic ruin to many. Are unwinding spirals and busting financial bubbles scandal stayed in the United States and was contained bubble was global in scope, but its effects were not felt much beyond those who either worked for or held stock within the banking industry. The bursting of the Internet in Internet start-ups. The current crisis, however, which began simply as a softening of the U.S. housing market, global financial meltdown that has brought down major country around the world. grew within an incredibly short span of time into a banks and affected almost every industry in every In pursuit of understanding how this happened and pursuit of innovation and profit? Is it possible to create risk management or other safeguards that can successfully re- the price that must be paid for harnessing self-interest in 1 The “LMX spiral” represents excess of loss reinsurance placed in the Lloyd’s and London Market in the 1980s where reinsurers participated in different layers of the same exposures, often unknowingly. As claims were reported and reinsurance recoveries were triggered, losses worked their way through the “spiral,” often passing back and forth through the same group of companies. 72
Slide 73: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The game of ‘Pass the Risk’: Then and now by Joy A. Schwartzman how the environment for risk has changed in only a few decades, it is helpful to look at some of the similarities— conditions that preceded and led to the aforementioned LMX spiral, and the collapse of the subprime residential-mortgagefinancial crisis. reinsurers. Adding to their popularity and quick growth, as 10 percent, thereby making them an attractive sell for and, more importantly, the differences—between the LMX deals also offered commissions to brokers as high brokers (very much like subprime mortgages). With risk free money, just as low interest rates and ever-higher home prices appeared to be a license to make money during the recent housing boom. backed-securities (RMBS) market that sparked the current Collateralized debt obligations (CDOs) and other debt ostensibly low and commissions high, it seemed a bit like securitization products were touted as way of diluting the poison of toxic risk by mixing it in with good, investment- grade debt. But instead of the good credit risk making during the ’80s with a period of expansion within Lloyd’s, The growing popularity of LMX deals coincided the bad credit risk harmless, the bad risk polluted and (XL) contracts of the 1980s placed in the Lloyd’s/London as it opened its doors for the first time in centuries to thousands more “Names” (individual investors), many of whom were unsophisticated when it came to the insurance industry. syndicates with a heavy concentration of LMX policies.2 ultimately froze the entire credit market. The excess of loss Market were seen initially as a way of spreading risk across a series of reinsurers with the financial capacity to carry it. Both strategies became instead vehicles for foisting severely underpriced, highly correlated risks onto others— transmitting and multiplying toxic risk, in the case of the subprime mortgage debacle, so that it proceeded to spread These newer Names were disproportionately shunted into Problems developed when a series of huge losses in- cited an escalating spiral of claims that pinged back and forth among the finite number of companies and syndicates that had spent the past few years writing excess of loss protection for each other. There were huge losses related to the Piper Alpha oil platform explosion in the North Sea, the Exxon Valdez oil spill in Alaska, Hurricane Hugo, the like an infecting virus throughout the entire financial system—or, in the case of the LMX spiral, spreading risk when losses materialized. at price levels that ultimately became toxic to some reinsurers The LMX spiral developed in the 1980s, during a time San Francisco earthquake and a devastating windstorm in Europe. Any one of these events may have been enough of relative softness and quiet in the reinsurance market— the 1960s, and it appeared that markets and home prices to start the spiral unraveling; taken together they prehistory. Lloyd’s lost nearly 8 billion pounds between 1988 and 1992, many of the losses due to negligent underwriting, according to British courts. there had been low frequency of major catastrophes since would only ever go in one direction—up. (Sound familiar?) To compensate for falling rates in a soft market, Lloyd’s greatly expanded its use of London market excess of loss cipitated the greatest financial crisis in Lloyd’s 300-year (LMX) policies. LMX policies reinsure the policies of another reinsurance company or syndicate in exchange for a share of the premium. LMX business was attractive RMBS debacles are many. In each case there was: • • participants The similarities between the LMX spiral and subprime An attempt to mitigate risk by spreading it to market A series of new and complicated financial instruments because it was easy to administer, had low overhead, and the aggregating exposures could be off-loaded to other 2 not understood by most people and not even well Along with asbestos exposures. 73
Slide 74: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The game of ‘Pass the Risk’: Then and now by Joy A. Schwartzman understood by market professionals • • A pool of unsophisticated investors not adequately advised of the risk they were taking on A collection of unscrupulous brokers (reinsurance/ mortgage) who took advantage of the situation to increase commissions by encouraging as many deals as in the future • • The LMX spiral continues to play out, but has not spread to other areas of the financial system. There in a manner that left the institution of Lloyd’s and the was time to digest what was happening and to respond London insurance market intact and at least functioning in its weakened state, despite the seriousness of the crisis and the depth of the damage caused by it. • Lloyd’s ultimately made good on its obligations and— possible with no concern as to how they might play out Huge profits that continued as long as nothing happened to change the situation on the ground. In the case of the LMX spiral, there were no issues as with the creation of Equitas, a facility created to offload the unprofitable business years—returned to Bear Stearns or the many other banks, mortgage profitability. This cannot be said for Lehman Brothers, companies and unrelated businesses that have either long as there were no catastrophic events to set off a series of back-and-forth claims among the finite number of reinsurers. In the over-heated RMBS market, everything was fine as long as housing prices continued to only go up. The differences between the two crises sparked by the failed, been sold or are teetering on the edge of insolvency in the wake of the subprime securitization meltdown. LMX spiral and the subprime meltdown are fewer and have more to do with how the two situations played out once similarities for purposes of “lessons learned.” the trouble began. They are also more important than the institutions in a state of crisis. And markets are so globally Time is no longer on the side of today’s financial interrelated today that the fallout from major financial problems can no longer be contained to one country or region of the world. Joy A. Schwartzman, FCAS, MAAA, is principal and consulting actuary at Milliman Inc., in New York, N.Y. She can be reached at joy.schwartzman@miliman.com. 74
Slide 75: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications From Liquidity Crisis to Correlation Crisis, and the need for ‘Quanls’ in eRM by Stéphane Loisel The liquidity crisis and the insufficient depth of the market led to a strong correlation crisis: many risks that could be considered as close to mutually independent in the classical regime suddenly became strongly positively dependent. More correlation crises may happen in the future. We need to be more careful with black-box tools and to train what I would define as ‘’quanlitative analysts’’ (‘Quanls’) in the Enterprise Risk Management (ERM) process, that is risk managers who are able to lead interdisciplinary ERM studies from a jointly qualitative and quantitative point of view, with an emphasis on dynamics. lack of the depth of the financial market to absorb liquidity by (too) easy access to credit and (too) low interest rates. As things went wrong, many risks, often considered to be The recent crisis may be regarded as a result of the within the market and as such is an example of endogenous risk, in analogy with the horizontal oscillations of the Millennium Bridge of London that forced the bridge to close for a battery of tests three days after its opening to the public, as noted by Danielsson and Shin (2002). insurance and financial risks. The consequences on the future earnings of insurers and the difficulties that financial institutions would have to maintain their activities are correlations are defined in QIS4 of Solvency II does not A pandemic could create a correlation crisis between often underestimated in Solvency II and in Basel II. The way really take into account correlation crises that could occur after a catastrophe, or just because of endogenous risk, for example with surrender options. How to value these options, as well as deposits in finance, remains a question that has to be addressed in a more sophisticated way. needs after a period of artificial additional growth generated close to independent, suddenly became strongly positively Fisher et al. (2008) and Biard et al. (2008). dependent: this is what we define as a correlation crisis in From the point of view of the insurance industry or of the mathematics. I am convinced that some products were far too simple. One often hears about the crisis that people got lost in far too complex, and the models to describe their dynamics To me, considering more sophisticated models does the equity derivatives market, the recent crisis would be a consequence of an external shock arising from the subprime crisis in credit risk. The fact that many companies defaulted or were downgraded almost simultaneously corresponds to what is often referred as a consequence of the smile of correlation: correlation has been known to be number of years. not mean replacing a Brownian motion with a more general Lévy process, or a Gaussian copula with a mixattention to the dynamics, and consider risk processes with with the goal to understand the main sources of risk. ture of Student copulas. I believe that we must pay more non-stationary increments and dynamic correlation models, If a pandemic occurred, the delay between its beginning larger in bad times than in the classical regime for quite a Nevertheless, after this exogenous risk appeared, once liquidity needs are there and as the market is not behave similarly, breeding a vicious cycle: because of mardeep enough to absorb it, most market participants tend to and the date at which insurers or reinsurers would have to pay, as well as the time elapsed before stock prices move back up after the epidemics, would be very important. gin calls and liquidity needs, investors are forced to sell valuable (on the long term) assets at the bad instant, which leads to adverse price moves, further margin calls, and so on… This copycat behavior generates and amplifies risk Similarly, in the equity market, the correlation crisis that end later on almost as suddenly as it appeared, and missing this dynamic would lead to bad hedging strategies. Identifying the main sources of risk and understanding their caused many basket options to be underpriced is likely to 75
Slide 76: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications From Liquidity Crisis to Correlation Crisis, and the need for ‘Quanls’ in eRM by Stéphane Loisel interactions is far more difficult, but it has to be done if moving back to Basel and Solvency 0. Both fundamental an integrated risk view and a correct ERM process, those be studied and managed at the same time. we want to move to Basel III and Solvency II.1, instead of and applied research is needed to tackle these issues. With external shocks and their endogenous consequences could Important difficulties to overcome concern IT, Pillar for an ERM process, for supervision and for communication to markets (Pillar III of Basel II and Solvency II). Fair value and risk neutral valuation techniques have also been too often blindly used without exercising critical judgment. There is currently a debate on the use of fair value and the freedom to use a different framework during a crisis. First, I think one must not mix up accounting, be useful to measure something and to take risk into does not show up. Third, in contrast to some people who 2 of Basel II and Solvency II and invisible barriers that regulatory and pricing tools. Second, a concept that should account is not suitable if you can only use it when risk recommend forgetting these valuation techniques, I think make it quite difficult to implement an ERM process that guarantees this view of risk at the macro level. Very often, to meet the constraints of some softwares and to maintain robustness and auditability of processes, models are simplibe ignored for some combinations of positions taken by the front office. Besides, some market participants, instead of from self-developed programs that are black boxes at the fied, and key risks like credit risk or exchange rate risk may they ‘’just’’ need to be adapted to take into account risks of temporary illiquidity, correlation crises and copycat behavior and the way transactions are made, in particular if there are linked securities, for example). using the official software, may take their decisions directly risk management level. One must absolutely avoid blindly to measure credit risk. In the insurance business, I am only a few market participants (in the case of insuranceAnother point to carefully address is the way brokers, trusting black-box models as one trusted rating agencies concerned by the fact that a few software developers have a monopoly on the quantification of financial consequences of natural disasters and are almost blindly trusted by many yearly price movements obtained for the same risk with the insurers and reinsurers, in spite of recent events and strange same software. Because these risks are complex and specific, traders, executives and others can maximize their salaries, and the perverse incentives this may create. The one-year horizon in Solvency II reinforces the preference for a short- term view too. It is clear that if no five- or 10-year indicator is added into the current project, most companies will mainly develop short-term capital models and not enough will consider long-term perspectives. The same reasoning applies to governments: Was there it is tempting for supervisors to use this black box model the middle office with the front office software because it an incentive for them to limit or to encourage easy access to credit? The answer is not the same with short-term and long-term views, but elections are a key factor to help the guarantees they would have to give to keep the finan- as well. Similarly, in finance, some controls are made by would be too expensive to develop another one. If market hard correlation crisis might occur if an unmodelled catastrophe breaks out. To implement a valuable ERM process would require more transparency of models and strategies, and we get to one of the main issues to address after the crisis: how to deal with the mismatch between confidentiality and competition on one side, and the need them choose their strategy! This led them to underestimate cial system up. The guarantee that governments provide to some financial institutions should be studied in detail some circumstances if one does not pay attention. because of the competitive advantage it may generate under After the threat of economic crisis in 2009, we may participants or insurers all use similar black box models, a face in the 2010s and in the 2020s other correlation cri- 76
Slide 77: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications From Liquidity Crisis to Correlation Crisis, and the need for ‘Quanls’ in eRM by Stéphane Loisel ses that could arise from illiquidity, pandemic, inflation, oil peak, climate change, pollution, a natural disaster, etc…. To limit their financial consequences and ensure the longterm viability of our financial system, I believe we need to References Fisher, Wayne, Loisel, Stéphane and Wang, Shaun. “On Some Key Research Issues in Enterprise Risk Management Group Level.” Bulletin Français d’Actuariat (BFA), No. 15, Vol. 9, 2008. put more emphasis on fundamental and applied research and use continuous professional development to train ERM These experts would be able to identify, quantify and manthe underwriting process to investment strategies. experts with both quantitative and qualitative expertise. age risks faced by insurance and financial institutions from For some risks, studies with mutualized data by related to Economic Capital and Diversification Effect at Biard, Romain, Lefèvre, Claude and Loisel, Stéphane. “Impact of Correlation Crises in Risk Theory: Asymptotics of Finite-Time Ruin Probabilities for Heavy-Tailed Claim Amounts when Some Independence and Stationarity Assumptions are Relaxed.” To appear in Insurance: Mathematics & Economics, 2008. Danielsson, Jon and Shin, H. “Endogenous Risk.” In ModBooks, 2003. researchers, federations of insurance companies and banks blind trust in black-box models. This would help us to find management and regulation. and international institutions should be carried out to avoid out which risk indicators would be relevant for better risk ern Risk Management: A History. Edited by Field, P. Risk Stéphane Loisel is associate professor at Institut de Science Financière et d’Assurances (ISFA), Université Lyon 1, France and research associate at Ecole Polytechnique, Palaiseau, France. He can be reached at loisel@univ-lyon1.fr. 77
Slide 78: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications societal themes: behavioral/personal economics essay on the financial crisis By ANDREw wINklER the human touch underlying the current financial crisis By VIVEk gupTA if it looks too Good to be true ... by steve MAleRich 78
Slide 79: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications essay on the Financial Crisis by Andrew Winkler The current crisis is catalyzing an array of responses, including searching for causes, reworking regulations, scapederstanding of the cause, the remedies may do more harm goating and a massive capital injection. Without a clear unthan good, innocents may be scapegoated, and valuable progress in financial tools may be lost. Worse, it will happen again. From a simple mathematical model of the underlying make housing prices inversely proportional to the interest rate. If interest rates are cut in half, house prices double. When those rates double, house prices are slashed in half. When interest rates are large, they are not likely to double or halve, but when interest rates are small, a small adjustment can be a big percentage change, and the danger of big swings in housing prices is appreciable, even inevitable. With no down payment, no amortization and closing economics, I first predicted this crisis in July of 2004. An economic dynamic relating very low interest rates to the this outcome foreseeable, indeed inevitable. The current full explanations; see mattersofinterestmatters.blogspot.com. costs folded into the loan, the only issue in affording a structure of the demand curve in the housing market made crisis had a mathematical cause. There isn’t space here for This much is clear to everyone—the crisis results from house is the monthly payment, which is the house price half, the house you can buy with a given monthly payment the same income distribution are competing for a fixed stock of housing. The house price is bid up until the new multiplied by the interest rate. If interest rates are cut in costs twice as much. But the same number of people with an epidemic level of mortgage defaults, in turn caused by ballooning monthly payments from variable rate mortgages, caused by a rise in interest rates from historically low levels. This made the monthly payment change quite large, because while the rise was small in absolute terms, it was huge in relative terms. monthly payment at the new interest rate matches the old monthly payment at the old interest rate. The house price varies inversely with the interest rate. The effect is somewhat mitigated, ironically, by property taxes, which effectively raise the interest rate, but historically low rates, and burst when those rates rose again. it’s no accident that the bubble occurred during a time of This perfect storm required the confluence of a number default the only option. This is the effect which we must real estate value, and why do rising interest rates burst that bubble? The simultaneous plummet in property values made understand—why do low interest rates cause a bubble in of factors, each one of which was at worst innocuous and at best virtuous. The traditional mortgage had several features which had recently been relaxed; fixed rates were forced by unpredictable inflation rates to become variable; sophisticated credit models and rising markets made down innovations, in isolation, represents a significant advance in rates were low for valid economic reasons. Taken together, however, they arm a trap which springs when interest rates dip by a significant factor, and then rise again. a demand curve for housing. Over the short term, the housing supply can’t change, so it’s the demand curve that’s discuss, buyers buy a monthly payment, not a house price, and buyers buy as much house as they can afford. crucial. The two central facts are these: for reasons we’ll Like all equilibrium pricing, there is a supply curve and payments and amortization less meaningful. Each of these making home ownership affordable and available. Interest require no down payment, and only interest payments, alter that is harmless enough when interest rates are high, but This brings us to the heart of the matter: mortgages that the structure of the demand curve for real estate, in a way which drives a bubble at low interest rates. Specifically, they owners walk into that trap? Why would mortgage lenders? Dr. Alan Greenspan recently testified that he discovered But that raises significant issues. Why would home- 79
Slide 80: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications essay on the Financial Crisis by Andrew Winkler a flaw in the model of how the world works, that he had relied on the self-interest of lenders to act rationally. The market worked, however. Supply met demand. to answer three questions: Why do buyers buy a monthly payment? Why do buyers buy as much house as they can short answer is that supply and demand are not magic; they’re Selection at work, which tells us the limits of the model. afford? Why didn’t lenders see the trap, and avoid it? The To fully understand what did happen, and why, we need worked better in the past than the alternatives did. Sexual selection forces it, as does natural selection. You don’t want your kids exposed to drive-by shootings or gang know that your display of wealth will have a real impact on lauding that undeniable fact. violence. You do want your kids to be attractive, and you their attractiveness. Don’t shoot the messenger—I’m not Why do buyers buy a monthly payment? There are Why do buyers buy as much house as they can afford? tivity is physical activity. And physics, the body of knowl- Economic activity is human activity is biological ac- two, related reasons. It reduces what is at heart a very comto a single, knowable, comprehensible number. The other reason has to do with personality, itself a manifestation of Sexual Selection. Estimates vary, but around 40 percent of temperament, characterized by a preferred reliance on the ness of sensory input from the external world, creating a strong orientation to the “here and now,” and a relative blindness to the past or the future. For such a person, the monthly payment is the “here and now”. plicated transaction full of unknowable future uncertainties edge, is simply a collection of technologies for calculating probabilities, with the key insight being the Principle of More. In biological systems, the Principle of Selection Natural Selection, and the less familiar, but more important, Principle of Sexual Selection. The Principle of Natural Selection, if you recall, states occurs in at least two distinct forms, the Principle of the population of the United States has the “Improviser” “extroverted sensing” cognitive function, profound aware- that a heritable trait which confers a higher degree of probability of survival to an individual, has a higher probability Selection states that a heritable trait which confers a higher probability of surviving in a population. of surviving in a population; while the Principle of Sexual degree of probability of having offspring, has a higher Seen in this light, maximizing utility, which drives characterized by a preferred reliance on “introverted sensing,” a deep awareness of sensory memory, creating a strong orientation to the past, and a keen awareness of standards work is going to continue to work, and a high level of trust in “the system”. For such a person, the fact that a mortgage product contains innovative elements would be counterbalanced by the source of those products—(formerly) large rich prestigious established institutions. of the United States has the “Stabilizer” temperament, Another approximately 40 percent of the population both supply and demand, means neither more nor less than Any economic behavior that raises the probability of surDNA or cell structure or ideas or skills, will predominate. of money. maximizing the long run number of surviving offspring. vival, or of offspring, which is also heritable, whether as This makes supply meet demand, and forces the time value But nobody knows the future. Biological selection and tradition, as well as a strong bias that whatever used to can’t (or at any rate hasn’t) given us the power to formulate decisions based on perfect knowledge of the future. Rather, it gives us tendencies and faculties that have, on average, missing patterns, and in being relatively unaware of the future. noticing details, as well as a vulnerability, in sometimes The two other temperaments, Conceptualizers and Both Stabilizers and Improvisers share a strength, in 80
Slide 81: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications essay on the Financial Crisis by Andrew Winkler Catalysts, share a strength in noticing patterns, as well as a vulnerability, in sometimes missing details, and in being overly future-focused. As such it is largely a test for identifying smart Stabilizers. Quant interviews lean heavily on “fact sheet” questions, or tricky problem solving. Einstein need not apply here! Asked what the speed of sound was, he wondered why he Wall Street has a strong bias for detail-focused rather than pattern-focused people. Businesses do in general. The mathematical component of the GMAT tests heavily Which brings us to why lenders failed to see the trap. would bother to memorize something he could look up in advance, businesses which rely solely on the Improviser’s here-and-now, real-time response have become just as resistance to innovation. an encyclopedia. In the modern world of rapid technical your knowledge of Euclidean geometry, which has been essentially useless since the days of Descartes, but draws heavily on your “extraverted thinking” faculties, largely ignoring your analytic “introverted thinking” capabilities. vulnerable as those which rely solely on the Stabilizer’s Andrew Winkler, Ph.D., is CEO at Data Risk Management, Inc. He can be reached at a@datariskmgmt.com. 81
Slide 82: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Human Touch Underlying the Current Financial Crisis by Vivek Gupta This article presents a hypothesis that the market cycles are primarily created by human actions, behavior and assumptions rather than by random variables. An analysis of main financial events over the last decade reveals that: • • • Every one of these events has had a human touch. The current financial crisis is an aftermath of the excessive economic boom during this period. and innovation. A new economy must emerge from human ingenuity The housing boom and the home equity impetus: Due to low interest rates, paying an additional $10,000 for a house only meant an increase of $40-$50 in monthly a higher mortgage amount, the homebuyers started bidding up the house prices. The bankers bundled up these all over the world to raise more cash. This iterative cycle through a never-ending supply of money resulted in evermortgages and sold mortgage-backed securities (MBS) mortgage payments. When the bankers were eager to lend increasing house prices. Many homeowners borrowed and travel and invested in the stock markets. This selfand uplifted the stock markets. The B!O!O!M! The Internet and Y2K: When former U.S. President Bill Clinton declassified the military technology of the Internet for public use, it created an unprecedented tech boom, against the equity of their homes and spent on renovations fulfilling prophecy created an enormous construction boom Extrapolation of historical data to make future as- millions of new jobs and a booming global economy. Microsoft President Bill Gates added fuel to this fire by hyping the concerns of Y2K. Y2K fears forced almost every business in the world, especially in the developed world, sumptions: All financial players—lenders, borrowers and never fall in the future because those prices had never fallen policymakers—believed that the U.S. house prices would in the past. The lenders kept lending indiscriminately, ascover their investment by selling the houses that were going role in creating the recent housing boom. either to upgrade or to replace their computer systems. No catalysts ever experienced. doubt, Y2K was one of the single most powerful economic 9/11: The horrific tragedy of 9/11 caused a major melt- suming even if some homeowners default, they could reto appreciate anyway. This assumption played a significant Mathematical mirage of index returns through the down of the stock markets as well as of investor confidence. At that time, the U.S. government felt compelled to lift the Federal Reserve Board Alan Greenspan started increasing ber 2003, federal funds rates touched as low as 0.96 per- American economy. To do so, former Chairman of the U.S. the supply of money to lower the interest rates. In Decemcent, the lowest level ever recorded in American history. by lifting the stock markets and the mood of America in due course. smoke and mirrors of mergers and acquisitions (M&A): Lately, a combination of low interest rates and low credit percent. To earn a 1 percent spread, they valued the targeted stock how M&A made the index returns look highly impressive. spreads allowed high-end borrowers to borrow at, say, 4 expecting a 5 percent yield. The following example illustrates When this stock started trading for $120, everyone who This unprecedented low rate did achieve the desired intent owned the stock in 2006 got an impression of 20 percent Time 2006 2007 Expected dividend $6/per year $6/per year Expected yield 6% 5% valuation Mathematics 6/.06 6/.05 stock price $100 $120 82
Slide 83: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Human Touch Underlying the Current Financial Crisis by Vivek Gupta appreciation in the “value” of their stock, whereas their income was still $6/per year. A simple analysis of historical stock prices cannot reveal how lowering the expectations of future income created a mirage of exuberant current ill-intentioned on anyone’s part. rates could not be pushed any lower, the house price boom The artificial economic impetus disappeared. stabilized and halted the free income from home equity. Super-saturated housing market: Due to many entic- income. In this process there is nothing illegal or even Fire and hay: Intermediaries and investors: Generally, ing incentives, people who would have bought a house over the next three to four years had already bought a house. When the high-risk mortgages started defaulting in large investors are averse to the guaranteed deposit returns and are so mesmerized by the historical surge in stock market indexes that they ignore the market risks and sales commissions. Intermediaries were easily steering such investors This raging bonfire was adding glow to the markets. numbers, lenders suddenly realized that they could not recover their investment through foreclosure in a super-satwas punctured! urated housing market. Lenders’ fundamental assumption The domino effect: When lenders’ overoptimistic as- toward equities as preferred investment for their retirement. Oversupply of money, yet no inflation: Theoretically, sumptions did not pan out as expected, lending retreated, demand retreated, house prices plummeted, MBS value fell, guarantors of MBS defaulted, credit swaps fell in the money, and the issuers of credit swaps disappeared or weakened. Rescue packages: The U.S. government could take the appropriate amount of money in the economy is that which keeps the inflation between 1 and 2 percent. The central banks have flooded the markets with money and have created enormous demand without disrupting the inflation charts. To achieve this most desirable outcome, China, India and other low-cost production regions. This tion, economic boom and vibrant stock markets. The BUsT Imbalance of production and consumption: The fundamental reason for this recent bust is a trillion dollar (fiscal + trade) deficit. In the most fundamental trading relationship for humans, the barter, the United States has a gap of one trilrest of world: “Give us goods and we will pay you later.” lion dollars. Essentially, the United States is saying to the So far the rest of the world is accumulating “savings” by exporting goods to the United States, confident one day it will get its savings back with interest. The housing bust and halt of the home equity impe- they increased supply by liberalizing import quotas from phenomenon resulted in an unprecedented era of globaliza- over the severely defaulted mortgages in an attempt to unlock the credit crisis. However, would the bankers find enough creditworthy homebuyers with 25 percent down payments? Sudden deleveraging: Eighty percent of U.S. GDP is dependent upon the consumer expenditure, which is in parts supported by borrowing. A $50 reduction in monthly pay means a reduction of $10,000 borrowing capacity. A slight slowdown in the job market will significantly reduce be seen how nimbly the U.S. economy can shrink. The … The ripple effects of pension shortfalls: Due to declining stock markets and low interest rates, pension asset values have fallen and liabilities have increased. Some of this gap a soft economy, profits have shrunk. Higher-than-expected has to be covered from company profits this year. Due to pension contributions will further lower profits. This may demand for everything, from cars to trinkets. It remains to tus: Prudent Alan Greenspan admitted that the housing boom was reaching the tipping point by saying “There is some froth in the housing market.” At a point when interest 83
Slide 84: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Human Touch Underlying the Current Financial Crisis by Vivek Gupta result in lower credit ratings and further drops in stock prices. Since company A’s stock is part of company B’s pension portfolio and vice versa, more pension contributions will have a domino effect on stock prices. The defined benefit pension plans established many years ago will be hit hard. The demographic shift: Developed countries are ex- resulting in loss of wealth. As far as I can see, innovation overcome concerns regarding environment, economy and is the only comprehensive solution that can simultaneously excessive capital. We need a paradigm shift in our thought dard of living for all humanity by adding enormous value— for example, harnessing gravity. Summary: No one is proposing ways to rectify the process to achieve innovations that can further lift the stan- periencing an aging population along with a slower growth in their intrinsic populations. A large segment of the working population will retire in the next few years. The ecosumption by the stable population only. trillion dollar deficit that is creating a serious geopoliti- nomic growth is assumed to come from the increasing conThe environmental scare: To some degree, most peo- cal shift. Optimism is good; however, too much optimism fosters complacency. It is time we realize the severity of the situation, implement a serious change in behavior and innovation and pension reforms, to overcome the curact urgently to find long-term solutions, mainly through ple believe something needs to be done to avoid looming lead to job reductions or are considered ineffective. Most policymakers tend to favor jobs over the environment. Innovation, the last hope: All the reasons that environmental catastrophe. The proposed solutions either rent challenges. Such solutions will not be fast and easy but rather will require virtues like leadership, hard work, courage, innovation, sacrifice, etc. If we fail to do so, either environmental or economic woes will lead to widespread some time to change course. tal cannot be deployed in new industries. Therefore, it is chasing a few opportunities, creating extreme volatility and caused the bust have one common thread—the available capi- catastrophe. After all, there is good news. We still have Vivek Gupta, FSA, FCIA, is an actuary at VG Actuarial Consulting Inc., in Kingston, Ontario. He can be reached at vg@vgac.ca. 84
Slide 85: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications If It Looks Too good to Be True, … by Steve Malerich Most people should know how to end that statement. Leading up to the current financial crisis, many looked at States and concluded that we were in an asset-price bubble. real estate prices in some significant areas of the United Yet, despite that appearance, we continued to lend money bubble has sent shock waves around the world. America has taken a step toward long-term fiscal responsibility. We have reduced spending. The next step is not to reverse completely that movement, though some reversal may be good. Rather, we need to start moving long-term needs. some of the increased savings into things that will promote Economic growth can come from investments targeted based on prices then prevalent. Now, the bursting of the Although real estate can be called a trigger of the current crisis, it is only part of the problem. For years, for long-term benefit. An obvious choice is in investments number of retirees. There are many other candidates for warnings about things that may be too good to be true. we have lamented that people weren’t saving enough. increased savings and investment. Our attention turns to that will help to meet the needs of a substantially growing such investment. The physical sciences give us many No doubt, some such investments will fail. Others may Yet, when spending slows, we forget about the need for the slowing economy and we demand that something be done. Invariably, that has led to some form of stimulus to encourage increased consumption (and discourage saving). prove to be unnecessary. But then, investments of all types fail and solutions often abound for problems that don’t to our failure to find an economic goal other than continuous “growth” as it’s been measured for decades? As painful as it may be, we can view the current crisis Could it be that the depth of our current crisis is related materialize. Imagine the casualties among traditional investments, however, if any portions of today’s warnings prove accurate. as a wake-up call. Rather than trying desperately to restore a “healthy” economy, perhaps it’s time to try building a truly healthy economy—one where we don’t keep promoting and investing for the future. consumption of any and all kinds at the expense of saving worked in the past just because they have worked in the some of these things are—too good to be true. We should not continue to promote things that have past. There is too much evidence to suggest that at least Steve Malerich, FSA, is assistant vice president and actuary at AEGON USA in Cedar Rapids, Iowa. He can be reached at smalerich@aegonusa.com. 85
Slide 86: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications effective Risk Modeling: credit crisis lessons for Modelers by pARR schoolMAn what Is a Robust level of Risk Capital? By lARRy RuBIN AND XIAokAI (VICToR) SHI 86
Slide 87: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Credit Crisis Lessons for Modelers by Parr Schoolman Does the credit crisis mean the heralded Age of the Quant because it increased the interest rate risk for these prepayment rate, as their refinancing tended to be less has passed? Much of the blame for the current credit crisis is being laid at the feet of the analysts responsible for modeling and evaluating the innovative debt securities driving the massive losses for financial institutions. How was the modeling of these securities so wrong? securities. Subprime loans demonstrated a more stable correlated with market interest rates, and more cor-- related with individual borrower financial difficulty. pool of independent borrowers. Furthermore, as much This source of prepayment was diversifiable for a large as 80 percent of subprime mortgages contained prethe mortgages to be refinanced if interest rates interest rate risk of subprime securities, making them security with the same credit rating. economists, “Making Sense of the Subprime Crisis,”1 pro- An article recently published by four Federal Reserve vides some insight into what information was available origination associated with the most toxic segment of the subprime securities. The falsely optimistic pitch to investors could have been based upon the following points: 1. The subprime market fundamentals were considered payment penalties,2 further reducing the likelihood of decreased. These features reduced the perceived arguably a safer investment than a prime mortgage 3. The strong housing market was expected to mini- for analysts during 2005 and 2006, the time period of loan to be strong. Lending in this market had evolved toward subsidiaries of large, reputable financial services companies, replacing the small, thinly capitalized lenders of the 1990s. Lenders were increasing the use of quantitative models based on credit scores for loan underwriting, which were demonstrating an borrowers. Furthermore, the historical performance of mize the downside risk of subprime loans. The data typically used to evaluate these securities went back to 1998. Data prior to 1998 was not thought to be as relevant due to the changes in the industry regarding loan originators and the more automated underwriting process. Unfortunately, that time period did not contain a recession, nor did it contain a period of sustained home quoted as stating: improvement in average FICO scores for subprime subprime mortgage securities had shown them to have bonds. With increased use of automated underwriting, price declines. A Citigroup December 2005 report is “the risk of national decline in home prices appears remote. The annual HPA has never been negative in the United States going back to 1992.” more stable credit ratings than similarly rated corporate improved credit score transparency and more reputable lenders, the performance of subprime securities was expected to remain strong. 2. Subprime securities were expected to have less Home price appreciation (aka HPA) all the way back to 1992 has not been negative. What could possibly go wrong? The basics of this story look very familiar to what oc- interest rate risk than prime mortgage securities. Prime mortgage borrowers had demonstrated a tendency to refinance their loan and pay off their existing loan when interest rates decreased. This correlation to interest rate changes was problematic for investors curred in the P&C insurance industry during the depths of the soft market of the late 1990s. Underwriters and brokers 1 2 K. Gerardi, A. Lehnert, S. Sherlund and P. Willen, “Making Sense of the Subprime Crisis,” Sept. 5, 2008. G. Gorton, “The Panic of 2007,” Aug. 4, 2008. 87
Slide 88: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications Credit Crisis Lessons for Modelers by Parr Schoolman were making assertions that the re-underwriting of books would mean that future results would be better than historical also expected to reduce the future development that stan- loss experience indicated. Changes in claim handling were dard actuarial loss triangle methods were predicting. Management teams were proclaiming that the diversifying of their portfolios into new lines of business would reduce the risk of loss as well. Wall Street errors of the current crisis and acknowledge the resulting uncertainty. A model built on five to 10 years of data provides limited information about a 100-year PML. Many analysts of subprime securities recognized that using data since 1998 was less than Extrapolating beyond the historical data, they made reason- Finally, understand the limits of the data being used ideal and not fully representative of all possible scenarios. able estimates of the potential losses to securities backed ever, their biggest mistake was to underestimate the probability of U.S. housing prices dropping nearly 20 percent from 2006 to 2008 in the largest metro areas. This error is more difficult than the quantification of possibilities. echo these soft market mistakes of the P&C industry. Both Wall Street and the insurance industry have demonstrated a propensity for underestimating risk, although the bankers or two at the end of their paychecks while doing so. Going living learn from these missteps? seemed to have discovered a way to receive an extra zero forward, what can those who attempt to quantify risk for a First, recognize that the accuracy of a model is limited by subprime loans if home prices were to decrease. How- demonstrates that the quantification of remote probabilities To further illustrate this point, Nassim Taleb presents to the accuracy of the input assumptions. Complex models that the entire range of indications may hinge on one or two key assumptions. Use data-driven assumptions, making the clever story of a turkey being raised on a farm in his can provide a false sense of security, hiding the evidence book The Black Swan. Every day of its life, when a turkey sees the farmer, it gets fed. Based upon that experience, when the turkey sees the farmer coming out of the farmhouse the day before Thanksgiving, it sees no reason to be risk posed by the farmer is driven by the fact that the turTo make sure the end users of model projections do concerned. This very big error in judgment regarding the key’s prior experience period did not include a Thanksgiving. sure the time series includes stressed environments when possible. If a model of underwriting risk indicates that the from 1998 to 2000 is remote, it is not a realistic model. probability of accident year combined ratios experienced Second, stress test key assumptions. In most insur- ance risk modeling exercises, the correlation assumptions drives the tail of the results. These correlation assumptions not make the same errors in judgment as the turkey, modelers should maintain the humility to document the limits of the data underlying their model, providing transparent summaries of the key assumptions and their impact to the abilities for real world results. uncertainty of the estimates. Don’t mistake modeled probWhat Thanksgiving is your model potentially missing? between lines of business and between other risk elements should be transparent, while the model needs to be able to stress test the impact of increased correlation between risk elements. Each new market crisis demonstrates that correlation in stressed environments is much higher than historical averages would indicate. What are you doing to address it? Parr Schoolman, FCAS, is vice president at Aon Benefield Analystics. He can be reached at parr.schoolman@aon.com. 88
Slide 89: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications What Is a Robust Level of Risk Capital?1 by Larry Rubin and Ziaokai (Victor) Shi Market observers have never been as skeptical on financial service firms’ capital standing as they are today. As The Hartford released its third quarter earnings on Oct. 30, call was dominated by questions on the company’s capitalization level. Their skepticism makes sense. As shown in the table 2008, its stock price fell almost 50 percent when it reported a $2.6 billion quarterly loss. Many believe this was driven by its failure to convince the market that it was sufficiently capitalized to survive the financial crisis. The conference below, some prominent firms have asset-to-equity ratios as high as or higher than 20, which means their $1 in capital could leverage more than $20 in assets. Wall Street firms and some other financial conglomerates were operating LEvERagE oF Top 25 FInanCIaL sERvICE CoMpanIEs In ThE unITEd sTaTEs ($ Billions, 03/31/2008) asset Rank 10 16 5 11 7 4 18 8 15 17 1 14 21 24 19 6 3 13 12 22 2 9 25 23 20 source: snl Company freddie Mac bear stearns Morgan stanley lehman brothers Merrill lynch Goldman sachs fhlb of san francisco fannie Mae prudential financial the hartford citigroup Metlife GMAc countrywide financial WaMu AiG JpMorgan Chase Wells fargo Ge capital u.s. bancorp bank of America Wachovia farm credit system bank of new york Mellon berkshire hathaway Industry specialty lender broker/dealer broker/dealer broker/dealer broker/dealer broker/dealer fhlb specialty lender insurance insurance bank insurance specialty lender thrift thrift insurance bank bank specialty lender bank bank bank specialty lender bank insurance asset-to-Equity 50 34 33 32 29 28 23 22 21 19 17 17 16 15 14 13 13 12 12 11 11 10 7 7 2 Q1 2008 assets 803 399 1,091 786 1,042 1,189 332 843 478 344 2,200 557 243 199 320 1,051 1,643 595 684 242 1,737 809 197 205 281 Q1 2008 Equity 16 12 33 25 37 43 14 39 23 18 128 33 15 13 22 80 126 48 58 22 156 78 27 28 119 1 The views in this article only represent the authors’ personal opinions. This article does not represent any statements from the organization where the authors are currently employed. 89
Slide 90: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications What Is a Robust Level of Risk Capital? by Larry Rubin and Ziaokai (Victor) Shi with high leverage ratios while at the same time they assurement generated from their internal risk capital models. Would those internal model generated figures be sufficiently robust for financial firms to withstand unexpected losses such as the credit crisis happening today? Also what has talization of financial firms? sumed that they were holding enough capital based on mea- Brothers, Washington Mutual, Fannie Mae and Freddie Mac had to face the reality of either filing for bankruptcy or This is largely because of the difficulties in funding their selling themselves (to the government or other companies). capital gaps. Regulated insurance companies, although also better capital strength to begin with and so far have been able to survive based on private sector solutions rather than purely relying on government bailouts. This is because 1) the are healthy and therefore attractive to private investors. really caused the financial crisis as well as the undercapi- lacking capital access under current market pressure, had What Has Caused the Problem? In the middle of this “hurricane” of the financial crisis, insurance companies generally stand in slightly better positions than investment banks (with a few exceptions that capital gap is smaller; and 2) their core insurance operations Regulation, especially of minimum capital require- mostly arise from businesses written by affiliates of U.S. insurers but were not regulated as insurance). The credit surance sector. One of the reasons might be because they crisis has impacted the banking sector more than the inare under different regulatory environments. The insur- ments, plays a key role in preventing the trains from moving off their tracks. However, is more regulation the answer, or is the current problem a result of flawed regulation? In the past, we occasionally heard complaints of the over-regulation of federal banking regulators on investment and non-investment banks. However could it be true that federal regulation enforced burdensome regulatory rules resulting in only increased workloads on principles of capital requirements? financial reporting processes, while neglecting some basic It is the authors’ contention that it was not greed ance industry differs from banking in terms of the regulatory requirements of capital required. In the United States, the regulators have enforced risk-based capital (RBC) law, which requires insurers to hold minimum capital requireprovided by regulators. Companies need to have an RBC ments according to calculations using a series of factors ratio (total capital after slight adjustments over minimum requirement) of greater than 150 percent to avoid any four times the minimum RBC requirements for a targeted by setting up one bottom line of capitalization level across the industry. While U.S. risk-based capital is a crude and regulatory actions. Insurers are generally holding two to rating. This regulation has helped the insurance industry that led to the financial crisis but inadequate capital that followed from a flawed risk management strategy. Without an industry-wide capital framework, financial firms have been overrelying on their internal economic capital (EC) models to make their capital funding/allocating decisions. But they made three faulty assumptions: 1) they are EC model has calculated; 2) EC models, which rely on to survive unexpected losses; 3) EC is the capital needed for adequately capitalized if they hold capital at the level their historical experience as input, are sufficient to enable them the company to survive until the company can recapitalize (one year). We believe if companies had adequately considcapital, the credit crisis might have been avoided. ered the market price of risk in determining their economic one-size-fits-all solution, the fundamental premise is that its future obligations. capital should be sufficient to enable a company to mature However, investment banks are not as regulated as insurance companies in the United States. The credit default swap (CDS) market was nearly unregulated before this crisis. Firms like Bear Sterns, Merrill Lynch, Lehman 90
Slide 91: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications What Is a Robust Level of Risk Capital? by Larry Rubin and Ziaokai (Victor) Shi Market Price of Risks Markets price risk even though risks may not be traded in a deep and liquid market. Investors require margins when they choose to lend or invest their money. This margin and optimism, as well as changes in their risk aversion. A-rated bond issuers). Using a cost of capital of 9 percent, results of the analysis are shown below. we solved for the market implied economic capital. The decreases or increases depending on investors’ pessimism Over the longer term, this margin becomes the excess of the company cost of capital over risk-free rates. A company creates additional value for its investors if the return on economic capital exceeds the cost of capital. Basic Points of Notional Amount Market price economic capital 99.5% percentile over one year 99.5% percentile over five years 265 65 122 under Solvency II is significantly lower than the market implied economic level of capital. Even under five years’ loss (Solvency II defines one year) time horizon, the internal The conclusion is clear. The economic capital defined historical loss data) to establish economic capital? To what extent does the market-priced risk impact economic capital? The “standard” or most popular definition of economic Can we purely utilize an internal model (that relies on EC is still lower than the market priced number although it is closer. We believe there are a number of reasons for this difference: capital is defined as the amount that an insurance company needs so that it can absorb all losses within a one-year time horizon with 99.5 percent probability. This definition is currently contained in the CFO principles for MCEV and in Solvency II. We considered how this level compares to market price of risks. In other words, is economic capital under this definition truly economic (i.e., consistent with the market)? 1. The historical data represents only one sample of potential outcomes that could have happened and is not necessarily the mean. 2. The market is pricing risks that are currently unknown (such as black swans and paradigm shifts). 3. Economic capital modeling may have failed to priced for in the market. adequately consider the level of liquidity risk that is Failure to reflect the market price of risk in economic to the market price of risk by analyzing an A-rated bond. Using historical default rates and rating transition probloss distribution of this bond. Capital was set equal to the abilities published in Moody’s study,2 we simulated the We compared the standard definition of economic capital capital calculations may have resulted in the undercapitalization of the banking industry. If the industry replaces RBC with the currently Solvency II defined risk capital, we may foresee an undercapitalization and increasing conclude the 265 basis points is the correct level of 99.5 percentile of this distribution over average loss (i.e., the 50th percentile of credit losses) over a one-year period. We further ran the model using a five-year time horizon. failures in the insurance sector in the future. While we do not economic capital, as this price can include other factors frictional costs and tax liabilities, we do conclude that the large disconnect between the market price of risk and the We then compared this to economic capital as determined by the market price of risk (described above). Under this ed defaults was the market-consistent return on capital (for approach, the average excess historical spread over expect- such as cost of funds for potential purchasers or the assets, 2 “Corporate Default and Recovery Rates, 1920-2007,” February 2008, Moody’s. 91
Slide 92: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications What Is a Robust Level of Risk Capital? by Larry Rubin and Ziaokai (Victor) Shi Solvency II definition should lead to risk managers questioning whether their economic capital models are properly reflecting all the risks. We believe that for the regulatory mote a healthier industry, the definition of economic capital under Solvency II needs to be an improvement over the one-size-fits-all definition contained in U.S. RBC. In order reliance on complicated capital models and arrogance in believing that the complex mathematical formula in risk models overcame the limitations on input availability and enabled companies to capture returns that were greater than the cost of capital, have led Wall Street (by creating framework to become more economic and thereby pro- CDOs and similar instruments) to repackage risks and take out the systematic excess risk charges as profits, and have ratios that were in hindsight over the limit. If some of the also allowed companies to run “prudently” on leverage for it to be an improvement, it should be modified to reflect as a key input the market price of risk, and it should adopt the RBC definition that capital exists to mature an obligation. establish Robust Risk Capital strategy The authors agree with Greenspan’s remarks recently made in the New York Times, “Bad data hurt Wall Street computer models,” Greenspan said, “… whole intellectubecause the data inputted into the risk management period of euphoria… .” broker dealers that were running over 30-to-1 leverage ratios had reflected the market price of risk, they might still be around today. Establishing an “economic” solvency and performance framework based on the presumption that we are smarter than the market is not economic and is both fatally flawed and fundamentally unsound. The objective of performance reporting should be to judge whether we up-front. al edifice, however, collapsed in the summer of last year models generally covered only the past two decades a Political figures and journalists have tended to blame truly were smarter than the market and not to presume it Almost every recently failed firm was perceived as a sophisticated risk manager. We wish those painful failure examples would evoke more consideration over a system that might be fundamentally flawed rather than blaming individuals who were unlucky in being the last executives of their fallen companies. deregulation and greed for the credit crisis. However, we of the credit crisis. The real cause of the crisis was faulty would like to suggest that these were not the primary cause regulation and arrogance. Flawed regulation in the over- Larry H. Rubin, FSA, is a partner at PricewaterhouseCoopers LLP in New York, N.Y. He can be reached at larry.rubin@ us.pwc.com. Xiaokai (Victor) Shi, FSA, is senior associate at PricewaterhouseCoopers LLP in New York, N.Y. He can be reached at victor.shi@us.pwc.com. 92
Slide 93: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications insights from the insurance industry: seven simple lessons from two Actuaries on the Mortgage crisis By RANDy RoTH AND JoHN pIERCE An Industry in Question, a profession with Answers By JAMES RAMENDA the financial crisis: An Actuary’s view by louise fRAncis Actuaries Would have Made a difference By w. JAMES MACgINNITIE 93
Slide 94: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications seven simple Lessons on the Mortgage Crisis from Two actuaries by Randy Roth and John Pierce We have been observing the mortgage crisis over the recent months. To some extent, we are like Will Rogers in that all we know about the crisis “is what we read in the newsreinsurance treaties, and will then justify its actions by saying “How can we be hurt too badly? We are only taking a small share.” In some cases, additional layers of reinsurers took a share of the first reinsurer’s business. This process papers.” Even though we lack the practitioner’s in-depth lessons we have learned in the property-casualty insurance lessons—based on admittedly imperfect understanding of the exact structure of the underlying mortgage industry— are listed below: knowledge of the mortgage industry, it does seem to us that industry have relevance to the mortgage crisis. These seven was labeled the “reinsurance spiral.” These naïve reinsurers often learned just how badly they could be hurt by these “small no substitute for making your own analysis of the ultimate profitability of the underlying reinsurance treaties. Similarly, investors who relied on the various rating shares” of the underlying business. In these cases, there was 1. Never Assume Someone Else Will Serve Your Own Interest Better Than You Will. In the property-casualty insurance industry, we know that managing general agents (MGAs) have interests which pany. Because of this, it is standard practice to establish are different than the interests of the insurance comincentives that align the interests of the MGAs and the “have some skin in the game”—either by retaining a share of agencies or the reputations of the originating banks to determine which mortgage-backed securities (i.e., combinations of small slices from many individual mortgages) doing one’s own analysis of an investment risk. to purchase now understand that there is no substitute for 3. Don’t Confuse Dispersion with Diversification. From our experience in the insurance industry, we know a diversified pooling of non-correlated similar exposures insurance companies. Commonly, an MGA is required to the business it produces through an agent-owned captive or loss experience. does reduce risk for the pool and does achieve greater pre- by a commission structure that is responsive to its business’ In this mortgage crisis, it appears mortgage brokers dictability of pooled outcome. However, mortgage securitization appears to have led to global dispersion of pools consisting of relatively little slices of mortgage risk that were and mortgage originating banks were similar to unchecked MGAs. These brokers and banks had an incentive to Those mortgages would then be sold to investors, gener- positively correlated. The risks in the pools of mortgages may have been dispersed—but they were not diversified in any meaningful sense. When trouble hit the entire U.S. real curities was impacted. produce mortgages that would generate fees for them. ating additional fees—but it appears there were often no estate market, each of the little slices in these mortgage se- incentives to align the interests of those mortgage producbanks “did not have any skin in the game.” 4. Never Assume That There Is Only One Cockroach. From our experience with loss reserve problems, we have proves to be optimistic. Often further emergence of loss learned that oftentimes the first estimate of a major problem reserve deficiencies shows that the initial estimates of the say “there is never only one cockroach.” ers with the purchasers of these mortgages. The brokers and 2. It Is Always a Recipe for Disaster When a Secondary Market Risk-Taker Fails to Adequately Comprehend and Evaluate the Risk Being Accepted from a Primary Marketplace. There have been instances where a naïve reinsurer will take small shares (1 percent or 2 percent) of several larger problem’s size were too low. In “actuarial” terminology, we It appears many financial firms did not understand this concept of “one cockroach”—as their initial estimates 94
Slide 95: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications seven simple Lessons on the Mortgage Crisis from Two actuaries by Randy Roth and John Pierce of losses due to the mortgage crisis generally have been revised upwards several times. 6. The Regulator Is Not Always Your Enemy While compliance with the insurance regulatory process is often burdensome, we know it may force us to recognize regulators of our banking industry had been as rigorous as the regulators of our own insurance industry, perhaps this mortgage crisis would never have happened. 5. You Can Fool Some of the People Some of the Time but... Many smart people have devised clever and complicated schemes to camouflage risk, but it always results in a game of musical chairs where someone loses when the music a problem that we would not otherwise recognize. If the stops. For example, Lloyds has a complicated three-year accounting practice with an assumed likelihood of a syncomplicated structure served Lloyds well for many years, ages of U.S. asbestos and pollution liabilities. dicate’s ability to purchase closing reinsurance. While this it eventually failed to protect naïve investors from the ravIt appears that many of the mortgage securities that 7. Risk Assessors Need a Proven Record of Professionalism, Credibility and Objectivity The current mortgage crisis plus earlier financial collapses rating firms, the banking industry and other risk-assessing for risk assessors to restore customer confidence and trust. this decade have severely damaged the reputations of credit professionals. As a result, it will take much time and effort While the actuarial profession is not free of any blem- were purchased by investors were relatively complicated in complicated structure camouflaged the underlying risk in these instruments. their structure. In retrospect, one can wonder whether this ishes in this arena, actuaries as a whole take pride in their professional accomplishments. Our profession has taken cent experiences demonstrate the need that we continue to do so. steps to protect and enhance our reputation, and these re- Randy J. Roth, FCAS, MAAA, is currently retired but does on occasion take on some consulting work. He can be reached at corandy@prodigy.net. John Pierce, FCAS, MAAA, M.B.A., is president of John Pierce Consulting Actuary in Park Ridge, Ill. He can be reached at zorbpierce@aol.com. 95
Slide 96: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications an Industry in Question, a Profession with answers by James Ramenda The current financial crisis is sometimes described as a pany’s ability to meets its obligations and produce a profit. claims or meeting solvency requirements to this point. process of “washing away” economic excesses. Yet for a widening range of industries, it seems that more than just excesses are being washed away. Major companies, longstanding ways of doing business and thousands of jobs are disappearing, seemingly overnight in some cases. Today’s constant flow of information and the globalization of marsheets are viewed as trading positions, marked to market with each new data point. kets reinforce the “real time” nature of the crisis. Balance Indeed, no major life insurer has reported difficulty paying In theory, the impact of a market-driven decline in equity is dampened by statutory accounting principles, measures. Even under GAAP, such a decline theoretically earnings due to widened spreads. However, the reality of which judge solvency based largely on historical cost is balanced to some extent by the prospect of higher future the marketplace is that a sizeable decrease in book value may be the least strategically suited to this environment Among all financial institutions, insurance companies brings other consequences: ratings downgrades, short-sellsales and discounted capital raise-ups. Under such stress, discounted assets and raising capital at highly dilutive terms, actions which sharply reduce future profitability. ing raids, potential runs on the company, distressed asset a company may need to sustain itself by selling deeply because of the potentially decades-long nature of their liabilities. And among insurance companies, life insurers face particular pressure owing to three characteristics of their business: (1) they have the longest duration liabilities, i.e., permanent life insurance and long-term annuities, (2) these are demand liabilities, and (3) these lines have the highest asset leverage relative to capital. The leveraged effect on balance sheets has been severe, with equity prove to be temporary, insurers will still need to incorporisk management scenarios, e.g., the VIX above 60 (and up to 500 basis points; performing AAA commercial mortgage spreads of 600 basis points. Perhaps these Moreover, even if the recent severe declines in equity rate the observed market data into their forward-looking surging to nearly 90); investment grade bond spreads accounts of many major publicly traded life insurers down to fall further by year-end. The S&P 500 life insurance component is off more than 50 percent in 2008 through mid-November, and several companies have suffered harsh judgment will prove correct in the long term is raising some new questions for the industry. 20 percent or more through nine months of 2008 and likely extremes reflect “fire sales” conducted by severely distressed sellers, but they cannot be discounted as 1per-1,000-year events, “black swans,” or “Great Depresdeleveraging of the life insurance business, as is occurring in other financial industries. substantially more severe declines. Whether the market’s anyone’s guess at this point, but at a minimum the crisis is The central question is that of viability. Specifically, sion” scenarios. Recognition of increased risk will require is the configuration of life insurers’ products and balance sheets economically viable under current accounting and investment guarantees in their life and annuity policies. The assets backing these products are chosen to mature the ongoing market conditions? Life insurers make long-term raising, but at the present time the public markets are not fractions of their book values, some as low as 20 percent Deleveraging can be accomplished through capital- accommodating. Many leading companies are trading at (of book value as stated, i.e., reflecting fair value accounting). junk bond spreads despite investment grade ratings. It may be that markets simply distrust insurers’ balance sheets, or liabilities over the long term, not to match them in value moment to moment. The result is that interim swings in equity can be large, but not necessarily meaningful to a com96 The debt of some household names in the industry trades at are discounting still worse days to come. Perhaps markets
Slide 97: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications an Industry in Question, a Profession with answers by James Ramenda are simply in a panic, but it is noteworthy that generally higher valuations are being awarded to property-casualty and certain health insurers vs. life insurers, these having shorter duration business, no demand liabilities and lower asset leverage. Markets may be rendering their verdict on the question of viability of the underlying business, i.e., ment-based liabilities underwritten by life insurers. ties, able to buy, hold and manage these assets when few others could or would. In the Great Depression, many insurers ended up owning vast swaths of land and innumerable commercial and apartment buildings, property which eventually proved highly valuable and contributed to the foundation of many industry giants. But even before they won’t provide the capital to support long-term investWith private capital not forthcoming, it is likely the the current economic crisis, the ability of insurers to act accounting measures (FAS 159 and its forerunner, FAS 115) due to the potential wide swings in market value such averse investors in the future. assets entail. Insurers seem destined to be even more riskAll of this is not to pass judgment as to whether fair- in this capacity had already become limited by fair-value life insurance industry will join the banking industry (and perhaps other industries, e.g., auto manufacturers) and Treasury. It has also been suggested that insurers receive recapitalize through direct investment from the U.S. FDIC-type backing to shore up policyholder confidence. If that the time for federal regulation over insurer solvency would be near. As with banks, if the federal government is going to backstop an industry, it is going to have a greater insurance industry, many companies would now welcome a these steps were taken, it would be reasonable to assume value accounting is inherently good or bad, or whether more observation that changes to accounting rules have tended to regulation is inherently good or bad. Rather, it is simply an diminish the industry’s ability to act as a long-term stabilizing force in the economy. The depth of the current crisis may put a permanent stamp on this trend. and the capital markets’ dire assessment of the industry For many of us in the industry, this seems like a sadly regulatory say. Though once resisted by much of the move toward this Canadian-type approach to regulation. It might also be logical to expect greater regulation of insurthe enterprise at that level. ance holding companies to monitor the build-up of risk to Whatever its form, future regulation will almost counterproductive outcome. But we have to answer risk-averse, it would not be verging on a bailout from actuarial profession needs to be heard in several key areas: critics who would argue that had the industry been more taxpayers. This is a pivotal moment for the industry. The 1. Determine whether the industry’s business model (i.e., certainly reinforce deleveraging. If so, insurers will need tees they provide, limit the risks they cover and attempt to to back their risks with more capital, reduce the guaranincrease their prices to maintain returns. For economists, this suggests a lower equilibrium quantity point for risk transfer. In everyday terms, it means less coverage for tion for businesses. current products, investment practices, capital ratios, holding company leverage standards, etc.) is viable in view of ongoing capital market conditions. This in- breadwinners, less income for retirees and less continuaOf course, in addition to their risk-bearing function, cludes examining the impact of fair value accounting and whether it has created additional risk or simply identified it. 2. Every effort must be made to see that the profession’s insurers are also investors. And being long-term businesses, remaining in the world. Historically, insurance companies were among the buyers of last resort for distressed proper- they are by definition among the few long-term investors analyses inform the development of future regulation, including the form, terms and amounts of potential capital infusions or guarantees from the government. 97
Slide 98: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications an Industry in Question, a Profession with answers by James Ramenda 3. Develop a system of reporting the financial condition of life insurers that would replace the tangle of GAAP statements, “non-GAAP” measures and statutory statements with a single set of highly transparent unified set of experience assumptions, a validation of industry’s business model. It is possible policyholders may someday follow suit. For the actuarial profession, this is Public markets doubt the viability of the life insurance financial statements. Such statements should include a the assumptions using actual experience and explicit means of reconciling measures used in the capital vency. This will necessarily mean the transformation to a set of transparent adjustments to prevailing GAAP. markets with those used in the determination of solof solvency accounting from a separate set of principles a challenge on par with the need for Social Security and Medicare reform. It must provide the objective tools and ity of the life insurance industry. analysis needed to either reaffirm or re-establish the viabil- James Ramenda, FSA, is managing director of Northington Partners Inc., in Avon, Conn. He can be reached at jr@northington.net. 98
Slide 99: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Financial Crisis: an actuary’s View by Louise Francis In a recent Casualty Actuarial Society VALCON1 list eWhen the data is transposed, so that rows represent year of origin, and columns represent development age (the number of years after the origin year, with one denoting the origin year), the loss development factor method, also known as the chain-ladder method, can be applied to estimate ultimate foreclosure rates for each origin year. An estimate of these ultimate rates may provide insight into the magnitude of the subprime mortgage problem. In order to apply the chain-ladder method, cumulative foreclosure rates are needed. These are derived from the calendar year incremental rates for each cohort and are shown in Table 2. mail, Gary Venter distributed foreclosure rates for cohorts of subprime mortgages organized by origination year. Venter noted that when the data are transposed, they have the form of a loss development triangle, a standard tool applied by property and casualty actuaries to estimate ultimate liabilities. He provided some qualitative insights and conclusions that could be drawn by an actuary from the information. Below is a further elaboration of insights that can be drawn by applying actuarial techniques to the data. The insights publications on the topic of subprime mortgages. The author’s conclusion is that subprime mortgages constituted a Ponzi scheme and could have been avoided. The foreclosure rate data is presented below with one derived from the data are augmented by results from recent needed to bring the cumulate rate as of a given age for a given year to a maturity of one year beyond the given age. At the bottom of Table 3 are the age-to-ultimate Table 3 displays the age-to-age factors, or the factor adjustment to the original data: the values on the diagonal, factors. These are the cumulative product of the age-to-age factors starting from the oldest maturity and working backwards to the youngest maturity. They are a key component which were evaluated as of September, and thus were divided by 0.75 to bring them to an annual basis, consistent with all the other entries. For the adjustment to be reasonable, the foreclosures must occur uniformly throughout the year. That this assumption may not hold is a limitation on the analysis affecting the uncertainty of results. (See Table 1 below). of the estimate of ultimate rates. As foreclosure rates as of nine years (the oldest year for which we have data) from origination do not appear to be at ultimate (i.e., further development will likely occur), a “tail factor” is needed. TaBLE 1 origination year Foreclosure Year 0 1 2 3 4 5 6 7 8 1999 0.013 0.063 0.055 0.049 0.023 0.021 0.008 0.006 0.003 2000 0.015 0.069 0.060 0.034 0.025 0.012 0.007 0.004 2001 0.019 0.072 0.058 0.042 0.019 0.012 0.006 2002 0.011 0.055 0.046 0.024 0.016 0.008 2003 0.008 0.041 0.031 0.022 0.011 2004 0.009 0.039 0.017 0.025 2005 0.010 0.064 0.062 2006 0.026 0.103 2007 0.040 1 The VALCON list is a list sponsored by the Committee on the Theory of Risk of the Casualty Actuarial Society and is a list that is subscribed to by actuaries and insurance professionals. The community of subscribers share research, ideas and musings related to the valuation of Contingent obligations. 99
Slide 100: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Financial Crisis: an actuary’s View by Louise Francis The inverse power curve described by Sherman (Sherman, 1987) was used to estimate this tail. Table 4 displays the application of the age-to-ultimate are very uncertain. The estimates are considered especially unstable for data of low maturity, such as that of the 2007 and 2006 years. Moreover, some of the assumptions additional uncertainty to the estimates. factors, to the diagonal (as of year-end 2007) cumulative underlying the chain ladder may be violated, adding yet Venter (1998) describes techniques that can be used to foreclosure rates to estimate ultimate foreclosure rates for foreclosure rates are estimated to be in excess of 40 percent for 2006 and over two-thirds for 2007. each origination year. Using the chain-ladder technique, test whether the chain-ladder assumptions are violated. One method, or any other actuarial development techniques, TaBLE 2 Estimated ultimates derived form the chain-ladder When this test was performed, the age 1–2 (also referred ladder assumptions. As a result,3 the analysis for the 2007 of the tests involves an application of regression analysis.2 to age 12 months to 24 months) factor violated the chain- Cumulative default Rates @12/31/07 Development Age Year 1999 2000 2001 2002 2003 2004 2005 2006 2007 TaBLE 3 1.000 0.013 0.015 0.019 0.011 0.008 0.009 0.010 0.026 0.040 2.000 0.076 0.084 0.090 0.066 0.050 0.048 0.074 0.128 3.000 0.131 0.144 0.148 0.111 0.081 0.064 0.136 4.000 0.179 0.177 0.191 0.135 0.103 0.089 5.000 0.202 0.202 0.209 0.151 0.114 6.000 0.223 0.214 0.221 0.158 7.000 0.231 0.221 0.228 8.000 0.236 0.225 9.000 0.239 age-to-age Factors Development Age Year 1999 2000 2001 2002 2003 2004 2005 2006 Average selected Age to ultimate 12-24 5.869 5.573 4.876 6.150 6.049 5.570 7.577 5.005 5.834 5.800 16.779 24-36 1.714 1.719 1.644 1.691 1.627 1.344 1.845 36-48 1.371 1.233 1.285 1.213 1.276 1.383 48-60 1.128 1.141 1.099 1.116 1.107 60-72 1.101 1.059 1.056 1.052 72-84 1.035 1.033 1.029 84-96 1.024 1.018 96-108 1.012 Tail 1.698 1.700 2.893 1.294 1.300 1.702 1.118 1.100 1.309 1.067 1.067 1.19 1.032 1.032 1.115 1.021 1.021 1.08 1.012 1.012 1.058 1.0453 1.0453 100
Slide 101: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Financial Crisis: an actuary’s View by Louise Francis year was adjusted. The results are shown in Table 5. Using this adjustment, the estimated rate for 2007 exceeds 50 percent. Note that the use of this adjustment addresses the violation of certain assumptions underlying the chain-ladder technique. It does not significantly reduce the uncertainty in the estimates, which, given the sparseness and variability 2007 foreclosure year’s rates to an annual basis, is quite large. (See Table 5 below). The estimates in Table 5 based on the chain ladder (with adjustment) show a dramatic increase between 2004 and 2007. Under a scenario of real estate price depreciation, such default rates could be expected to be ruinous. Accord- of the data and the crude assumptions needed to adjust the TaBLE 4 default Rates developed to ultimate year Current year End default Rate (1) 1999 2000 2001 2002 2003 2004 2005 2006 2007 0.239 0.225 0.228 0.158 0.114 0.089 0.136 0.128 0.040 age To ultimate (2) 1.058 1.058 1.080 1.115 1.190 1.309 1.702 2.893 16.779 ing to Demyanyk and Hemert (2008), the deterioration in ultimate default Rate (3)=(1)*(2) 0.253 0.238 0.246 0.177 0.136 0.117 0.231 0.371 0.673 notes: (1) All rates adjusted to 12 month basis by dividing by .75 TaBLE 5 default Rates developed to ultimate year adj Current year End default Rate (1) 1999 2000 2001 2002 2003 2004 2005 2006 2007 0.239 0.225 0.228 0.158 0.114 0.089 0.136 0.128 0.187 age To ultimate (2) 1.058 1.058 1.080 1.115 1.190 1.309 1.702 2.893 2.893 ultimate default Rate (3)=(1)*(2) 0.253 0.238 0.246 0.177 0.136 0.117 0.231 0.371 0.540 notes: (1) 2007 rate adjusted to age 24 using: .02 + 3.129 * Age 1 rate + age 1 rate 2 3 The incremental rates for a given maturity are regressed on the cumulative rates for the prior maturity. If the constant is significant, and/or the coefficient is not significant (typically at the 95 percent level), the assumptions are likely to be violated. The fitted age 1–2 regression parameters were used to adjust the 2007 rates to age 24, and then the chain-ladder technique was applied. 101
Slide 102: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications The Financial Crisis: an actuary’s View by Louise Francis foreclosure rates should have been known to the mortgage lenders as early as 2005. Their analysis applied logistic regression to loan level data and found that the quality of loans declined for six consecutive years: “Problems could have been detected long before the crisis, but they were masked by house price appreciation.” unseen, motivated by the belief that it would be resold at a handsome profit. In Galbraith’s words, the real estate investors “proceeded to build a world of speculative makebelieve. This is a world inhabited by people who do not have to be persuaded to believe, but by people who want debacle was Mr. Charles Ponzi, and the scenario came to be known as a “Ponzi scheme.” an excuse to believe” (p. 8). One of the principals in the appears to be inherent in their design, as they were not being completely discharged by the debtor. According to Gorton, serial refinancing was intended and built into the Moreover, the problem with subprime mortgages designed to be held to maturity, with interest and principal mess was none other than a Ponzi scheme repackaged into 21st century financial engineering clothes. What makes this scheme particularly disastrous is that the 21st century Ponzi mortgages were packaged and sold to investors and then It is the belief of this author that the subprime mortgage product when the mortgages were sold. To protect the lender from the “risky borrower,” the loans were structured to be then refinanced. As price appreciation of the underlying cur before the rates of an ARM or of a mortgage with an held for a relatively short period (two to three years) and asset was expected, the refinancing was anticipated to ocinitial teaser rate were adjusted upwards and the mortgage payment exceeded the debtors’ resources. However, the refinancing was at the option of the lender, so if houses stuck in a mortgage that under any realistic scenario trillions of dollars of derivatives were constructed based on the underlying mortgages, magnifying the problem by orders of magnitude. value-at-risk and dynamic analysis models to their hearts’ content. If the founding principle underlying an investment is that of a speculative bubble scheme, the scenario is virtually guaranteed to come to a bad end. References Demyanyk and Hemert. 2008. Understanding the Subprime Mortgage Crisis, SSRN paper, August, ssrn.com/ abstract=1020396. Gorton, G. 2008. “The Subprime Panic,” National Bureau of Economic Research, October. Sherman, R. 1984. “Extrapolating, Smoothing and InterpoActuarial Society. The most brilliant analysts can run their option pricing, failed to appreciate, the borrower faced the risk of being exceeded his/her ability to pay. According to Gorton, “The every two to three years.” appreciation of the house became the basis for refinancing The scenario is reminiscent of another speculative bubble based on the expectation of real estate price appre- ciation without end, and the anticipation of fantastic wealth based on the appreciation. The scheme is described in some detail by John Kenneth Galbraith in his landmark book, The Great Crash. The real estate bubble occurred in Florida estate bubble), a state with a congenial winter climate, of an improved transportation system and spend their (one of the states most seriously affected by the latest real where people of means were expected to avail themselves winters there in increasing numbers. Land was bought sight lating Development Factors,” Proceedings of The Casualty Venter, G. 1997. “Testing the Assumptions of Age-to Age Factors,” Proceedings of The Casualty Actuarial Society. Louise Francis, FCAS, MAAA, is the founder and consulting principal of Francis Analytics and Actuarial Data Mining, Inc., in Philadelphia, Pa. She can be reached at louise_francis@msn.com. 102
Slide 103: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications actuaries Would Have Made a Difference by W. James MacGinnitie In the current financial markets turmoil, it is tempting to ask whether things might have turned out differently. ‘What if’ questions were a favorite of a history professor under whom I once studied. They of course are speculative, but have been less if more actuaries had been involved. I offer 10 reasons: 1. Actuaries understand that the distribution function for pects of operations that might have an impact on experience. Because they are trained in all aspects of the to ask of whom, and they quantify model drift and current relevance. 4. enterprise, they are well-grounded in what questions in this case I think the pain in our financial markets would A recent example is the business of worker’s compen- Actuaries understand spirals, and seek to avoid them. most risks is not the bell curve or normal distribution, but rather one of several distribution functions that in capital markets used value-at-risk, which is based have longer, fatter tails. Too much of the risk analysis on the normal distribution. That makes it easy to sation ‘carve out,’ when much of the medical and time loss coverage in worker’s compensation policies was the derivatives of derivatives that are part of the curreinsured in increasingly complex structures, akin to rent problems. The basic proposition was to take a face, and by packaging, slicing, dicing and repackagof the way, turn lead into gold. It didn’t work in that manipulate mathematically, but it usually understates the chances of bad outcomes, both in frequency and in amount. Actuaries working in catastrophe reinsurance, and in coverages impacted by large verdicts, underneed to be included in the model. 2. stand that large but infrequent events (‘Black Swans’) Actuaries understand that while choosing the right business that was marginally profitable at the mine ing, with managers and brokers getting paid every step case, but it did lead to common prohibitions of reinsur- ance on reinsurance (derivatives on derivatives). Had similar prohibitions existed in the derivatives markets, a great deal of pain could have been avoided. 5. Actuaries are accustomed to developing values for lia- model is very important, it’s even more important to needs to be analyzed, not just the trades from the last calibrate it appropriately. A rich, long-term data set few months. And extreme events shouldn’t be excluded some actuarial models, such as catastrophe reinsurthat encompasses those extreme events. 3. bilities where no deep liquid market exists, such as lawsuit liability. They frequently develop values for pension obligations, long-term care insurance and claims that have not yet been reported to the insurer. on the basis that “that will never happen again.” In ance, the only thing that really matters is the fat tail Actuaries understand ‘model drift.’ Most accounts of Similar techniques would be useful for many of the market values. While such values would necessarily alistic than those quoted by someone who wants to avoid acquiring the asset at any price. 6. Actuaries are used to taking a long-term view. With assets that are currently being marked to nonexistent be uncertain approximations, they would be more re- the subprime meltdown report that mortgages issued before 2005, and into 2006, have performed as ing standards deteriorated. Actuaries know that when modeled. Later issues have not, because underwritunderwriting standards are lowered, worse experience results, and should be reflected in the price, and in any reserves set aside to pay losses. And they are trained pension obligations extending for decades, as do life insurance policies and benefits, and also long-tailed things will play out over the long term. casualty coverages, actuaries have to think about how to inquire about changes in underwriting and other as- 103
Slide 104: R I s k M a n a g e M e n T: the current financial crisis, lessons learned and future implications actuaries Would Have Made a Difference by W. James MacGinnitie 7. concept that would be helpful in the capital markets. pectations, then measure actual results and use those The Actuarial Control Cycle is a well-developed 10. Actuaries accept a quasi-fiduciary obligation. Since In simple terms, it requires the actuary to model exmeasurements to recalibrate the model. This kind of the sand. The model can be complex, with different tal markets is based on daily procedures, which can cause one to lose sight of both the forest and the trees by focusing on twigs. 8. the pension plans and insurance companies they customarily serve will need to deliver on their promises many years into the future, paying benefits to survi- vors and retirees, actuaries understand that they have an obligation to do their best to make sure that those benefits will be paid when they are needed. The contrast to the trader’s mentality is stark. feedback loop helps adjust course before the ship hits feedback loops and frequencies. Too much of the capi- surers and pension plans that failed, but the frequency is actuary’s advice. Actuaries aren’t perfect. There are examples of in- relatively small, and in many cases it was in spite of the As regulators, legislators and central banks seek to Actuaries are accustomed to transparency. Their reguan actuarial report to back their opinion, and it must appraise the conclusions. lators require it. Their professional standards require contain sufficient detail so that another actuary can design a better future, it would be helpful to include more actuarial training and thinking. 9. Actuaries have professional standards. They should only do work for which they are qualified. They should follow professional guidance from their accrediting organizations. They must continue their professional discipline code. education to maintain currency. They are subject to a W. James MacGinnitie, FSA, MAAA, FCAS, is an actuary and consultant in Atlanta, Ga. He can be reached at jimmacg@ soa.org. 104