Reading About the Financial Crisis: A Twenty-One-Book Review

The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters.

Citation Lo, Andrew W. "Reading about the Financial Crisis: A Twenty- One-Book Review." Journal of Economic Literature, 50(1): 151- 78 (2012). As Published http://dx.doi.org/10.1257/jel.50.1.151 Publisher American Economic Association

Version Final published version Accessed Mon Jun 19 17:49:15 EDT 2017 Citable Link http://hdl.handle.net/1721.1/75360 Terms of Use Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. Detailed Terms Journal of Economic Literature 2012, 50:1, 151–178 http:www.aeaweb.org/articles.php?doi=10.1257/jel.50.1.151

Reading About the Financial Crisis: A Twenty-One-Book Review

Andrew W. Lo*

The recent financial crisis has generated many distinct perspectives from various quarters. In this article, I review a diverse set of twenty-one books on the crisis, eleven written by academics, and ten written by journalists and one former Treasury Secretary. No single narrative emerges from this broad and often contradictory collection of interpretations, but the sheer variety of conclusions is informative, and underscores the desperate need for the economics profession to establish a single set of facts from which more accurate inferences and narratives can be constructed. (JEL E32, E44, E52, G01, G21, G28)

1. Introduction which completely satisfies our need for redemption and closure. Although the movie n Akira Kurosawa’s classic 1950 film won many awards, including an Academy IRashomon, an alleged rape and a mur- Award for Best Foreign Language Film in der are described in contradictory ways by 1952, it was hardly a commercial success in four individuals who participated in various the United States, with total U.S. earnings of aspects of the crime. Despite the relatively $96,568 as of April 2010.1 This is no surprise; clear set of facts presented by the differ- who wants to sit through 88 minutes of vivid ent narrators—a woman’s loss of honor and story-telling only to be left wondering who- her husband’s death—there is nothing clear dunit and why? about the interpretation of those facts. At Six decades later, Kurosawa’s message the end of the film, we’re left with several of multiple truths couldn’t be more rel- mutually inconsistent narratives, none of evant as we sift through the wreckage of the worst financial crisis since the Great Depression. Even the Financial Crisis * MIT Sloan School of Management and AlphaSimplex Group, LLC. I thank Zvi Bodie, Jayna Cummings, Janet Inquiry Commission—a prestigious biparti- Currie, Jacob Goldfield, Joe Haubrich, Debbie Lucas, Bob san committee of ten experts with subpoena Merton, Kevin Murphy, and Harriet Zuckerman for help- power who deliberated for eighteen months, ful discussions and comments. Research support from the MIT Laboratory for Financial Engineering is gratefully acknowledged. The views and opinions expressed in this article are those of the author only, and do not necessarily represent the views and opinions of MIT, AlphaSimplex, 1 See http://www.the-numbers.com/movies/1950/0RASH. any of their affiliates or employees, or any of the individuals php. For comparison, the first Pokemon movie, released in acknowledged above. 1999, has grossed $85,744,662 in the United States so far.

151 152 Journal of Economic Literature, Vol. L (March 2012) interviewed over 700 witnesses, and held much higher yields than straight corporate nineteen days of public hearings—presented bonds with identical ratings, apparently for three different conclusions in its final report. good reason.4 Disciples of efficient markets Apparently, it’s complicated. were less likely to have been misled than To illustrate just how complicated it can those investors who flocked to these instru- get, consider the following “facts” that have ments because they thought they had identi- become part of the folk wisdom of the crisis: fied an undervalued security. As for the second point, in a recent study 1. The devotion to the Efficient Markets of the executive compensation contracts Hypothesis led investors astray, causing at 95 banks, Fahlenbrach and Stulz (2011) them to ignore the possibility that secu- conclude that CEOs’ aggregate stock and ritized debt2 was mispriced and that the option holdings were more than eight times real-estate bubble could burst. the value of their annual compensation, and 2. Wall Street compensation contracts were the amount of their personal wealth at risk too focused on short-term trading prof- prior to the financial crisis makes it improb- its rather than longer-term incentives. able that a rational CEO knew in advance of Also, there was excessive risk-taking an impending financial crash, or knowingly because these CEOs were betting with engaged in excessively risky behavior (exces- other people’s money, not their own. sive from the shareholders’ perspective, that 3. Investment banks greatly increased is). For example, Bank of America CEO Ken their leverage in the years leading up Lewis was holding $190 million worth of com- to the crisis, thanks to a rule change pany stock and options at the end of 2006, by the U.S. Securities and Exchange which declined in value to $48 million by the Commission (SEC). end of 2008,5 and Bear Stearns CEO Jimmy Cayne sold his ownership interest in his com- While each of these claims seems perfectly pany—estimated at over $1 billion in 2007— plausible, especially in light of the events of for $61 million in 2008.6 However, in the 2007–09, the empirical evidence isn’t as clear. case of Bear Stearns and Lehman Brothers, The first statement is at odds with the fact Bebchuk, Cohen, and Spamann (2010) have that, prior to 2007, collateralized debt obliga- argued that their CEOs cashed out hundreds tions (CDOs),3 the mortgage-related bonds at of millions of dollars of company stock from the center of the financial crisis, were offering 2000 to 2008, hence the remaining amount

2 “Securitized debt” is one of the financial innovations at 4 For example, in an April 2006 publication by the the heart of the crisis, and refers to the creation of bonds Financial Times, reporter Christine Senior (2006) filed of different seniority (known as “tranches”) that are fixed- a story on the enormous growth of the CDO market in income claims backed by collateral in the form of large Europe over the previous years, and quoted Nomura’s portfolios of loans (mortgages, auto and student loans, estimate of $175 billion of CDOs issued in 2005. When credit card receivables, etc.). asked to comment on this remarkable growth, Cian 3 A CDO is a type of bond issued by legal entities that O’Carroll, European head of structured products at Fortis are essentially portfolios of other bonds such as mort- Investments replied, “You buy a AA-rated corporate bond gages, auto loans, student loans, or credit-card receiv- you get paid Libor plus 20 basis points; you buy a AA-rated ables. These underlying assets serve as collateral for the CDO and you get Libor plus 110 basis points.” CDOs; in the event of default, the bondholders become 5 These figures include unrestricted and restricted owners of the collateral. Because CDOs have different stock, and stock options valued according to the Black- classes of priority, known as “tranches,” their risk/reward Scholes formula assuming maturity dates equal to 70 characteristics can be very different from one tranche percent of the options’ terms. I thank Kevin Murphy for to the next, even if the collateral assets are relatively sharing these data with me. homogeneous. 6 See Thomas (2008). Lo: Reading About the Financial Crisis 153

Assets-to-equity ratio 35 to 1

30 to 1

25 to 1

20 to 1

15 to 1

Year and quarter 10 to 1 Goldman Sachs Merrill Lynch Lehman Brothers 5 to 1 Morgan Stanley

0 to 1 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 1. Ratio of Total Assets to Equity for Four Broker-Dealer Holding Companies from 1998 to 2007 Source: U.S. Government Accountability Office Report GAO–09–739 (2009, figure 6).

of equity they owned in their respective com- Like World War II, no single account of panies toward the end may not have been this vast and complicated calamity is suffi- sufficiently large to have had an impact on cient to describe it. Even its starting date is their behavior. Nevertheless, in an extensive unclear. Should we mark its beginning at the empirical study of major banks and broker- crest of the U.S. housing bubble in mid-2006, dealers before, during, and after the finan- or with the liquidity crunch in the shadow cial crisis, Murphy (2011) concludes that the banking system7 in late 2007, or with the Wall Street culture of low base salaries and bankruptcy filing of Lehman Brothers and outsized bonuses of cash, stock, and options actually reduces risk-taking incentives, not unlike a so-called “fulcrum fee” in which portfolio managers have to pay back a portion 7 The term “shadow banking system” has developed sev- of their fees if they underperform. eral meanings ranging from the money market industry to And as for the leverage of investment the hedge fund industry to all parts of the financial sector banks prior to the crisis, figure 1 shows much that are not banks, which includes money market funds, investment banks, hedge funds, insurance companies, higher levels of leverage in 1998 than 2006 for mortgage companies, and government sponsored enter- Goldman Sachs, Merrill Lynch, and Lehman prises. The essence of this term is to differentiate between Brothers. Moreover, it turns out that the SEC parts of the financial system that are visible to regulators and under their direct control versus those that are outside rule change had no effect on leverage restric- of their vision and purview. See Pozsar, et al. (2010) for an tions (see section 4 for more details). excellent overview of the shadow banking system. 154 Journal of Economic Literature, Vol. L (March 2012) the “breaking of the buck”8 by the Reserve To that end, I provide brief reviews of Primary Fund in September 2008? And we twenty-one books about the crisis in this have yet to reach a consensus on who the essay, which I divide into two groups: principal protagonists of the crisis were, and those authored by academics, and those what roles they really played in this drama. written by journalists and former Treasury Therefore, it may seem like sheer folly Secretary Henry Paulson. The books in the to choose a subset of books that econo- first category are: mists might want to read to learn more about the crisis. After all, new books are • Acharya, Richardson, van Nieuwerburgh, still being published today about the and White, 2011, Guaranteed to Fail: Great Depression, and that was eight Fannie Mae, Freddie Mac, and the decades ago! But if Kurosawa were alive Debacle of Mortgage Finance. Princeton today and inclined to write an op-ed piece University Press. on the crisis, he might propose Rashomon • Akerlof and Shiller, 2009, Animal as a practical guide to making sense of Spirits: How Human Psychology Drives the past several years. Only by collecting the Economy, and Why It Matters for a diverse and often mutually contradic- Global Capitalism. Princeton University tory set of narratives can we eventually Press. develop a more complete understanding • French et al., 2010, The Squam Lake of the crisis. While facts can be verified Report: Fixing the Financial System. or refuted—and we should do so expedi- Princeton University Press. tiously and relentlessly—we must also • Garnaut and Llewellyn-Smith, 2009, recognize the possibility that more com- The Great Crash of 2008. Melbourne plex truths are often in the eyes of the University Publishing. beholder. This fact of human cognition • Gorton, 2010, Slapped by the Invisible doesn’t necessarily imply that relativ- Hand: The Panic of 2007. Oxford ism is correct or desirable; not all truths University Press. are equally valid. But because the par- • Johnson and Kwak, 2010, 13 Bankers: ticular narrative that one adopts can color The Wall Street Takeover and the Next and influence the subsequent course of Financial Meltdown. Pantheon Books. inquiry and debate, we should strive at • Rajan, 2010, Fault Lines: How Hidden the outset to entertain as many interpre- Fractures Still Threaten the World tations of the same set of objective facts Economy. Princeton University Press. as we can, and hope that a more nuanced • Reinhart and Rogoff, 2009, This Time Is and internally consistent understanding of Different: Eight Centuries of Financial the crisis emerges in the fullness of time. Folly. Princeton University Press. • Roubini and Mihm, 2010, Crisis Economics: A Crash Course in the Future of Finance. Penguin Press. • Shiller, 2008, The Subprime Solution: How Today’s Global Financial Crisis 8 This term refers to the event in which a money mar- ket fund can no longer sustain its policy of maintaining a Happened and What to Do About It. $1.00-per-share net asset value of all of its client accounts Princeton University Press. because of significant market declines in the assets held • Stiglitz, 2010, Freefall: America, Free by the fund. In other words, clients have lost part of their principal when their money market fund “breaks the buck” Markets, and the Sinking of the World and its net asset value falls below $1.00. Economy. Norton. Lo: Reading About the Financial Crisis 155 and those in the second category are: natural. The decision to include other books in the mix was motivated by the fact that, • Cohan, 2009, House of Cards: A Tale of as economists, we should be aware not only Hubris and Wretched Excess on Wall of our own academic narratives, but also of Street. Doubleday. populist interpretations that may ultimately • Farrell, 2010, Crash of the Titans: Greed, have greater impact on politicians and pub- Hubris, the Fall of Merrill Lynch, and lic policy. Whereas the academic authors the Near-Collapse of Bank of America. are mainly interested in identifying under- Crown Business. lying causes and making policy prescrip- • Lewis, 2010, The Big Short: Inside the tions, the journalists are more focused on Doomsday Machine. Norton. personalities, events, and the cultural and • Lowenstein, 2010, The End of Wall political milieu in which the crisis unfolded. Street. Penguin Press. Together, they paint a much richer pic- • McLean and Nocera, 2010, All the Devils ture of the last decade, in which individual Are Here: The Hidden History of the actions and economic circumstances inter- Financial Crisis. Portfolio/Penguin. acted in unique ways to create the perfect • Morgenson and Rosner, 2011, Reckless financial storm. Endangerment: How Outsized Ambition, Few readers will be able to invest the time Greed, and Corruption Led to Economic to read all twenty-one books, which is all the Armageddon. Times Books/Henry Holt more motivation for surveying such a wide and Company. range of accounts. By giving readers of the • Paulson, 2010, On the Brink: Inside the Journal of Economic Literature a panoramic Race to Stop the Collapse of the Global perspective of the narratives that are avail- Financial System. Business Plus. able, I hope to reduce the barriers to entry to • Sorkin, 2009, Too Big to Fail: The Inside this burgeoning and important literature. In Story of How Wall Street and Washington section 2, I review the books by academics; Fought to Save the Financial System in section 3, I turn to the books by journalists from Crisis—and Themselves. Viking. and former Treasury Secretary Paulson; and • Tett, 2009, Fool’s Gold: How the Bold I conclude in section 4 with a brief discus- Dream of a Small Tribe at J.P. Morgan sion of the challenges of separating fact from Was Corrupted by Wall Street Greed and fantasy with respect to the crisis. Unleashed a Catastrophe. Free Press. • Zuckerman, 2009, The Greatest Trade 2. Academic Accounts Ever: The Behind-the-Scenes Story of How John Paulson Defied Wall Street Academic accounts of the crisis seem to and Made Financial History. Broadway exhibit the most heterogeneity, a very posi- Books. tive aspect of our profession that no doubt contributes greatly to our collective intelli- I didn’t arrive at this particular mix of books gence. By generating many different narra- and the roughly even split between aca- tives, we’re much more likely to come up demic and journalistic authors with any par- with new insights and directions for further ticular objective in mind; I simply included research than if we all held the same con- all the books that I’ve found to be particu- victions. Of these titles, Robert J. Shiller’s larly illuminating with respect to certain The Subprime Solution: How Today’s aspects of the crisis. Reviewing the books Global Financial Crisis Happened, and authored by our colleagues is, of course, What to Do about It was the first out of the 156 Journal of Economic Literature, Vol. L (March 2012) gate. Written for the educated layperson, it mistaken beliefs about economic outcomes, appears from internal evidence that Shiller’s then the cure must be inoculation against short book was completed by April 2008, and further mistaken beliefs and eradication of published in August of that year. This book currently mistaken ones. Much as the gov- captures the view, which became current at ernment plays a vital role in public health the time, that the crisis was principally about against the spread of contagious disease, the unraveling of a bubble in housing prices. Shiller recommends government subsidies Shiller ought to know about such things: to provide financial advisors for the less years ago, he and his collaborator Karl E. wealthy, and greater government moni- Case pioneered a new set of more accurate toring of financial products, analogous to home-price indexes based on repeat sales the consumer product regulatory agencies rather than appraisal values, now known as already in existence in the United States. the “S&P/Case–Shiller Home Price Indices” More speculatively, he also suggests using and maintained and distributed by Standard financial engineering to create safer financial & Poor’s. Thanks to Case and Shiller, we can products and markets. Finally, since bubbles now gauge the dynamics of home prices both represent a failure of the correct informa- regionally and nationally. tion to propagate to the public, Shiller calls Much of Shiller’s exposition on real estate for greater transparency, improved financial bubbles will be familiar to readers of the sec- databases, and new forms of economic mea- ond edition of Irrational Exuberance. Rather surement made more intuitive for the gen- than scarcity driving up real estate prices— eral public. a theory that he demonstrates is incomplete Shiller’s stylized description of the hous- at best—he postulates a general contagion ing bubble largely passes over how its of mistaken beliefs about future economic bursting transmitted ill effects to the rest of behavior, citing Bikhchandani, Hirshleifer, the economy. In August 2008, however, at and Welch’s (1992) theoretical work on infor- the same time that his book was released, mational cascades to support this notion, but a much more detailed account of the also John Maynard Keynes’s famous concept mechanics behind the crisis in short-term of “animal spirits.” Overall, Shiller’s discus- credit markets was presented at the annual sion of underlying causes is rather thin, Jackson Hole Conference sponsored by the perhaps due to his writing for a general audi- Federal Reserve Bank of Kansas City. The ence. Shiller would expand more fully on his paper by Gary Gorton, simply titled, “The theory of animal spirits in his 2009 book with Panic of 2007,” quickly became a hot topic George Akerlof (reviewed below), as Shiller of discussion among economists and policy­ mentions in his acknowledgements, so per- makers, and—something new under the haps a little intellectual “crowding out” took sun—a samizdat for interested laypeople on place as well. the Internet. This paper was republished in With the benefit of three short years March 2010 with additional material and of hindsight, Shiller’s policy prescriptions analysis on the shadow banking system as appear laudable but almost utopian. Past the Slapped by the Invisible Hand: The Panic necessity of some bailouts, Shiller proposes of 2007. “democratizing finance—extending the Much of Gorton’s account is descriptive. application of sound financial principles to a Among other things, it’s a crash course (no larger and larger segment of society” (115). pun intended) in several specialized areas This follows from his theoretical premise: of financial engineering. Gorton begins if bubbles are caused by the contagion of with the basic building block, the subprime Lo: Reading About the Financial Crisis 157 mortgage,9 describing each of the layers of a view dominated the market. As a device for tall layer cake that we call securitized debt: aggregating information, the market was how those subprime mortgages were used very slow to come up with an answer in this to create mortgage-backed securities, how case. those securities were used to create CDOs, When the answer came to the market, why those obligations were bought by inves- it came suddenly. Structured investment tors, who those investors were, and why their vehicles and related conduits, which held specific identities were important. a sixth of the AAA CDO tranches,10 sim- What Gorton describes is a machine dedi- ply stopped rolling over their short-term cated to reducing transparency. Even today, debt. This wasn’t due to overexposure in it’s still striking how the available statistics in the subprime market: Gorton estimates his account dwindle as one gets to the upper that only two percent of structured invest- layers of the cake. There are estimates, ment vehicle holdings were subprime. guesstimates, important numbers with one Rather, as Gorton states, “investors could significant figure or less, and admissions of not penetrate the portfolios far enough to complete ignorance. Even the term “sub- make the determination. There was asym- prime” represents a reduction of transpar- metric information” (125). At each step in ency—Gorton details at some length the the chain, one side knew significantly more heterogeneity of the underlying mortgages than the other about the underlying struc- in this category, a term that wasn’t part of the ture of the securities involved. At the top financial industry’s patois until recently. layer of the cake, an investor might know With this description in hand, Gorton absolutely nothing about the hundreds walks us through the panic of 2007. It begins of thousands of mortgages several layers with the popping of the housing bubble in below the derivative being traded—and in 2006: house prices flattened and then began normal situations, this does not matter. In to decline. Refinancing a mortgage became a crisis, however, it clearly does. The ratio- impossible and mortgage delinquency rates nal investor will want to avoid risk; but as rose. Up to this point, this account parallels Gorton analogizes, the riskier mortgages Shiller’s basic bubble story. Here, however, in mortgage-backed securities had been Gorton claims the lack of common knowl- intermingled like salmonella-tainted frost- edge and the opaqueness of the structures ing among a very small batch of cakes that of the mortgage-backed securities delayed have been randomly mixed with all the the unraveling of the bubble. No one knew other cakes in the factory and then shipped what was going to happen—or rather, many to bakeries throughout the country.11 To people thought they knew, but no single continue Gorton’s analogy, the collapse of the structured investment vehicle market, and the consequent stall in the repurchase 9 The term “subprime” refers to the credit quality of the mortgage borrower as determined by various con- sumer credit-rating bureaus such as FICO, Equifax, and Experian. The highest-quality borrowers are referred to as 10 The term “AAA” refers to the bond rating of the “prime,” hence the term “prime rate” refers to the inter- CDO, which is the highest-quality rating offered by the est rate charged on loans to such low-default-risk individu- various rating agencies. als. Accordingly, “subprime” borrowers have lower credit 11 Gorton actually uses the analogy of E. coli-tainted scores and are more likely to default than prime borrow- beef in millions of pounds of perfectly good hamburger. ers. Historically, this group was defined as borrowers with I’ve exercised poetic license here by changing the refer- FICO scores below 640, although this has varied over time ence to tainted frosting to maintain consistency with my and circumstances, making it harder to determine what layer-cake analogy, but I believe the thrust of his argument “subprime” really means. is preserved. 158 Journal of Economic Literature, Vol. L (March 2012)

(repo) market, represented the market a market with little to no liquidity, exacer- recalling the contaminated cakes. bated the crisis. Certainly there was a sub- Here the story becomes more familiar stantial premium between mark-to-market to students of financial crises. Dislocation values and those calculated by actuarial in the repo market was the first stage of a methods. These lowered asset prices then much broader liquidity crunch.12 Short- had a feedback effect on further financing, term lending rates between banks rose dra- since the assets now had much less value as matically, almost overnight, in August 2007, collateral, creating a vicious circle. as banks became more uncertain about Gorton strongly disagrees with the which of their counterparties might be “originate-to-distribute” explanation of holding the cakes with tainted frosting and the crisis. This term, which became com- possibily shut down by food inspectors, i.e., mon in the summer of 2008, contrasts the which banks might be insolvent because of previous behavior of financial institutions, declines in the market value of their assets. which retained the loans and mortgages Fears of insolvency will naturally reduce they approved, i.e., “originate-to-hold,” to interbank lending, and this so-called “run the relatively new behavior of creating and on repo” (Gorton’s term) caused temporary packaging loans as products for further sale. disruptions in the price discovery system The originate-to-distribute explanation of short-term debt markets, an important places the blame on the misaligned incen- source of funding for many financial insti- tives of the underwriters, who believed tutions. In retrospect, the events in August they had little exposure to risk; on the rat- 2007 were just a warm-up act for the main ing agencies, which didn’t properly repre- event that occurred in September 2008 sent risk to investors; and on a decline in when Lehman failed, triggering a much lending standards, which allowed increas- more severe run on repo in its aftermath. ingly poor loans to be made. Here Gorton Gorton believes that the regulatory insis- becomes much less convincing, especially tence of mark-to-market pricing,13 even in in light of later information, and he argues as if proponents of the originate-to-dis- tribute explanation are directly attack- 12 The term “repo” is short for “repurchase agreement,” ing the general process of securitization a form of short-term borrowing used by most banks, bro- itself (which may have been the case at kerage firms, money market funds, and other financial the Jackson Hole conference). But there is institutions. In a typical repo transaction, one party sells a security to another party, and agrees to buy it back at a little in Gorton’s account—or for that mat- later date for a slightly higher price. The seller (borrower) ter, the recent historical record—to suggest receives cash today for the security, which may be viewed that the originate-to-distribute explanation as a loan, and the repurchase of the same security from the buyer (lender) at the later date may be viewed as the is excluded by the asymmetric information borrower repaying the lender the principal plus accrued hypothesis. Simply because many lenders interest. went under after the fact doesn’t mean that 13 “Mark-to-market pricing” is the practice of updat- ing the value of a financial asset to reflect the most recent their incentives were necessarily aligned market transaction price. For illiquid assets that don’t correctly beforehand. However, there is trade actively, marking such assets to market can be quite some anecdotal evidence to suggest that a challenging, particularly if the only transactions that have occurred are “firesales” in which certain investors are des- number of the most troubled financial insti- perate to rid themselves of such assets and sell them at tutions ran into difficulties in 2007–08 pre- substantial losses. This has the effect of causing all others cisely because they did not distribute all of who hold similar assets to recognize similar losses when they are forced to mark such assets to market, even if they the securitized debt they created, but kept have no intention of selling these assets. a significantportion ­ on their own balance Lo: Reading About the Financial Crisis 159 sheets instead.14 Perhaps with the benefit of blindness. In their view, the worst offenses more hindsight and data collection, we can took place at the first link of the chain, among get to the bottom of this debate in the near the subprime lenders who took advantage of future. borrower ignorance. Later links in the chain With asymmetric information in the air, had little incentive to investigate, and greater one might have expected Akerlof and Shiller’s incentives to overlook or spin away flaws in Animal Spirits: How Human Psychology earlier links. Drives the Economy, and Why It Matters for These are serious allegations, and while Global Capitalism, released in January 2009, there is no doubt that certain lenders did to have touched on the topic, especially since take advantage of certain borrowers, some Akerlof’s classic 1970 paper, “The Market for empirical support would have been par- ‘Lemons’,” launched this entire literature. ticularly welcome at this point, especially Instead, Animal Spirits, which Akerlof and because the reverse also occurred. During Shiller began writing in 2003, attempts to the frothiest period of the housing market, rehabilitate John Maynard Keynes’s concept stories abounded of homeowners flipping of “animal spirits” into a broad interpretive properties after a year or two, generating framework for studying less quantitative eco- leveraged returns that would make a hedge- nomic phenomena, among them confidence, fund manager jealous. Loose lending stan- fairness, corruption, the money illusion, and dards also benefited first-time homebuyers stories, i.e., the power of narrative to shape who couldn’t otherwise afford to purchase, events. Like Shiller’s The Subprime Crisis, and many of these households haven’t this is also meant for the advanced general defaulted and are presumably better off. reader, although earlier drafts were used in Moreover, even among the households Shiller’s course on behavioral economics at who have defaulted, while many are cer- Yale. As a result, the book is variegated, but tainly worse off, there are also those who sometimes unfocused. While the insertion can afford to pay their mortgage payments of material pertaining to the economic crisis but have chosen to “strategically default” isn’t an afterthought, in some places, it feels because it’s simply more profitable to do like a ninety-degree turn away from the main so. Are we certain that predatory lending thrust of their argument. was more rampant than predatory borrow- Akerlof and Shiller clearly hold to the ing, and that the cumulative benefits to all originate-to-distribute theory. Tellingly, homeowners are less than the cumulative they describe the run-up to the financial cri- costs? I’m not advocating either side of this sis in their chapter on corruption and bad debate—in fact, it’s difficult to formulate a faith in the markets. Where Gorton sees sensible prior as to which is more likely— opaqueness dictated by the structure of the but I believe this is a sufficiently important securities in question, Akerlof and Shiller issue to warrant gathering additional facts to see concealment, deception, and willful support a particular conclusion. In the end, Akerlof and Shiller believe, there was “an economic equilibrium that encompassed the whole chain” (37), where 14 These were presumably the “troubled assets” that the government’s $700 billion Troubled Asset Relief Program no one had any incentive to rock the boat— (TARP) were meant to relieve. For example, on October until housing prices began to drop. As with 28, 2008, Bank of America, BNY Mellon, Citigroup, Shiller’s earlier book, their policy recom- Goldman Sachs, JPMorgan Chase, Morgan Stanley, State Street, and Wells Fargo received a total of $115 billion mendations for the financial crisis appear under the TARP program (see GAO 2009). almost naively optimistic with the passage 160 Journal of Economic Literature, Vol. L (March 2012) of time. They suggest two stimulus targets. position of the American economy—the First, the proper fiscal and monetary stimu- largest engine of growth in human history— lus needed to bring the American economy didn’t render it immune to basic forms of back to full employment. The proper target, financial calamity. Nor, more disappointingly, they believed, would be easy to administer: did the expertise of its financial professionals “The Federal Reserve, the Congress, and or the strength of its financial institutions. the Council of Economic Advisers are all Nor did the forces of globalization or inno- experienced in making such predictions” vation prevent the financial crisis—in fact, (89). Second, they propose a target for the they may have provided it with new channels proper amount of credit needed to keep the through which to propagate. economy at full employment. In retrospect, To respond to future crises, Reinhart and this—the more speculative of their propos- Rogoff suggest the further development als—is the one that has been most fully real- of informational “early warning” systems ized. In January 2009, it wasn’t yet clear that and more detailed monitoring of national the political economy of the financial crisis financial data, perhaps through a new inter- would favor the rebuilding of the credit mar- national financial institution, similar to kets over the pursuit of full employment. the development of standardized national By the fall of 2009, the outlines of the account reporting after World War II. Their early stages of the financial crisis were data appendices and analytics pave the clear, although the exact causation (or the way for such an initiative. They also warn blame) remained a point of vigorous con- about the recurrence of “this time is differ- tention. With the September publication of ent” syndrome, something that observers This Time Is Different: Eight Centuries of since Charles Kindleberger (if not Charles Financial Folly, Carmen M. Reinhart and Mackay) have warned against. Moreover, Kenneth Rogoff provided invaluable histori- they preemptively dismiss future state- cal data and context for understanding the ments of “this time is different” based on the crisis. Among all the books reviewed in this Lucas critique, Robert E. Lucas’s famous article, theirs is the most richly researched macroeconomic dictum against historical and empirically based, with almost 100 pages prediction because simple linear extrapo- of data appendices. If all authors of crisis lations of the past don’t take into account books were required to support their claims the sophistication of rational expectations. with hard data, as Reinhart and Rogoff do Reinhart and Rogoff argue that because the most of the time, readers would be consider- historical record shows that some nations ably better off and our collective intelligence have “graduated” from perennial financial would be far greater. instability to financial maturity, there is This vast compendium of financial crises reason to hope that improved forms of self- showed that the 2007 subprime meltdown monitoring and institutional advances can was neither unprecedented nor extraordi- keep certain types of financial crises from nary when compared to the historical record. happening, despite the implication of the Reinhart and Rogoff briefly document the Lucas critique that such predictions are “this time is different” thinking among inves- futile. tors, academics, and policymakers. They link An unusual perspective of the finan- the rise of the housing bubble in particular cial crisis appeared in the United States in and the rise of the financial industry in gen- November 2009 from the Australian econo- eral to the large increase in capital inflows to mist Ross Garnaut in a book coauthored with the United States. The great size and central journalist David Llewellyn-Smith. Written Lo: Reading About the Financial Crisis 161 originally for an Australian audience, The America, Free Markets, and the Sinking Great Crash of 2008 gives a somewhat jour- of the World Economy in January 2010. nalistic account of the events of the crisis Expanded in part from two earlier arti- through the summer of 2009, but one in cles in Vanity Fair magazine, this book is which the authors describe the many firms Stiglitz’s jeremiad as well as his explanation and personalities involved in the crisis by of the financial crisis. He begins his story in name and by anecdote, with obvious rel- 2000 with the bursting of the Internet bub- ish. This was a necessity for them because ble. In his view, the housing bubble and the most of the primary actors were unfamiliar subprime mortgage crisis cannot truly be to Australians, but the authors’ specificity separated from the earlier dot.com boom contrasts starkly with the greater abstraction and bust, but rather represent symptoms of and distance of most American academics in a deeper systemic crisis among our policy- their formal accounts of the crisis (though makers and institutions. Instead of address- not necessarily in op-ed pieces and less for- ing the root problems underlying the mal articles). earlier bubble, a dismantling of the regula- Australia’s position as an English-speaking tory apparatus, regulatory capture, and an advanced economy, yet one still peripheral explosion in untested financial innovations to the core global economies of the North, set the stage for the next crisis. Stiglitz fears closely informs Garnaut and Llewellyn- that the pattern will repeat: that govern- Smith’s account. Like Reinhart and Rogoff, ment half-measures—or actively bad policy they immediately tie the housing bubble decisions—in response to the subprime to increased capital flows, especially those crisis will set up the conditions for an even from China. They largely agree with the greater crisis. originate-to-distribute hypothesis, and they In many ways, Stiglitz’s polemical tone believe that regulatory capture and a cul- belies the mainstream nature of his expla- ture of greed aided and abetted the develop- nation. It is a variation of the originate-to- ment of the crisis. Where The Great Crash distribute theory, made rhetorically sharper of 2008 is most valuable for an American with the revelations of venality and outright reader, however, is through its descriptions criminality among intermediate links in the of parallel innovations in the Australian subprime chain. The largest misaligned financial industry and in Australian political incentives, however, in Stiglitz’s view, were economy. Here, the authors postulate a con- found among the “too big to fail” financial tagion of ideas through the English-speaking institutions, which Stiglitz argues took exces- world—the “Anglosphere”—causing econo- sive risk because they were too big to fail; mies such as Australia, the United States, that is, they were so large and essential to the and Great Britain to experience similar con- functioning of the financial systems of the sequences, e.g., securitization, the shadow American (and global) economy that their banking system, housing price booms, and a managers behaved as though they would be rise in executive remuneration, rather than bailed out despite making poor decisions. such developments arising naturally and While such vitriol accurately channels a independently in response to local economic significant portion of the public’s reaction conditions. to the crisis, there’s not much new in the If American academics had previously way of data or economic analysis. It seems been circumspect in their accounts of the eminently plausible that “too big to fail” financial crisis, the gloves came off with and implicit government guarantees could the publication of Joseph Stiglitz’s Freefall: affect corporate strategy to some extent, but 162 Journal of Economic Literature, Vol. L (March 2012) quantifying the impact seems less obvious. or tedious, depending on their prior beliefs, In particular, to determine the effect that but at least he’s explicit about his convic- government bailouts might have on cor- tions. However, he sometimes loses clarity porate risk-taking, it matters a great deal with respect to his assertions of bad faith whether the bailouts are intended to res- among principal players during the crisis. cue bondholders, equityholders, or both. Stiglitz was certainly in a position to hear This is where new economic analysis could privileged information about private policy have added real value. For example, given discussions—he credits the Obama admin- the empirical evidence in Fahlenbrach and istration’s economic team with sharing their Stulz (2011) and Murphy (2011) that CEOs’ perspectives with him, despite his often pro- incentives seem highly aligned with share- found disagreement with them. Still, many holders, do implicit government guarantees readers will have their curiosity piqued about cause shareholders to take on too much risk, the circumstances behind some of these dis- in which case we need to focus on reduc- closures; unfortunately, they may not get ing the sizes of large financial institutions, much satisfaction until Stiglitz publishes his as Johnson and Kwak (2010) propose (see memoirs. below)? Or is this a reflection of deeper con- Several attempts to place the financial cri- cerns regarding corporate governance and sis into a larger framework emerged in the whether CEOs should be maximizing stake- spring of 2010. First published among these holder wealth instead of shareholder wealth? attempts was Simon Johnson and James Maximizing shareholder wealth is currently Kwak’s Thirteen Bankers: The Wall Street the focus of most U.S. CEOs and their exec- Takeover and the Next Financial Meltdown, utive compensation plans. However, some of released in March. Johnson and Kwak frame the rhetoric in this debate suggests an unspo- the financial crisis as another swing of the ken desire for more inclusive policies, which pendulum of the American political economy would be quite a departure from the corpo- and its financial institutions. In their view, rate governance structures of most Anglo- the concentration of power by financial elites Saxon and common-law countries such as in the American system—whom Johnson the United States and United Kingdom.15 A and Kwak characterize as “oligarchs”— more detailed fact-based analysis would have leads to governmental financial institutions been particularly valuable in this instance. with strong private cross-interests and weak The proper solution according to Stiglitz regulatory oversight, producing a financial is a wholesale reformation of the American environment prone to recurrent crises. On financial system on a scale not seen since the the other hand, when the government has Great Depression. Much of Freefall laments played an aggressively hostile role against the missed opportunity for such a reforma- the concentration of financial power (as dur- tion. Here, however, Stiglitz’s account of the ing the Andrew Jackson administration), its political economy behind the stimulus pack- actions have resulted in a fragmented, weak, ages and bailouts becomes much too vague. and vulnerable financial system. In their It may fall to the political scientists rather opinion, the most successful course has been than the economists to give us the complete the middle course, taken by Franklin Delano story of what happened. Readers will likely Roosevelt and his advisors in the early 1930s, find Stiglitz’s moral fervor either refreshing which led to a half-century of strong finance without major financial crises. Johnson and Kwak mark the turning point 15 See Allen and Gale (2002). away from the older, safer, “boring” banking Lo: Reading About the Financial Crisis 163 regime to today’s bigger, “exciting,” more capitalization limits on the size of financial crisis-prone regime with the election of institutions. This, they believe, would cause Ronald Reagan. Financial innovation and a these problems to unwind, piece by piece, wave of financial deregulation, made pos- initially by decreasing the threat of “too big sible in the new political climate, reinforced to fail” banks. As the financial sector becomes each other, leading to increased profits and less “exciting” under these new rules, the a rapid expansion of the financial sector. incentives for pursuing risky behavior will Banks also grew under deregulation—here, diminish. Eventually, this virtuous cycle ends Johnson and Kwak’s account doesn’t fully with changes to the institutional culture of explain their reasoning behind the resulting the financial sector, returning to its earlier concentration, although the facts are hardly norms. in dispute. By the 1990s, the American finan- Nouriel Roubini and Stephen Mihm’s cial sector was able to exert further influence Crisis Economics: A Crash Course in the on the political process in a number of ways: Future of Finance was published in May lobbying, campaign contributions, and pro- 2010, shortly after Johnson and Kwak’s viding official Washington with a cadre of account. Roubini by this point had achieved financial professionals who had internalized a certain measure of notoriety outside of much of the new, “exciting” ethos of Wall academia as the prophetic “Doctor Doom” Street. of the financial media; his early warnings that According to Johnson and Kwak, this the housing bubble could lead to systemic renewed regulatory capture by America’s financial collapse led Roubini to become one new masters of the universe set the stage for of the few financial economists nicknamed the boom and bust cycles of the late 1990s after a comic book super-villain (a nickname and onward. Moves toward greater finan- in fact popularized by his coauthor in a New cial regulation were actively driven back York Times profile). by the so-called oligarchs—in one of their Roubini and Mihm give a crisp exposition examples, Brooksley Born, then head of the of the underlying mechanisms of the crisis. Commodity Futures Trading Commission, In Roubini’s view, the financial crisis wasn’t was blocked from issuing a concept paper a rare, unpredictable “black swan” event, on new derivatives regulation by the “thir- but rather a wholly predictable and under- teen bankers” of Johnson and Kwak’s title. standable “white swan.” Comparing it to Financial institutions became “too big to recent crises in developing economies and fail,” taking additional risk with the implicit historical crises in developed ones, Roubini (and possibly not-so-implicit) knowledge and Mihm present a short primer on conta- that should the worst happen, the United gion, government intervention, and lender States government would likely rescue them of last resort theory, using them to set up from their financial folly. Once again, this the heart of the book: its policy prescrip- glosses over the critical question of whether tions. They propose a two-tier approach it is the bondholders or equityholders who of short-term patches and long-term fixes. get bailed out, and where more careful eco- Most of the short-term proposals have to nomic analysis is needed. do with reforms to the financial industry, Johnson and Kwak diagnose a systemic including increased transparency, changes problem of consolidation and influence, not to compensation structure, and increased merely of a small number of large financial regulation and monitoring of the securitiza- institutions, but of an entire financial sub- tion process, the ratings agencies, and capi- culture. Their solution is quite simple: hard tal reserve requirements. 164 Journal of Economic Literature, Vol. L (March 2012)

In contrast, Crisis Economics prescribes enterprises like Fannie Mae and Freddie much stronger medicine for the long term. Mac.16 Political pressure caused these pro- Bubbles should be actively monitored and grams to extend easier credit to less suit- proactively defused by monetary authori- able applicants and private firms followed ties. Lobbying and the “revolving door” the government’s lead, culminating in the between finance and government should housing bubble of 2006 and its aftermath. be severely restricted to prevent regula- Each link in Rajan’s causal chain is a com- tory capture. To prevent what Roubini pelling idea worthy of further consideration, and Mihm call “regulatory arbitrage” by characteristic of Rajan’s method of argument. banks—what lawyers often refer to as But does the chain truly hold? As with the “jurisdiction shopping”—a single, uni- well-known property of probabilities, even if fied national authority should regulate and each link has a high likelihood of being the monitor financial firms, and strong inter- “correct” causal relationship, a sufficiently national coordination is needed to prevent long chain of independent events may still be banks from engaging in regulatory arbitrage extremely unlikely to occur. Of course, Rajan on a global scale. “Too big to fail” institu- realizes the solution to this conundrum, and tions should be broken up, whether under uses multiple chains of reasoning to create a antitrust laws, or under new legislation that stronger cable of analysis. He considers other defines such institutions as a threat to the “fault lines” such as the global capital imbal- financial system. Finally, the separation ance, the traditionally weak social safety net between investment banking and com- in the United States, and the separation of mercial banking, which had existed under business norms in the financial sector from the Glass–Steagall Act, should return in an those in the real economy, which Rajan wit- even stronger form. Given their premises, nessed firsthand. these suggestions make sense, but Roubini He proposes a three-pronged attack against and Mihm avoid the difficult political ques- the conditions that made the financial crisis tions of implementation. possible. First, he suggests a set of strong May 2010 was also the month in which social policies to lower inequality in the Raghuram G. Rajan’s Fault Lines: How United States, among them increasing educa- Hidden Fractures Still Threaten the World tional access, universalizing health care, and Economy was released. Rajan’s arguments decreasing the structural risks to personal on the causes of the financial crisis are labor mobility. Second, he recommends that multiple and complicated, but they are all international multilateral institutions develop variations on the same theme: systematic relationships with the constituencies of their economic inequalities, within the United component nations, rather than functioning States and around the world, have created merely as a ­top-down council of ministers. deep financial “fault lines” that have made crises more likely to happen than in the past. Rajan begins with the United States, 16 “Fannie Mae” is the nickname of the Federal National where there has been a long-term trend, Mortgage Association, a government-sponsored enterprise he argues, of unequal access to higher edu- created by Congress in 1938 to “support liquidity, stability, and affordability in the secondary mortgage market, where cation creating growing income inequal- existing mortgage-related assets are purchased and sold.” ity. To address the political effects of this “Freddie Mac” refers to the Federal Home Loan Mortgage inequality, leaders from both parties have Corporation, another government-sponsored enterprise created by Congress in 1970 with a charter virtually iden- pursued policies to broaden home owner- tical to Fannie Mae’s. See http://www.fanniemae.com and ship, e.g., through government-sponsored http://www.freddiemac.com for further details. Lo: Reading About the Financial Crisis 165

More democratic input and greater transpar- ­professionals who take an active role in their ency should, in Rajan’s opinion, improve the firms’ operation. quality of the decision-making process among Rajan believes the discipline of the mar- the multilateral institutions on the one hand, ket will not be enough, however. Other and make their policy recommendations governmental regulation must simultane- more palatable to their member nations on ously become more comprehensive and the other. This would allow greater interna- less sensitive to political over- or under- tional and domestic coordination regarding reaction. In contrast to Johnson and Kwak, the global capital imbalance (and other press- Rajan believes that fixed limits on bank size ing international issues). or activity are too crude and easily evaded, Rajan proposes a complex set of carrots creating a new set of misaligned incentives and sticks to defuse the bad incentives that for financial institutions. Rajan sees an active have accumulated in the American financial role for bank regulators and supervisors. sector. He believes risk was systematically Public transparency and bank supervision underpriced in large part because of the would serve as a check to excessive risk-tak- financial sector’s expectations of govern- ing by corporate governance. Like Roubini ment intervention. Removing the implicit and Mihm, Rajan favors a modern version promise of intervention and the explicit of the Glass–Steagall Act and other forms of promise of subsidies would eliminate this asset segregation: this would diminish risk distortion. The government should espe- and eliminate a potential channel for a panic. cially remove itself from the secondary Rajan admits that this would also increase a mortgage market as soon as possible, and bank’s borrowing costs, but he believes the reduce its role in the primary mortgage tradeoff might be worthwhile. He also favors market. Even the role of deposit insurance, a prohibition against proprietary trading, not usually thought of as one of the center- for its increased risks, but because of the pieces of American bank regulation, should potential abuse of asymmetric information be reconsidered according to him. by the banks. Meanwhile, financial corporate gover- In May 2010, a third crisis book was pub- nance must reduce the amount of risk taken lished, authored by fifteen financial econ- on by traders and companies. Instead of omists including Rajan and Shiller: The immediate compensation for investment Squam Lake Report: Fixing the Financial strategies that might have hidden tail risk, System. This bipartisan group originally Rajan proposes that a significant fraction of met in the fall of 2008 at Squam Lake, the bonuses generated by finance workers New Hampshire, to discuss the long-term and management be held in escrow subject reform of the world’s capital markets. This to later performance. This would have the report cuts across a representative (but not effect of extending the time horizon used to necessarily complete) section of the politi- calculate profit. If the traders and managers cal and ideological spectrum; as a result, are acting rationally, this should, in theory, many passages resemble carefully worded diminish tail risk.17 At the highest ­levels, public statements released by an ecumeni- boards should choose prudent financial cal group on a controversial tragedy. This report doesn’t propose any consensus view among academic policymakers, but is more of an extended brainstorming session to 17 It’s worth noting that AIG had a broadly similar plan in place for its top executives during the run-up to the find new policy solutions for an unprece- crisis. dented crisis. 166 Journal of Economic Literature, Vol. L (March 2012)

Many of the Squam Lake group’s pro- financial institutions to use a single, strongly posals will already be familiar to readers of regulated clearinghouse.19 On other ques- this review. The group proposes that each tions, such as the problem of runs on large nation set up a systemic financial regulatory brokers due to their unsegregated asset agency run by the central bank. In terms of structure, the group cannot decide on a solu- transparency, these regulators should collect tion based on existing research. Interestingly, much broader standardized data on financial the group attempts to walk through how institutions, and this data should become specific failures during the financial crisis, public after an interval. Capital require- such as the collapse of Bear Stearns, would ments should increase with the size, risk, and have played out had their recommendations liquidity of assets. Governments shouldn’t been in place. Candidly enough, they see a impose limits on executive compensation, modest improvement at the firm level, and a but they should impose rules that financial reduced cost to the taxpayer, but they make institutions withhold full compensation for no claims that the financial crisis itself would a fixed time period. Simply put, the govern- have been averted. ment should be used to universalize regula- Finally, in April 2011, Acharya, tion, but institutions should internalize the Richardson, van Nieuwerburgh, and White’s cost of their own failures. Guaranteed to Fail: Fannie Mae, Freddie Other proposals of the Squam Lake group Mac, and the Debacle of Mortgage Finance are more novel. To maintain bank solvency, was published. This is a key contribution to the group proposes that the government pro- one of the most vexing problems from the mote banks to issue a long-term convertible epicenter of the crisis: the future of Fannie bond that converts to equity at very specific Mae and Freddie Mac. The authors trace triggers during a crisis. In this way, instead the origin of their problems to Fannie Mae’s of ad hoc government recapitalization during flawed privatization during the Johnson a banking crisis, the costs of recapitalization administration (made largely for account- will be put on the bank’s investors. To expe- ing reasons). Fannie Mae, and later Freddie dite a recovery, the group recommends that Mac, had the ability to participate as a pub- financial institutions maintain “living wills” licly traded company on the one hand, but to help regulators restructure them quickly maintained the privileges granted by its fed- in worst-case scenarios. eral charter on the other. Financial markets For problems specific to the recent cri- believed that Fannie Mae and Freddie Mac sis, however, the Squam Lake group offers had implicit guarantees on their holdings fewer panaceas. The problem of systemic from the federal government, apparently risk in credit default swaps (CDSs) is a dif- with good reason. Following the deregulation ficult one,18 but the Squam Lake Report can of the mortgage industry during the Reagan only suggest that the government encourage administration, investors naturally preferred to invest in them rather than in truly private

18 A “credit default swap” is an agreement between two parties in which one party agrees to pay the other party a prespecified amount of money in the event of a default on a 19 A “clearinghouse” is a legal entity that serves as an third party’s bond. Essentially a type of insurance contract, intermediary between two counterparties so that if either CDSs were used to provide credit protection for various one defaults on its obligation, the clearinghouse will fulfill mortgage-backed securities like collateralized debt obliga- that obligation. The presence of a clearinghouse greatly tions (CDOs), which was particularly popular among the reduces “counterparty risk” and enhances the liquidity of most conservative investors in CDOs such as money mar- the contracts traded, which is especially relevant for credit ket funds. default swaps. Lo: Reading About the Financial Crisis 167 mortgage companies. Bipartisan policy goals name. The authors attempt to split the dif- made the enterprises politically untouch- ference by proposing a private/public part- able, even while the evidence of their mis- nership for the mortgage guarantee business management grew. In effect, as the authors only, the lower levels of the mortgage indus- of Guaranteed to Fail point out, Fannie Mae try becoming fully private (although highly and Freddie Mac were run as the world’s regulated). Finally, the authors believe the largest hedge funds, and badly at that. root cause of the mortgage finance debacle, How to unwind this trillion-dollar prob- and by extension, the entire global financial lem? If much smaller institutions were crisis from 2007—the American “addic- already “too big to fail,” Fannie Mae and tion” to homeownership—should be treated Freddie Mac must represent a class unto posthaste. themselves in terms of sheer size and the dollar-value of their implicit guarantees 3. Journalistic Accounts (estimated to be between $20 to $70 billion in present-value terms according to Lucas While often overlooked by academic read- and McDonald (2011), depending on the ers, the journalistic accounts of the financial assumptions used). Drawing on the example crisis are complementary in many ways to of the savings and loan crisis in the United their academic counterparts. If we return to States in the late 1980s and early 1990s, the analogy of the financial crisis as a major the authors propose that the government war, then in the same way that the academic establish a “resolution trust corporation” to writers acted as the strategists, diplomats, manage the slow liquidation of Fannie Mae and gadflies of the crisis, the financial report- and Freddie Mac assets—slow, so as not to ers were the war correspondents. These destabilize the remaining mortgage-backed journalists documented the campaigns, securities market. As the housing market battles, and exceptional acts of courage and improves, eventually the process can be cowardice among individuals and battalions. accelerated. A similar procedure can take Moreover, they describe elements of the cri- place with those Fannie Mae and Freddie sis that, as a scientific discipline, economics Mac assets now held by the Federal Reserve. has difficulty capturing: the role of motives, The other half of this trillion-dollar prob- psychology, personality, and strong emotion. lem, the authors agree, is to never let a simi- We have seen how Akerlof, Shiller, Stiglitz, lar situation arise again. The authors believe Roubini, and others have touched upon the that the problem is inherent to government- role of greed, fear, and anger in the hous- sponsored enterprises with laudable social ing bubble, the financial crisis, and its policy goals, especially in the housing market, responses. By breaking down the macro- and they point to similar but smaller fail- events of the crisis into many different per- ures in Germany and Spain. They reject full sonal stories, these accounts are actually nationalization due to its enormous liabil- literary attempts to make sense of the crisis ity—Johnson had partially privatized Fannie from a micro-foundational level. It’s difficult Mae for much less—and for the likely politi- to speak of rational behavior in the aggregate cal capture of its management. In a similar when major economic decisions are made spirit, they are agnostic about full privati- by an unrepresentative handful of people. zation, foreseeing that the largest private While journalistic accounts of the crisis have mortgage originators would simply induce the flaws of their genre—they are necessar- enough regulatory capture to become gov- ily subjective, often moralistic, and they may ernment-sponsored enterprises in all but attempt to shape a narrative beyond what 168 Journal of Economic Literature, Vol. L (March 2012) the facts will strictly bear—the accounts of best players in the world and notorious for economists and policymakers may have their his presence at tournaments and absence at own form of biases. Bear Stearns during the crisis. Cohan makes William Cohan’s House of Cards: A Tale of the intriguing implication that the cognitive Hubris and Wretched Excess on Wall Street skills involved in playing world-class bridge was the first major journalistic account out of might distort the skills involved in making the gates, published in March 2009, almost a financial decisions at their highest levels. year to the day after the fall of Bear Stearns, The spring of 2009 also saw the release of which it recounts in great detail. Cohan, a for- Gillian Tett’s Fool’s Gold: The Inside Story mer finance professional turned investigative of J.P. Morgan and How Wall St. Greed reporter, documents the harrowing final days Corrupted Its Bold Dream and Created a of the firm, and this morbidly fascinating tale Financial Catastrophe in May. Tett, the for- reminds us that economics has few answers mer global markets editor and current U.S. to liquidity crises, thin markets, and other sit- managing editor of the Financial Times, uations where the price discovery mechanism reconstructs the early history of the develop- fails to perform. As the financial analyst A. ment of the credit derivatives market, which Gary Shilling (1993, 236) put it, “Markets can played a key role in the subprime crisis. If remain irrational a lot longer than you and I Cohan’s account was the view from Bear can remain solvent.” In those circumstances, Stearns, Tett’s account is very much the view economic actors will necessarily fall back from J.P. Morgan (now formally JPMorgan onto procedures which, almost by definition, Chase). Tett traces the origin of credit deriv- will produce suboptimal outcomes, e.g., the atives to an initiative of Morgan’s swaps team fate of Bear Stearns. Cohan is also very strong at a Palm Beach resort hotel in 1994 (Tett in his portrayal of economic decision-making mentions, in passing, earlier, less success- under stress and decision-making by small ful innovations in default-risk derivatives at groups, two areas which have recently begun Merrill Lynch and Bankers Trust). In a heady to receive more scholarly attention.20 intellectual atmosphere of Friedrich von Bear Stearns was the first of the major Hayek and Eugene Fama, this young team American banking firms to fall during the sought to create a successful derivative prod- financial crisis, and it’s commonly believed uct that would protect against default risk, that it was also the weakest in terms of over- something all lending institutions have to sight, incorrectly aligned incentives, and deal with. This product would combine the organizational culture to handle the crisis. virtuous motive of helping to expand capital While this might be an example of fallacious into the greater economy with the self-inter- post hoc reasoning, Cohan presents a case ested motive of helping to expand Morgan’s that Bear Stearns’s dysfunctional manage- share of the derivatives market. Banks for ment and aggressive corporate culture— the first time would be able to make loans even by the standards of Wall Street—made without carrying the associated credit risks of it particularly vulnerable. Unusually, several those loans, which would be transferred to figures in Bear Stearns’s management were the buyers of the derivative. tournament-caliber bridge players, including At the cutting edge of financial engineer- its last chairman, Jimmy Cayne, one of the ing for its time, these new derivatives were “technically sweet,” to borrow J. Robert Oppenheimer’s postwar description of the 20 For the former, see Kowalski-Trakofler, Vaught, and atomic bomb. As a product, their design Scharf (2003); for the latter, see Woolley et al. (2010). principles were similar to other consumer Lo: Reading About the Financial Crisis 169 success stories: they were easy for the inves- critical events of September 2008 (speaking tor to buy and sell; they could use a wide generally for all these books, one has to be variety of starting materials in their bundled impressed by their speed from crisis to print). loans through the securitization process; Sorkin’s account is perhaps the best single and they conformed to (or, more strictly descriptive narrative of the top levels of the speaking, evaded) government and industry 2008 phase of the crisis that we have, and standards. Morgan’s first BISTROs—broad as memories fade, self-justifications harden, index secured trust offerings—were issued and participants leave the scene, it’s likely to in December 1997, and the product quickly remain the best anecdotal summary of these became a hot item. events. As one of the New York Times report- Tett’s later story is primarily one of corpo- ers covering the crisis, Sorkin had an unusual rate culture and intellectual contagion. Tett, amount of access to participants and observ- who began her career as a social anthropolo- ers both during and after the events of 2008. gist, has a fine eye for the group dynamics Too Big to Fail must represent the distilla- behind these processes. Financial firms tion of hundreds, if not thousands of hours of throughout the United States and Europe off-the-record interviews, tapes, videos, and quickly adopted the basic forms of Morgan’s more conventional sources. innovations, resulting in a Cambrian explo- However, Sorkin’s wide scope and multiple sion of new derivatives—to use Tett’s termi- viewpoints of the crisis represent a tradeoff nology, derivatives “perverted” from their with respect to deeper analysis. His book is original form and intent. At the same time, probably best read in conjunction with other however, Morgan kept its original wor- accounts as a reference point. For example, ries about “super-senior” risk and the lack it throws former Treasury Secretary Henry of provenance within mortgage bundles to M. Paulson’s memoir (see below) into an itself. The merger of Morgan with Chase entirely different light when Sorkin reveals Manhattan introduced a new, risk-seeking that Paulson’s deputy would routinely warn element to the culture of the new JPMorgan visitors that Paulson had no “social emotional Chase, driving away most of J.P. Morgan’s quotient” at all. Too Big to Fail will also likely earlier talent, and paradoxically spreading be used for later memoirists to craft their new financial innovations to much less risk- own accounts of events, an influence that averse corporate cultures. A later merger future historians of the crisis should keep in with Bank One introduced new management mind. Along those historical lines, one wishes headed by Jamie Dimon to JPMorgan Chase, there was a convenient date- and time-stamp which consequently became concerned of the events in the page margin—or in the again about hidden risk within its derivative corner of the viewing screen—as one follows products. Tett makes the case that Dimon’s individual threads of the complicated deci- skills—including his famous insistence on a sion-making processes Sorkin recounts. In “fortress balance sheet”—allowed JPMorgan fact, Sorkin’s narrative would make an excel- Chase to survive the crisis when some of its lent front end to a multimedia database of largest competitors did not. materials pertaining to the crisis. During the autumn of 2009, New York That fall also saw the publication of a Times columnist Andrew Ross Sorkin pub- book about the other “Paulson,” Gregory lished Too Big to Fail: The Inside Story of Zuckerman’s The Greatest Trade Ever: How Wall Street and Washington Fought to The Behind-the-Scenes Story of How John Save the Financial System—and Themselves. Paulson Defied Wall Street and Made Its release came a little over a year after the Financial History, published in November 170 Journal of Economic Literature, Vol. L (March 2012)

2009. Despite their eight-hundred year his- the Global Financial System—was released tory, bubbles are still rather mysterious eco- in February 2010. At first glance, Paulson’s nomic phenomena. One deep mystery of memoir appears to be derived from his per- bubbles is their asymmetry. Why do so few sonal diary of the crisis, revised and edited investors try to take advantage of an obvi- for publication. In fact, On the Brink is an ous bubble? And why do even fewer inves- almost wholly synthetic day-by-day account tors manage to profit once a bubble bursts? of the escalating series of crises during Zuckerman, a reporter for the Wall Street Paulson’s time at Treasury, based on his pro- Journal, tells the riveting story of the largest digious memory, incomplete phone logs, single beneficiary of the collapse of the hous- and personal conversations with many of its ing bubble, a previously unknown hedge- participants after the fact (Paulson states he fund manager named John Paulson. does not use email.) Why did John Paulson succeed? Paulson’s It’s become a truism that one should read rare (but not unique) insight was to purchase memoirs by people at the center of great CDS insurance on the most risky slices of historical events with a careful eye toward mortgage bonds, the BBB tranches. These score-settling, self-justification and, more bonds would be the first to be hit in the rarely, self-blame; On the Brink would be event of default, which Paulson saw as inevi- unique if it lacked those elements. For the table in the collapse of the housing bubble. most part, however, Paulson presents him- Derivative contracts like CDSs were gen- self as a competent man dealing with events erally unpopular because they represented almost beyond his control, often mistaken “negative carry” trades, a situation in which or uncertain about the magnitude of each buyers of such contracts are subject to a impending phase of the crisis taking place steady stream of sure losses. Its payoffs are while he and the Treasury Department man- similar to playing a slot machine, constantly aged to weather the collapse of Bear Stearns, putting in coins in the hopes of an enor- Lehman Brothers, Fannie Mae, and Freddie mous but uncertain jackpot some time in the Mac, and the financial near-apocalypse of future. In a normal market, someone obses- September 2008. sively buying CDS insurance would have a In that respect, On the Brink is very much similar financial fate as someone obsessively the Treasury view of the crisis of 2008. playing the slots. Astonishingly, Paulson’s Similar memoirs from Timothy Geithner or initial purchases not only failed to run up Ben Bernanke will probably be some time in the price of the insurance contracts, but the coming. In the meantime, however, policy- sellers tried to convince him he was making minded readers will find much to think about a mistake. The information-gathering func- regarding the formal and informal constraints tion of the price discovery mechanism was on the power of the United States’s monetary clearly awry. Paulson’s uniqueness came institutions. One striking example is the pol- from his conviction, his deep pockets, and icy aversion at the time to any cost figure near his ability to get out of his position. Without a trillion dollars or higher. Paulson and his any single one of those qualities, Zuckerman colleagues believed that legislators would be implies, Paulson’s record-breaking $4 billion too hostile to a trillion-dollar estimate for the payout in 2007 would have been much less Troubled Assets Relief Program, and instead spectacular. chose $700 billion as the least-bad figure Former Secretary of the Treasury Henry that might accomplish their goals. As it hap- M. Paulson’s account of the crisis—On the pened, Paulson was still surprised at the hos- Brink: Inside the Race to Stop the Collapse of tility he received from lawmakers. Was this Lo: Reading About the Financial Crisis 171 a case of political timidity or Hayekian local Street, and the enormous energies of innova- knowledge? Overall, Paulson’s account of the tion and profit which they unleashed, embed- crisis isn’t particularly analytical, being more ded false assumptions deep into the culture akin to a boxer’s account of a fight the morn- of Wall Street, assumptions that blinded the ing after, but it provides much raw material vast majority of its participants to the possi- for subsequent analysis by others. bility that they might be mistaken. In Lewis’s Of all the financial journalists in this opinion, the financial crisis marked the pass- review, best-selling author Michael Lewis ing of a fascinating but flawed cultural era on is probably the best known. A former bond Wall Street. salesman at Salomon Brothers in the 1980s, In another coincidence of timing, finan- his memoir of his short time on Wall Street, cial journalist Roger Lowenstein’s The Liar’s Poker, has become a financial classic. End of Wall Street was published in April More recently, his book on the economics of 2010, shortly after Lewis’s elegy to the old baseball team development, Moneyball, has Wall Street appeared. Lowenstein is per- catalyzed popular interest in the use of statis- haps best known for When Genius Failed, tical innovation in professional sports—per- his account of the collapse of Long-Term haps the first time in history a bestseller has Capital Management. The End of Wall Street made statistics cool. The Big Short: Inside is a similar chronicle of the top levels of the the Doomsday Machine, published in March financial crisis, from large mortgage firms to 2010, examines the crisis from a similar per- banks to official Washington. Unlike Sorkin’s spective to Zuckerman’s, by profiling a group account, Lowenstein’s narrative presents a of people who profited from the crisis. This highly linear view of the crisis, with banks is apparently something of a coincidence: and institutions falling down like dominoes Lewis read the coverage of John Paulson in in a row. This is a legitimate approach, but it the Wall Street Journal, while Zuckerman fails to capture the sense of a tectonic shift in had read Lewis’s elegy for the old Wall Street the markets in 2007 and 2008. Lowenstein’s in Portfolio magazine which became the first later publication date, however, allows him and last sections of The Big Short. to explore the continuation of economic Only a very few contrarians, outsiders, policy in the new Obama administration, the malcontents, and naifs bet against the hous- beginnings of the new low-lending, high- ing bubble in Paulson’s manner—Lewis unemployment era that followed, and the estimates between ten and twenty people early political conflicts over governmental worldwide—and The Big Short, a rather economic stimulus. short book itself, describes a significant frac- Lowenstein views the financial crisis as a tion of them.21 This is an extraordinary level failure of the market system and postindus- of uniformity of opinion, and it’s no surprise trial capitalism, a sentiment that manages that the dissidents from the mainstream to sound surprisingly conventional in his view were, at first glance, marginal figures at hands—a measure, perhaps, of the depths best, and more often considered crackpots. of the crisis. Intriguingly, he considers the The reasons for this uniformity are complex. crisis a natural consequence of a financial Lewis believes the past successes of Wall system that, rather than extracting Marxist super-profits from society, extracted risk from its investments and dumped it on those members of society least able to handle it. 21 Two of these figures, the Deutsche Bank trader Greg Lippmann and the neurologist turned hedge fund manager The individual firm reduces its risk, but soci- Michael Burry, were also profiled in Zuckerman’s book. ety as a whole has its risk increased. There 172 Journal of Economic Literature, Vol. L (March 2012) are several economics and finance Ph.D. November also saw the publication of theses that need to be written to sort out this Bethany McLean and Joe Nocera’s All the one idea. Devils Are Here: The Hidden History of In November 2010, Greg Farrell’s book the Financial Crisis. McLean is, of course, Crash of the Titans: Greed, Hubris, the Fall best-known for her breaking reportage of of Merrill Lynch, and the Near-Collapse of the Enron scandal, and Nocera is currently Bank of America came out. Farrell, a cor- an op-ed columnist at the New York Times. respondent for the Financial Times, has Their book is an ensemble portrait of the written a strong narrative business history subprime crisis, clearly of the second (or per- of Merrill Lynch in its final months, and haps third) publishing cycle after the original the peculiar merger with Bank of America event; in fact, many books mentioned earlier that followed in late 2008. Unlike earlier in this article are acknowledged as important accounts described here, Farrell’s book lacks sources of insight. Its strengths, however, are a strong analytical focus, perhaps because in its grounding in the nuts and bolts of the by this time the basic narrative of the finan- relevant industries and government organi- cial crisis seemed like well-trodden ground. zations—most notably, in the bond rating Farrell employs a personality-driven model firms and the mortgage originators—all the regarding the behavior of firms: personalities way up to the actions of the Federal Reserve at the top create incentives (or disincentives) Board. for its employees to follow, rather than the McLean and Nocera tell a story of lead- firm following the dictates of the market. For ing personalities in representative indus- example, Merrill’s adoption of a heavy load tries responding to incentives, especially of CDOs is presented as a consequence of its to changes in the regulatory environment. chief executive Stanley O’Neal’s dismantling These changes induced a coarsening in stan- of Merrill’s earlier corporate culture rather dard business practice. Established firms than market competition or opportunities became corrupt in their pursuit of profit; per se. corrupt firms became criminal. McLean and If a rising tide lifts all boats, a perfect Nocera also tell a parallel story of regula- storm will sink even the soundest. In Farrell’s tory capture, evasion, inundation, and inef- account, once again we see how the financial fectiveness. With few exceptions, official crisis exacerbated preexisting dysfunctions Washington is excoriated for its inaction and in the management structure, oversight, and complicity in this process. Local officials at corporate governance of financial institu- the city and state level, on the other hand, are tions. According to Farrell, Merrill Lynch’s praised for their attempts to curb or halt the final CEO, John Thain, appears to have mis- excesses at the ground floor of the crisis— calculated the length and depth of the storm although these attempts were often quashed of the crisis. Thain’s guarded optimism that by active lobbying and federal intervention. the crisis would pass and the market would Avoiding policy prescriptions, McLean and rebound led him to make incorrect deci- Nocera’s account concludes with a series of sions on the size of the repairs needed by the open-ended questions about the future of company—although Farrell also keeps open the government’s role in mortgage finance. the possibility that Merrill Lynch was an Gretchen Morgenson and Joshua Rosner’s irreparable cause without an outside buyer. book, Reckless Endangerment: How In the end, Bank of America, with its insular, Outsized Ambition, Greed, and Corruption regional corporate culture, became Merrill’s Led to Economic Armageddon, published last resort. in May 2011, extends this inquiry into the Lo: Reading About the Financial Crisis 173 government’s past role in mortgage finance cited. One hopes that future editions will and in creating the conditions for the hous- rectify this glaring omission. ing bubble to begin. Morgenson, a Pulitzer Prize-winning financial journalist at the 4. Fact and Fantasy New York Times, and Rosner, an indepen- dent Wall Street analyst who spotted early There are several observations to be made problems among the government-sponsored from the number and variety of narratives enterprises, trace the origins of the crisis to that the authors in this review have proffered. a program of systematic regulatory capture The most obvious is that there is still signifi- of Fannie Mae and Freddie Mac beginning cant disagreement as to what the underlying in the early 1990s. The authors are particu- causes of the crisis were, and even less agree- larly suited to this task: Rosner was an ana- ment as to what to do about it. But what may lyst of the industry as the regulations were be more disconcerting for most economists implemented, while Morgenson specializes is the fact that we can’t even agree on all the in financial scandals and conflicts of interest. facts. Did CEOs take too much risk, or were In many ways, Reckless Endangerment is they acting as they were incentivized to act? a necessary work of regulatory archaeology. Was there too much leverage in the system? The Clinton administration’s pursuit of a Did regulators do their jobs or was forbear- policy of low-income home ownership was ance a significant factor? Was the Fed’s low captured, often willingly and far too easily, interest-rate policy responsible for the hous- by profit interests. Fannie Mae and Freddie ing bubble, or did other factors cause hous- Mac, as government-sponsored enterprises, ing prices to skyrocket? Was liquidity the used their status as quasi-governmental issue with respect to the run on the repo organizations to gain business advantage, market, or was it more of a solvency issue and used their business profits to gain politi- among a handful of “problem” banks? cal advantage, in a round-robin of influence For financial economists—who are used peddling. Cronyism became the rule of to dealing with precise concepts such as no- the day, as with the Countrywide “Friends arbitrage conditions, portfolio optimization, of Angelo” program to offer “sweetheart” linear risk/reward trade-offs, and dynamic loans to influential political figures, a pro- hedging strategies—this is a terribly frustrat- gram whose blatant nature one might expect ing state of affairs. Many of us like to think to see in a developing nation or a corrupt of financial economics as a science, but com- municipality, rather than at the highest lev- plex events like the financial crisis suggest els of the American government. As paired that this conceit may be more wishful think- reading with Acharya et al.’s Guaranteed to ing than reality. Keynes had even greater Fail, this is especially illuminating. One sig- ambitions for economics when he wrote, “If nificant scholarly problem with Morgenson economists could manage to get themselves and Rosner’s account, however, is its lack thought of as humble, competent people on a of sourcing. Major assertions are left hang- level with dentists, that would be splendid.”22 ing in the text without an independent way Instead, we’re now more likely to be thought to verify them. There is no footnote or end- of as astrologers, making pronouncements note apparatus, and the index is poorly con- and predictions without any basis in fact or structed. Much of Reckless Endangerment empirical evidence. is apparently based on earlier reporting by Morgenson or Rosner dating back to the mid-1990s, but the individual articles aren’t 22 Keynes (1932, 373). 174 Journal of Economic Literature, Vol. L (March 2012)

To make this contrast more stark, com- memos, emails, text messages, and vague pare the authoritative and conclusive acci- impressions that a CEO is bombarded with dent reports of the National Transportation almost 24/7, not all of which is true; where Safety Board (NTSB)—which investigates the “flight controls” are often human subor- and documents the who–what–when– dinates, not mechanical devices or electronic where–and–why of every single plane switches; and where there is no single “flight crash—with the twenty-one separate and data recorder,” but rather hundreds of dis- sometimes inconsistent accounts of the tinct narratives from various stakeholders financial crisis we’ve just reviewed (and with different motivations and intentions, more books are surely forthcoming). Why generating both fact and fantasy. If we want is there such a difference? The answer is to determine whether or not the failure of simple: complexity and human behavior. Lehman Brothers was due to “pilot error,” While airplanes often crash because of like the NTSB, we need to reconstruct the human behavior or “pilot error,” the causes exact state of Lehman prior to the accident, of such accidents can usually be accurately deduce the state of mind of all the execu- and definitively determined with sufficient tives involved at the time, determine which investigatory resources. Typically there are errors of commission and omission they a small number of human actors involved— made, and rule out all but one of the many the pilots, an air traffic controller, and per- possible explanations of the realized course haps some maintenance crew. Also, the of events. nature of accidents in this domain is fairly Given that we can’t even agree on a set of tightly constrained: an airplane loses aero- facts surrounding the financial crisis, nor do dynamic lift and falls to the ground. While we fully understand what the “correct” oper- there may be many underlying reasons for ation of a financial institution ought to be in such an outcome, investigators often have a every circumstance, the challenges facing pretty clear idea of where to look. In other economists are far greater than those faced words, we have sufficiently precise models by the NTSB. However, the stakes are also for how airplanes fly so that we can almost far higher, as we’ve witnessed over the past always determine the specific causal fac- four years. There is a great deal to be learned tors for their failure through relatively lin- from the NTSB’s methods and enviable track ear chains of physical investigation and record, as Fielding, Lo, and Yang (2011) logical deduction. Human behavior is just illustrate in their case study of this remark- one part of that chain, and thanks to flight able organization. And one of the most basic data recorders and the relatively narrow elements of their success is starting with a set of operations that piloting an aircraft single set of incontrovertible facts. In other involves—for example, the pilot must lower words, we need the equivalent of the “black the landing gear before the plane can land, box” flight data recorder for the financial and there’s only one way to lower it—the industry, otherwise we may never get to the complexity of the human/machine inter- bottom of any serious financial accident.23 face isn’t beyond the collective intellectual horsepower of the NTSB’s teams of expert investigators. Now compare this highly structured con- 23 This was precisely the motivating logic behind the text with piloting an investment bank, where Dodd Frank Act’s creation of the Office of Financial Research, but its future is unclear given the current politi- the “instrument panel” is the steady stream cal stalemate that has brought a number of important leg- of news reports, market data, internal islative initiatives to a standstill. Lo: Reading About the Financial Crisis 175

An instructive example of the importance rules. At Bear Stearns, the leverage ratio— of getting the facts straight is the role that a measurement of how much the firm was borrowing compared to its total assets—rose financial leverage played in the crisis, which sharply, to 33 to 1. In other words, for every is described in Lo and Mueller (2010, 50–51). dollar in equity, it had $33 of debt. The ratios at On August 8, 2008, the former director of the the other firms also rose significantly. SEC’s Division of Market Regulation (now the “Division of Markets and Trading”), Lee Pickard (2008), published an article in the The reports of sudden increases in leverage American Banker with a bold claim: a rule from 12-to-1 to 33-to-1 seemed to be the change by the SEC in 2004 allowed broker- “smoking gun” that many had been searching dealers to greatly increase their leverage, for in their attempts to determine the causes contributing to the financial crisis.24 In par- of the Financial Crisis of 2007–09. If true, ticular, Mr. Pickard (2008, 10) argued that it implied an easy fix according to Pickard before the rule change, (2008, 10): “The SEC should reexamine its . . . the broker-dealer was limited in the net capital rule and consider whether the amount of debt it could incur, to about 12 traditional standards should be reapplied to times its net capital, though for various reasons all broker-dealers.” broker-dealers operated at significantly lower While these “facts” seemed straightfor- ratios . . . If, however, Bear Stearns and other ward enough, it turns out that the 2004 SEC large broker-dealers had been subject to the typical haircuts on their securities positions, an amendment to Rule 15c3–1 did nothing to aggregate indebtedness restriction, and other change the leverage restrictions of these provisions for determining required net capital financial institutions. In a speech given by under the traditional standards, they would not the SEC’s director of the Division of Markets have been able to incur their high debt lever- and Trading on April 9, 2009 (Sirri 2009), age without substantially increasing their capi- tal base. Dr. Erik Sirri stated clearly and unequivo- cally that “First, and most importantly, the He was referring to a change in June 2004 Commission did not undo any leverage to SEC Rule 15c3–1, the so-called “net restrictions in 2004.”25 He cites several docu- capital rule” by which the SEC imposes net mented and verifiable facts to support this capital requirements and, thereby, limits the leverage employed by broker-dealers. This story was picked up by a number of news- 25 SEC Rule 15c3–1 is complex, and not simply a lever- age test. The rule does contain a 15-to-1 leverage test with papers, including the New York Times on a 12-to-1 “early warning” obligation. However, this com- October 3, 2008 (Labaton 2008, A1): ponent of the rule only limits unsecured debt, and did not In loosening the capital rules, which are sup- apply to large broker-dealers, who were subject to net capi- posed to provide a buffer in turbulent times, tal requirements based on amounts owed to them by their customers, i.e., a customer-receivable or “aggregate debit the agency also decided to rely on the firms’ item” test. This test requires a broker-dealer to maintain own computer models for determining the net capital equal to at least 2 percent of those receivables, riskiness of investments, essentially outsourc- which is how the five large investment banks had been able ing the job of monitoring risk to the banks to achieve higher leverage ratios in the 1990s than after themselves. the 2004 rule change (see figure 1). Similarly, their bro- Over the following months and years, each of ker-dealer subsidiaries (which were the entities subject to the net capital rule) had long achieved leverage ratios far the firms would take advantage of the looser in excess of 15-to-1. The historical leverage ratios of the investment banks were readily available in their financial reports, and the facts regarding the true nature of the SEC net capital rule were also available in the public domain. I 24 I thank Jacob Goldfield for bringing this example to thank Bob Lockner for decoding the intricacies of the SEC my attention. net capital rule. 176 Journal of Economic Literature, Vol. L (March 2012) surprising conclusion,26 and this correction time these news stories were published, and was reiterated in a letter from Michael easily accessible through company annual Macchiaroli, Associate Director of the SEC’s reports and quarterly SEC filings. Division of Markets and Trading to the Of course, the arcane minutiae of SEC General Accountability Office (GAO) on net capital rules may not be common knowl- July 17, 2009, and reproduced in the GAO edge, even among professional economists, Report GAO–09–739 (2009, 117). accountants, and regulators. But two aspects What about the stunning 33-to-1 leverage of this story are especially noteworthy: (1) ratio reported by the press? According to the the misunderstanding seems to have origi- GAO (Report GAO–09–739, 2009, 40): nated with Mr. Pickard, a former senior In our prior work on Long-Term Capital SEC official who held the very same position Management (a hedge fund), we analyzed from 1973 to 1977 as Dr. Sirri did from 2006 the assets-to-equity ratios of four of the five to 2009, and who was directly involved in broker-dealer holding companies that later drafting parts of the original version of Rule became CSEs and found that three had ratios 15c3–1; and (2) the mistake was quoted as equal to or greater than 28-to-1 at fiscal year- end 1998, which was higher than their ratios fact by a number of well-known legal schol- 27 at fiscal year-end 2006 before the crisis began ars, economists, and top policy advisors. Lo (see figure 6). and Mueller (2010) conjecture that these interpretations of Rule 15c3–1 emerged In footnote 68 of that report, the GAO through the apparent consistency and coin- observes that its 1999 report GAO/ cidence between the extraordinary losses GGD–00–3 (1999) on Long-Term Capital of Bear, Lehman, and Merrill and the 2004 Management “ . . . did not present the assets- SEC rule change—after all, it seems per- to-equity ratio for Bear Stearns, but its ratio fectly plausible that a loosening of net capital also was above 28 to 1 in 1998.” The GAO’s rules in 2004 could have caused broker-deal- graph of the historical leverage ratios for ers to increase their leverage. When new Goldman Sachs, Merrill Lynch, Lehman information confirms our priors, we usually Brothers, and Morgan Stanley is reproduced don’t ask why. in figure 1 (see section 1). These leverage This example underscores the critical numbers were in the public domain at the need to collect, check, and accumulate facts

26 So what was this rule change about, if not about relax any requirements at the holding company level changing leverage restrictions? It was meant to apply because previously there had been no requirements. In only to the five largest U.S. investment banks which were fact, the Commission increased its supervisory access to at a competitive disadvantage in conducting business the CSE investment bank holding companies.” Now with in Europe because they didn’t satisfy certain European respect to the net capital rule, Sirri (2009) explains that regulatory requirements dictated by the Basel Accord. it had nothing to do with leverage constraints: “The net By subjecting themselves to broader regulatory super- capital rule requires a broker-dealer to undertake two vision—becoming designated “Consolidated Supervised calculations: (1) a computation of the minimum amount Entities” or CSEs—these U.S. firms would be on a more of net capital the broker-dealer must maintain; and (2) equal footing with comparable European firms. As Sirri a computation of the actual amount of net capital held (2009) explains: “Thus the Commission effectively added by the broker-dealer. The ‘12-to-1’ restriction is part of an additional layer of supervision at the holding com- the first computation and it was not changed by the 2004 pany where none had existed previously. While certain amendments. The greatest changes effected by the 2004 changes were made in 2004 to the net capital rule to amendments were to the second computation of actual conform more closely with the methods of computing net capital.” capital adequacy that would be applied at the holding 27 See, for example, Coffee (2008), Blinder (2009), company, the changes were unrelated to the ‘12-to-1’ Reinhart and Rogoff (2009, 213–14), Stiglitz (2009; 2010, restriction Thus, the Commission did not eliminate or 163), and Woodward (2009). Lo: Reading About the Financial Crisis 177 from which more accurate inferences can may not be news, but it does make for good then be drawn.28 Without the immutable economic science. hard platform of objective facts on which we can build an accurate narrative of the crisis References that stands the test of time, there’s little hope Acharya, Viral V., Matthew Richardson, Stijn Van Nieu- for scientific progress as the waves of public werburgh, and Lawrence J. White. 2011. Guaranteed opinion toss our perspective in one direction to Fail: Fannie Mae, Freddie Mac and the Debacle of or another. This is one of the most compel- Mortgage Finance. Princeton and Oxford: Princeton University Press. ling reasons to read more than one account Akerlof, George A. 1970. “The Market for ‘Lemons’: of the financial crisis, and to seek out those Quality Uncertainty and the Market Mechanism.” books that may not agree with our precon- Quarterly Journal of Economics 84 (3): 488–500. Akerlof, George A., and Robert J. Shiller. 2009. Ani- ceptions, just in case we’ve been inadver- mal Spirits: How Human Psychology Drives the tently misinformed. Readers will find the 21 Economy, and Why It Matters for Global Capitalism. books reviewed in this article to be useful Princeton and Oxford: Princeton University Press. Allen, Franklin, and Douglas Gale. 2002. “A Compara- but not unbiased or flawless inputs to their tive Theory of Corporate Governance.” University of own critical thinking about the crisis. Given Pennsylvania Wharton Financial Institutions Center the complexity of the events surrounding Working Paper 03-27. Bebchuk, Lucian A., Alma Cohen, and Holger this debacle, the best hope for arriving at a Spamann. 2010. “The Wages of Failure: Executive deeper understanding of financial crises and Compensation at Bear Stearns and Lehman 2000– how to respond to them is through the col- 2008.” Harvard University John M. Olin Center for Law, Economics, and Business Discussion Paper lective intelligence of all economists, each of 657. us laboring to develop our own interpreta- Bikhchandani, Sushil, David Hirshleifer, and Ivo tion that can inform and improve the con- Welch. 1992. “A Theory of Fads, Fashion, Custom, and Cultural Change in Informational Cascades.” sensus. Like the characters in Rashomon, we Journal of Political Economy 100 (5): 992–1026. may never settle on a single narrative that Blinder, Alan S. 2009. “Six Errors on the Path to the explains all the facts; such a “super-narrative” Financial Crisis.” The New York Times, January 24. Coffee, John C. 2008. “Analyzing the Credit Crisis: Was may not even exist. But by working with a the SEC Missing in Action?” New York Law Journal, common set of facts, we have a much better December 5. chance of responding more effectively and Cohan, William D. 2009. House of Cards: A Tale of Hubris and Wretched Excess on Wall Street. New preparing more successfully for future crises. York: Random House, Doubleday. As of October 19, 2011, the New York Farrell, Greg. 2010. Crash of the Titans: Greed, Hubris, Times has yet to print a correction of its the Fall of Merrill Lynch, and the Near-Collapse of Bank of America. New York: Random House, Crown original stories on the 2004 change to Rule Business. 15c3–1, nor did the Times provide any cov- Fielding, Eric, Andrew W. Lo, and Jian Helen Yang. erage of Dr. Sirri’s April 9, 2009 speech. 2011. “The National Transportation Safety Board: A Model for Systemic Risk Management.” Journal of Correcting mistaken views and factual errors Investment Management 9 (1): 17–49. French, Kenneth R., et al. 2010. The Squam Lake Report: Fixing the Financial System. Princeton and Oxford: Princeton University Press. 28 The unintentional propagation of pseudo-facts with Garnaut, Ross, and David Llewellyn-Smith. 2009. The its subsequent impact on general beliefs and actions is Great Crash of 2008. Carlton, Victoria: Melbourne hardly unique to financial crises. The great sociologist University Press. Robert K. Merton (1987, 3), father of the economist, Gorton, Gary B. 2010. Slapped by the Invisible Hand: observed more than two decades ago that “establishing the The Panic of 2007. Oxford and New York: Oxford phenomenon” cannot be taken for granted and provided University Press. several vivid examples drawn from the sciences and sociol- Johnson, Simon, and James Kwak. 2010. 13 Bankers: ogy in which mistaken beliefs were subsequently accepted The Wall Street Takeover and the Next Financial and cited as fact by several experts before eventually being Meltdown. New York: Random House, Pantheon corrected. Books. 178 Journal of Economic Literature, Vol. L (March 2012)

Keynes, John Maynard. 1932. Essays in Persuasion. Folly. Princeton and Oxford: Princeton University New York: Harcourt Brace. Press. Kowalski-Trakofler, Kathleen M., Charles Vaught, and Roubini, Nouriel, and Stephen Mihm. 2010. Crisis Ted Scharf. 2003. “Judgment and Decision Making Economics: A Crash Course in the Future of Finance. under Stress: An Overview for Emergency Manag- New York: Penguin Press. ers.” International Journal of Emergency Manage- Shiller, Robert J. 2008. The Subprime Solution: How ment 1 (3): 278–89. Today’s Global Financial Crisis Happened, and What Lewis, Michael. 2010. The Big Short: Inside the Dooms- to Do about It. Princeton and Oxford: Princeton Uni- day Machine. New York and London: Norton. versity Press. Lo, Andrew W., and Mark T. Mueller. 2010. “Warning: Shilling, A. Gary. 1993. “Scoreboard.” Forbes February Physics Envy May Be Hazardous to Your Wealth!” 15: 236. Journal of Investment Management 8 (2): 13–63. Sirri, Erik R. 2009. “Securities Markets and Regula- Lowenstein, Roger. 2010. The End of Wall Street. New tory Reform.” Speech delivered to the National York: Penguin Group, Penguin Press. Economists Club, April 9. http://www.sec.gov/news/ Lucas, Deborah, and Robert McDonald. 2011. “Valu- speech/2009/spch040909ers.htm. ing Government Guarantees: Fannie and Freddie Sorkin, Andrew Ross. 2009. Too Big to Fail: The Inside Revisited.” In Measuring and Managing Federal Story of How Wall Street and Washington Fought to Financial Risk, edited by Deborah Lucas, 131–54. Save the Financial System—and Themselves. New Chicago and London: University of Chicago Press. York: Penguin Press, Viking. McLean, Bethany, and Joe Nocera. 2010. All the Devils Stiglitz, Joseph E. 2009. “Capitalist Fools.” Vanity Fair, Are Here: The Hidden History of the Financial Cri- January. sis. New York: Penguin Group, Portfolio/Penguin. Stiglitz, Joseph E. 2010. Freefall: America, Free Mar- Merton, Robert K. 1987. “Three Fragments from a kets, and the Sinking of the World Economy. New Sociologist’s Notebooks: Establishing the Phenom- York and London: Norton. enon, Specified Ignorance, and Strategic Research Tett, Gillian. 2009. Fool’s Gold: How the Bold Dream of Materials.” Annual Review of Sociology 13: 1–28. a Small Tribe at J. P. Morgan Was Corrupted by Wall Morgenson, Gretchen, and Joshua Rosner. 2011. Reck- Street Greed and Unleashed a Catastrophe. New less Endangerment: How Outsized Ambition, Greed, York: Simon and Schuster, Free Press. and Corruption Led to Economic Armageddon. New Thomas, Landon, Jr. 2008. “Down $900 Million or York: Henry Holt and Company, Times Books. More, the Chairman of Bear Sells.” New York Times, Murphy, Kevin J. Forthcoming. “Pay, Politics and the March 28. Financial Crisis.” In Economic Lessons from the U.S. Government Accountability Office. 2009. “Trou- Financial Crisis, edited by Alan S. Blinder, Andrew bled Asset Relief Program: Capital Purchases Pro- W. Lo, and Robert Solow. New York: Russell Sage gram Transactions for the Period October 28, 2008 Foundation. through September 25, 2009 (GAO-10-24SP): An Paulson, Henry M. 2010. On the Brink: Inside the Race e-Supplement to GAO-10-16.” http://www.gao.gov/ to Stop the Collapse of the Global Financial System. special.pubs/gao-10-24sp/file1.html. New York: Business Plus. Woodward. Susan. 2009. “The Subprime Crisis of 2008: Pickard, Lee A. 2008. “Viewpoint: SEC’s Old Capital A Brief Background and a Question.” ASSA Session Approach Was Tried—and True.” American Banker, on Recent Financial Crises, January 3. August 8. Woolley, Anita Williams, Christopher F. Chabris, Alex Pozsar, Zoltan, Tobias Adrian, Adam Ashcraft, and Pentland, Nada Hashmi, and Thomas W. Malone. Hayley Boesky. 2010. “Shadow Banking.” Federal 2010. “Evidence for a Collective Intelligence Factor Reserve Bank of New York Staff Report 458. in the Performance of Human Groups.” Science 330 Rajan, Raghuram G. 2010. Fault Lines: How Hidden (6004): 686–88. Fractures Still Threaten the World Economy. Princ- Zuckerman, Gregory. 2009. The Greatest Trade Ever: eton and Oxford: Princeton University Press. The Behind-the-Scenes Story of How John Paulson Reinhart, Carmen M., and Kenneth S. Rogoff. 2009. Defied Wall Street and Made Financial History. New This Time Is Different: Eight Centuries of Financial York: Random House, Crown Business.