Guaranteed to Fail: Fannie Mae, Freddie Mac and the Debacle of Mortgage Finance
The Authors’ Revolutionary Indictment of Systemically Dangerous Institutions (SDIs)
Overview of the Authors’ Logic
The title of the book is a pun. Fannie and Freddie, unlike Ginnie Mae, were privately-owned and managed entities whose liabilities were not guaranteed by the U.S. Treasury prior to the bailout. Indeed, their debt issuances explicitly warned purchasers that the debt was not guaranteed by the U.S. Treasury. Fannie and Freddie, however, were “too big to fail” so the markets correctly perceived that the U.S. would not permit the bonds they issued to default. Fannie and Freddie, like other systemically dangerous institutions (SDIs), therefore had an implicit Treasury guarantee of their debts even though all of the U.S. SDIs were entirely privately owned and managed. Guarantees normally protect against the cost of failure, but the authors’ thesis is that the implicit guarantees induce perverse behavior by the senior managers who control the SDIs and cause them to make business decisions that ensure that the SDIs will fail. The perverse incentives arising from this implicit governmental guarantee are the authors’ core concept.
The twenty largest U.S. banks are typically considered to be SDIs. The SDIs, in every highly developed large “free market” economy, dominate the financial system and the financial systems typically dominate the national economy. The Stern authors cite approvingly a study claiming that 59% of all liabilities in the financial system have explicit (typically deposit insurance) or implicit (because they are held by SDIs) Treasury guarantees (p. 134). The SDIs’ domination of finance and the overall economy has increased substantially over the last 30 years. The authors emphasize that the SDIs are even more dominant than the bare statistics reflect, for small banks are incapable of competing with the SDIs. The authors argue that the SDIs ability to borrow more cheaply due to the implicit (underpriced) Treasury guarantee and their crippling political power produces a regulatory “race to the bottom” that favors SDIs by (typically) allowing them much lower capital requirements, which allows them far greater leverage and higher (reported) returns on equity (p. 134). The SDIs inevitably become ever more dominant.
The authors are four finance professors at NYU’s Stern School. New York City was ground zero of the financial crisis. The Stern School is one of the leading educators of the SDIs’ officers. The Stern authors’ logic makes this a revolutionary book. Their logic sounds the death knell of “modern finance” and much of microeconomic and macroeconomic theory. The book is a case study of the perverse behavior of the managers controlling two SDIs, but the authors explicitly generalize the perverse incentives as controlling all SDIs (here and abroad).
The Stern authors’ logic invites the reader to look behind the edifices and see that the modern “free market” economy is a Potemkin prank. The authors provide a superb example of why James Galbraith’s titled his most recent book The Predator State: How Conservatives Abandoned the Free Market and Why Liberals Should Too. The irony of modern “capitalism” is that its leaders are the CEOs running the SDIs and, as the authors demonstrate, SDIs inherently make it impossible for markets to be “free”, efficient, or sound. The Stern authors found that the the SDIs produced “a highly distorted market with two types of institutions – LCFI King Kongs and GSE Godzillas – both implicitly backed by the government….” (p. 55). (LCFI: “large complex financial institutions” – the authors’ polite euphemism for SDIs.) The Stern authors conclusion that “there was nothing free about these [housing finance] markets” applies to all the SDIs, for all SDIs are “government sponsored enterprises” (GSEs) under the authors’ logic (p. 21). It is illusory to talk of the twenty largest banks in the U.S. as if it were not GSEs. The authors acknowledge this point:
“But when one scratches below the surface, the failure of the LCFIs and the GSEs is quite similar – a highly leveraged bet on the mortgage markets by firms that were implicitly backed by the government with artificially low funding rates only to differing degrees” (p. 49).
The authors also endorse the Congressional Budget Office’s (CBO’s) metaphor for the SDIs. The CBO said that the Fannie and Freddie had become so large that dealing with them was like sharing a canoe with a bear. The authors then urged:
“Because the GSEs are currently under the conservatorship of the government, it would be crazy not to kill off the “bear” and move forward with a model that did not again create a too-big-to-fail – and, more likely, a too-big-to-reform – monster” (p. 74).
If we’re going bear hunting, then the authors’ logic requires us to “kill off” all the “bears.” That means getting rid of all the “too-big-to-fail”, “too-big-to-regulate”, and “too-big-to-prosecute” banks. Again, the Stern authors’ logic is revolutionary. It would indeed “be crazy not to kill off the bear[s]” for they are inherently “monster[s]” who hold our economy hostage and block real markets and real democracy.
The SDI CEOs that dominate our economy and government are implacable foes of real markets. The Stern authors show that the CEOs’ bleating about the wonders of the markets are engaged in a cynical sham designed to block effective regulation. The SDIs’ CEOs oppose effective regulation precisely because it would create more competitive markets. They hate free markets. They want rigged markets that ensure they will destroy any honest competitor. The Stern authors are explicit on this point; stating that it is impossible for non-SDIs to compete with SDIs. Indeed, their argument is that it impossible for smaller (but still giant) SDIs to compete with the largest SDIs. Their simile is the largest SDIs’ advantages are so great that it is “like bringing a gun to a knife fight” (p. 22).
The Stern authors also confirm that the SDIs’ CEOs are not risk-takers in any sense that we normally use that term. The CEOs seek a sure thing. Risk-taking is supposed to be the essence of capitalism. The Stern authors’ case study provides yet another example supporting Robert Prasch’s insights about risk. His classic 2004 article is entitled “Shifting risk: the divorce of risk from reward in American capitalism.” In 1993, George Akerlof and Paul Romer authored the most important economics paper on financial crises, entitled Looting: the Economic Underworld of Bankruptcy for Profit. Akerlof & Romer explained the ultimate separation of risk and reward – what criminologists now term “accounting control fraud.” Akerlof & Romer explained how financial CEOs used accounting fraud to make record reported profits a “sure thing” (p. 5). The record, albeit fictional, reported profits were certain to make the CEO and his confederates wealthy through modern executive compensation. The bank was guaranteed to fail. The only question was how soon it would fail.
The key word in the Stern authors’ book title is “guaranteed.” The authors have confirmed Akerlof & Romer’s central insight. Absent effective private market discipline, bank CEOs have perverse incentives to make (and purchase) loans that have a negative expected value. In plain English that means that they deliberately make and purchase loans that will lose money. (They also sell loans to others that they know have a negative expected value, but this book does not deal with that form of fraud.) The CEOs’ perverse incentive creates three concurrent guarantees: the bank will report high (albeit fictional) short-term profits, the controlling officers will extract large increases in wealth through modern executive compensation, and the bank will suffer large losses. The bank fails, but the controlling officers can walk away rich. In economics jargon, “control frauds” represent the ultimate form of destructive “rent-seeking” behavior. The Stern authors do not note the fact that the SEC explicitly charged Fannie’s controlling officers with engaging in accounting and securities fraud for the purpose of inflating Fannie’s reported income so that they could extract greater personal income from Fannie by maximizing their bonuses. (Disclosures: I was an expert for OFHEO in its enforcement action against Fannie’s controlling officers and opined that they were engaged in accounting control fraud. I worked under one of the authors of the book, Lawrence White, when I was a financial regulator. I consider him a personal friend.)
Akerlof & Romer confirmed the findings of the National Commission on Financial Institution Reform, Recovery and Enforcement, the S&L regulators, prosecutors (and juries), and criminologists – looting drove the second phase of the S&L debacle. (The first phase was caused by interest rate risk.) The Stern authors’ case study of two of the many SDIs that made or purchased large amounts of “liar’s” loan paper finds that the same looting dynamic drove the current crisis. This confirms the findings of the nation’s top economists and criminologists as to why we suffer from recurrent, intensifying financial crises. (With one major caveat: criminologists recognize that an implicit governmental guarantee is not necessary to create these perverse incentives. The Enron-era accounting control frauds are simply the clearest example of this point. The authors’ assumption that private market discipline is effective in preventing accounting control frauds absent a governmental guarantee has long been falsified by criminologists.)
The Authors Related Tenets: “Tail Risk” and the “Race to the Bottom”
The Stern authors never use the “f-word” (fraud), but they are not lawyers, prosecutors, or criminologists. They do not cite any of the research on the fraud, even the relevant economics literature. It follows that they do not use the term or describe the key concept of a “criminogenic environment.” Nevertheless, what they describe is a accounting control fraud and what they seek to explain is why the combination of the three “de’s” (deregulation, desupervision, and de facto decriminalization), the implicit Treasury guarantee of the SDIs, and modern executive compensation produced an intensely criminogenic environment that generated an epidemic of accounting control fraud by the SDIs. Their explanation is often inexplicit. For example, I do not think this book discusses executive compensation even though understanding modern executive and professional compensation is essential to understanding the authors’ logic.
The Stern authors do not explain their concept of an extreme tail gamble very explicitly in this book, but they do say that Fannie and Freddie followed that strategy and that doing so guaranteed their failure. The authors also argue that it was Fannie and Freddie’s purchase of nonprime loan paper that guaranteed their failure. Those purchases, however, were not honest “bets” and they were not subject to loss only in “rare” circumstances. As I explain in the technical appendix, it was pervasively fraudulent “liar’s” loans that sank the SDIs, hyper-inflated the housing bubble, and caused the great recession.
I believe that the authors’ logic chain is as follows:
- SDI executives caused “their” banks to make investments that had a negative expected value (because they would eventually suffer severe losses)
- In violation of generally accepted accounting principles (GAAP), the SDIs did not provide remotely adequate allowances for those future losses (transmuting a real economic loss into a fictional accounting gain)
This created a “sure thing” in which the bank was guaranteed to report high (fictional) short-term profits because they received relatively high yield for purchasing riskier assets without providing the allowances for loan losses required to provide for future losses
This guaranteed fictional income led to the guaranteed payment of exceptional income to the executives due to modern executive compensation
The officers controlling the SDIs also used their power over professional compensation (e.g., of auditors, appraisers, and rating agencies) to create a “race to the bottom” that led to widespread appraisal fraud and suborned auditors and credit rating agencies (who are supposed to serve as external “controls”) and transformed them into the most valuable allies of the accounting control frauds (the SDIs did the same thing with their loan officers and loan brokers, creating what criminologists call “echo” fraud epidemics – but the authors do not discuss these fraud epidemics)
The accounting control fraud mechanism spread rapidly through the SDIs because as soon as any SDI adopted the fraud strategy it created a “race to the bottom” that inexorably drove other SDIs to engage in accounting control fraud
This hyper-inflated the housing bubble because the fraudulent officers controlling the SDIs found that making and purchasing vast numbers of mortgage loans to the uncreditworthy (at a premium yield) optimized the creation of fictional short-term income
The hyper-inflation of the bubble allowed the SDIs to continue their frauds for many years because they could simply refinance bad loans or sell the collateral for minimal loss even if the loan defaulted (the saying in the trade is: “a rolling loan gathers no loss”)
The SDIs’ creditors did not exercise effective private market discipline to prevent these frauds because it is expensive to provide effective discipline and creditors do not find it worthwhile to bear that expense where they are the beneficiaries of a governmental guarantee against loss
The SDIs’ regulators did not act as the regulatory “cops on the beat” to break this private sector “race to the bottom” because the SDIs’ used their political power and ideological “capture” to create a regulatory “race to the bottom” (p. 191, n. 3) that led allowed the fraudulent SDI managers to act with impunity from regulatory and criminal sanctions
The result was that SDIs following this fraud strategy were guaranteed to suffer massive loan losses and fail (unless they were bailed out by the government) because of the losses inherent in making high risk loans and the absence of any meaningful allowances for losses
These fraud epidemics, SDI failures, and the collapse of the hyper-inflated bubble caused multiple markets to collapse and triggered the Great Recession
The Authors’ Proposed Reforms are Criminogenic
The Stern authors appear to believe that they can put the Genie back in the bottle and de-guarantee the SDIs. This is a dangerous illusion. There is no way to make an SDI “private.” By definition, it will carry an implicit governmental guarantee of at least some of its liabilities. Government claims that they will not bail out the SDIs’ creditors are not, and cannot be, credible. The Bush administration, for example, claimed that the government did not and would not guarantee repayment of Fannie and Freddie’s bonds. As soon as Fannie and Freddie’s problems became public the Bush administration rushed to guaranteed Fannie and Freddie’s bonds.
The “tail risk” strategies they describe as guaranteeing the SDIs’ failures are fraudulent accounting acts taken at the direction of the firms’ controlling officers. Any analysis that ignores the fraudulent nature of the managers’ actions is certain to distract us from the reforms essential to remove successfully the perverse incentives that cause our recurrent, intensifying financial crises. Indeed, ignoring fraud can lead to proposing reforms that are criminogenic.
Reforms that might work for honest firms, such as the Stern authors’ suggestion that the Treasury charge the SDIs a fee equivalent to the value of the implicit Treasury subsidy to the SDIs, will fail to deter accounting control fraud because the very nature of the fraud causes the risks the lender or purchaser is taking (and those risks drive the value of the Treasury subsidy to the SDIs under the authors’ logic) to be hidden by the deceit. The result is that the risk and the fees are massively understated and the fraud is not deterred. Indeed, the unexpected consequences of authors’ proposed reform would be an intensified Gresham’s dynamic because the accounting frauds would have an even greater advantage (due to the grossly inadequate charge for their implicit Treasury subsidy) over their honest competitors. The Stern authors’ term for this Gresham’s dynamic is a “race to the bottom.” They agree that this perverse dynamic leads to the frauds driving their ethical competitors out of the marketplace absent effective regulation. It is essential, therefore, for economists to determine whether the underlying problem is control fraud and whether the reforms the economists are proposing would be effective against control frauds. The punch line of the standard economics joke is “assume a can opener” and disregarding control fraud makes that assumption and has repeatedly led neoclassical economists to propose criminogenic anti-regulatory policies. Fortunately, the Stern authors’ own logic demonstrates that the problem they have identified is the perverse incentive to engage in accounting control fraud to which that the officials controlling SDIs are inherently subject. Their logic explains why their proposed reform would increase control fraud. I urge them to reconsider their proposed reform in light of their logic and the relevant economics and criminology literature on fraud.
I have mentioned a second difficulty with the proposed reform – governmental guarantees are not necessary to create the criminogenic environments that drive our recurrent, intensifying financial crises. Enron and its ilk had no governmental guarantees and were not SDIs. Creditors rush to lend to, rather than disciplining, accounting control frauds because the frauds report record income and low losses. Indeed, in Enron’s case, the major banks deliberately and enthusiastically aided and abetted Enron’s frauds. The allied danger is that the Stern authors’ would counsel regulatory complacency if their reforms were adopted. They are strong advocates of relying on private market discipline and claim, without citation, that it is effective absent governmental guarantees. If their proposed reforms were adopted, their arguments would be cited as calling for minimal regulation. As I have explained, however, accounting control fraud targets the Achilles’ heel of private market discipline even in the absence of governmental guarantees so the proposed reforms would be criminogenic because they would lead to complacency and weaker regulation while failing to provide effective private market discipline where it was most essential.
Under the authors’ own logic (and metaphors and similes) the only reforms with any chance of success require us to shoot the bears – all of the bears. There is no reason for SDIs to exist. They are inefficiently large and would be more efficient if they were shrunk to the size that they no longer posed systemic risk. Why would we let a pack of fraudulent “King Kongs” control our financial and political systems? As I have argued, the Stern authors’ logic is revolutionary. Their remedies lack the courage of their logic and verbal imagery.
The “tails” of a statistical distribution are the extreme values, but the term is inexact because the shape of the tails can vary considerably and how wide and how “fat” the tails are is critical for analytical purposes. (More fundamentally, there is no exogenously determined distribution of economic events. The policies we follow create the distribution (the distribution is endogenous). The reason we have recurrent, intensifying financial crises is that we have adopted policies that are so criminogenic that they produce recurrent, intensifying crises. The “perfect storm”, “black swan”, and “tail risk” explanations of the crisis all crumble due to this basic statistical error.) But let us ignore this fundamental error for the sake of considering the Stern authors’ argument. Tail events can range from extremely probable (a tail can be so “fat” that it includes the great bulk of the distribution) to so improbable that Sol is likely to become a white dwarf billions of years from now before the event occurs. It’s a bit like saying that someone lives in America – the phrase covers a vast amount of territory. The Stern authors are also referring to only one tail of the distribution – the tail where bad things happen to the investor.
The Stern authors do not explain what kind of statistical distribution they believe to have existed for nonprime home loans or how extreme a tail event they are claiming that the SDIs’ officers chose to invest in. The authors’ discussion is not only vague, it is also internally inconsistent. The ambiguity and inconsistency are vital. One of the consistent problems with the neoclassical economics’ discussion of “moral hazard” has been the failure to articulate explicitly the business strategies that the scholars claim that moral hazard produced. One cannot test a hypothesis unless it is operationalized explicitly, clearly, and in a manner that is internally consistent.
At one juncture, the Stern authors assert, that losses would occur under the “tail” strategy only in unusual (but hardly rare) economic circumstances. The authors’ description of the strategy, in the case of Fannie and Freddie, was their CEOs’ decision to purchase “low quality mortgages that households will be unable to pay in a recession or a severe housing downturn” (p. 4). The U.S., since World War II, has suffered recessions roughly every five years. A U.S. home mortgage typically has a 30 year term, though prepayments are common. The typical home mortgage will be outstanding during at least one recession. If either a severe housing downturn or a recession will cause large numbers of defaults on nonprime loans, then a nonprime lender or purchaser that makes or buys large numbers on nonprime loans is guaranteed to fail. We don’t know which year the recession or fall in housing values will occur, but we know that such events occur fairly frequently.
When the Stern authors refer to “tail events” and apply a descriptor to the probability the word they choose is “rare” (p. 157). They also state: “When housing prices declined in 2007 and the tail risk that these institutions betted against materialized, they all experienced substantial stress” (p. 55).
The SDIs did not “bet” and they certainly did not make a bet that “rare” events would not occur. They also did not experience stress – they died. The government, however, resuscitated many of them through bail outs and perverting the accounting rules to hide their losses. The SDIs made, purchased, and sold massive amounts of “liar’s” loans. The Stern authors’ most obvious technical mistake is the failure to understand liar’s loans (p. 191, n. 2). There are four evident mistakes. First, the authors assert that liar’s loans and subprime loans are dichotomous (mutually exclusive categories). There are no official definitions of either “subprime” or any of the euphemisms for liar’s loans (“stated income”, “low and no doc”, and “alt-a”), but the best estimate is that by 2006 roughly 50% of loans called “subprime” were also liar’s loans. Similarly, loans called “alt-a” deteriorated in their (nominal, e.g., grossly inflated) stated credit quality over time. By 2006, roughly 30% of U.S. mortgage originations were liar’s loans.
Second, the authors do not even mention that the industry called “alt-a” loans “liar’s” loans. How can this not be relevant to a discussion of the risk of alt-a loans? This crisis is a mortgage fraud crisis. In September 2004, the FBI testified that mortgage fraud was “epidemic” and predicted that it would cause a financial “crisis” if it were not contained. No one claims it was contained. In 2006, the Mortgage Bankers Association (the trade association of the “perps”) sent its members the eighth report of its anti-fraud group (MARI). MARI reported three key research findings with regard to stated income loans:
- They were “open invitations to fraudsters”
- The incidence of fraud was 90%, and
- The loans deserved the description that the industry used: “liar’s” loans
After the FBI’s and MARI’s warnings the industry substantially increased the origination, sale, and purchase of liar’s loans. Note two other crucial facts:
It was lenders and their agents who overwhelmingly put the lies in “liar’s” loans and induced the endemic fraudulent (inflated) appraisals
Making liar’s loans in the mortgage context inherently creates intense “adverse selection”, which creates a “negative expected value” to lending. No honest lender would make “liar’s” loans – and the SDIs that purchased liar’s loans knew that, which is why they, and the credit rating agencies, followed the financial version of “don’t ask; don’t tell” when it came to (not) underwriting mortgage quality by reviewing samples of mortgage loan files.
Third, the Stern authors state that “liar’s” loans are “considered riskier than prime mortgages and less risky than subprime” (p. 191, n. 2). Note the false dichotomy between liar’s loans and subprime. The worst losses are on loans that are on subprime liar’s loans, but all “liar’s” loans have extreme fraud incidences.
Fourth, the authors claim that “Alt-A mortgages may have excellent credit but may not meet underwriting criteria for other reasons” (p. 191, n. 2). Liar’s loans weren’t simply overwhelmingly fraudulent; they were also promoted on the basis of this bright shining lie of “excellent credit” quality. Nobody with “excellent credit” and financial sophistication pays a premium yield. The claim that self-employed individuals could not get prime loans is false. They simply had to fill out the IRS form that allowed the borrower to check their tax returns. There may have been some liar’s loans made to people with “excellent credit”, but those are cases of extreme predatory lending. There were doubtless borrowers who did not fill out the IRS forms because they were involved in tax evasion and seeking to deceive their former spouses and their children as to their true net worth and income in order to minimize alimony and child support payments. Such individuals do not have “excellent credit” – if they’re willing to cheat others they represent extreme credit risks to a lender.
Liar’s loans are particularly important in the context of Fannie and Freddie because they provide a “natural experiment” that resolves the question of why Fannie and Freddie purchased so many nonprime loans after the SEC and OFHEO cracked down on their fraud schemes premised on rapid portfolio growth. The issue is whether Fannie and Freddie were reluctantly forced to purchase vast amounts of nonprime loans by Congressional mandate. The answer is no. Fannie and Freddie purchased the loans because they had a premium yield. Fannie and Freddie had no statutory mandate to purchase liar’s loans. Indeed, because liar’s loans endemically involved grossly (and fraudulently) inflated incomes and home prices they were less likely to qualify for affordable housing credit under the statute. Fannie and Freddie did not take public credit for purchasing vast amounts of liar’s loans. They would have done so had the purchases been made in response to political or statutory pressure. Instead, they engaged in further deceit by reporting that the liar’s loans that they purchased were “prime” loans.
The truth, which the authors ultimately acknowledge, is that Fannie and Freddie lost tremendous market share for the first half of the decade because they were reluctant to purchase large numbers of nonprime loans. Other SDIs were far larger purchasers of nonprime loans. Fannie and Freddie relied on a different form of accounting fraud. Their controlling officers caused them to grow their portfolios dramatically to take dramatically increased interest rate risk. Their controlling officers also caused them to adopt executive compensation plans that allowed extreme bonuses if Fannie and Freddie reported extreme income. Fannie guessed that interest rates would rise. Freddie guessed that they would fall. Rates fell. Fannie’s controlling officers engaged in accounting fraud to hide its losses (calling them hedges). Freddie’s controlling officers engaged in accounting fraud to hide its gains (creating “cookie jar” reserves they could draw on whenever needed to hit the income targets that would maximize their bonuses). The SEC and OFHEO ordered an end to both forms of accounting fraud and sharply limited the growth of their portfolios. OFHEO also changed Fannie and Freddie’s managers, but let the perverse executive pay practices in place that encouraged the accounting fraud. The result was that that there were two obvious paths to maximize Fannie and Freddie’s fictional incomes. They could increase substantially their purchase of loans and sales of mortgage backed securities (MBS) and they could reorient there relatively small permitted net growth in loans held in portfolio to higher yield mortgages. Nonprime loans were ideal purchases under both paths.
I hosted the Firedoglake forum for Dr. Rajan’s discussion of his recent book (also premised on the assertion that banks failed because they took extreme tail risk) and explained in my review and comments during the forum why investment scams involving assets that face only a “rare” risk of loss are inferior fraud strategies. (They are so inferior that they are vanishingly rare. Like the Stern authors, Rajan used repeated gambling metaphors, but his hypothetical example of an extreme tail “bet” actually required two frauds. Real frauds do not use his inferior hypothesized fraud scheme precisely because it requires would cover a risk that is rarely manifested.) The optimal fraud scheme is to obtain, promptly, a premium yield (which is only available if the risk is not rare, but rather highly probable), while providing trivial allowances for loan loss reserves that are grossly inadequate to cover the highly probable losses. This form of accounting control fraud, if combined with extreme growth and leverage, produces guaranteed, record (albeit fictional) short-term reported income. It guarantees that the controlling officers can extract massive rents and it guarantees that the firm will fail.
Why would the SDIs’ CEOs purchase or originate nonprime mortgage paper they knew would often default in such numbers that they guaranteed that the bank would fail? The Stern authors do not really address this question. They focus instead on why the CEOs were able to get away with imposing a business strategy that was suicidal for the bank. Most years, the U.S. economy is not in recession.
Indeed, they show that Fannie and Freddie followed the four-part accounting control fraud recipe for a loan purchaser that is guaranteed to maximize short-term (fictional) reported income, real losses, and managerial compensation.
Grow extremely rapidly by
- Purchasing bad loans at a premium yield while employing
- Extreme leverage and
- Providing grossly inadequate allowances for loan and lease losses (ALLL)
Two of the Stern authors were among the co-authors of a paper entitled “Manufacturing Tail Risk” that explained their concept of extreme tail risk a bit more detail. I have drawn on that paper to infer their concept of extreme tail risk in Guaranteed to Fail but I caution that neither the article nor the book is clear on what they argue was the business strategy that guaranteed failure. The Stern authors of the book do not discuss executive compensation and professional compensation, which (along with hiring, firing, promoting, and praising) are the primary means by which CEOs who adopt “guaranteed to fail” strategies create the incentives that pervert business behavior. Elegant control frauds suborn supposed internal and external “controls” into becoming fraud allies in a manner that also adds to the CEO’s deniability. The art is to generate a “Gresham’s” dynamic in which bad ethics drives good ethics out of the marketplace. The public result can be a fraud “epidemic,” hyper-inflation of a financial bubble, catastrophic losses, guaranteed bank failures, and financial crises. The private result is often that the CEOs and suborned executives and professionals become wealthy by looting the lender. The title of Akerlof & Romer’s 1993 article stressed this dichotomy between the private and public results of accounting control fraud.
The authors of “Manufacturing Tail Risk” address executive compensation at some length, but not professional compensation. They do not discuss the powerful and perverse Gresham’s dynamics that CEOs generate to assist and spread their frauds. This is curious given the fact that Fannie is the text book case for demonstrating these perverse dynamics.
Mr. Raines, Fannie’s CEO, explained in response to a media question what was causing the repeated scandals at elite financial institutions:
We’ve had a terrible scandal on Wall Street. What is your view?
Investment banking is a business that’s so denominated in dollars that the temptations are great, so you have to have very strong rules. My experience is where there is a one-to-one relation between if I do X, money will hit my pocket, you tend to see people doing X a lot. You’ve got to be very careful about that. Don’t just say: “If you hit this revenue number, your bonus is going to be this.” It sets up an incentive that’s overwhelming. You wave enough money in front of people, and good people will do bad things.
Under Raines, Fannie’s bonus program created “overwhelming” incentives. Raines learned that the unit at Fannie that should have been most resistant to this “overwhelming” financial incentive, Internal Audit; had succumbed to the perverse incentive. Mr. Rajappa, Senior Vice President for Operations Risk and Internal Audit instructed his internal auditors in a formal address in 2000 (and provided the text of the speech to Raines). ($6.46 refers to the earnings per share (EPS) number that will trigger maximum bonuses.)
By now every one of you must have 6.46 branded in your brains. You must be able to say it in your sleep, you must be able to recite it forwards and backwards, you must have a raging fire in your belly that burns away all doubts, you must live, breath and dream 6.46, you must be obsessed on 6.46…. After all, thanks to Frank [Raines], we all have a lot of money riding on it…. We must do this with a fiery determination, not on some days, not on most days but day in and day out, give it your best, not 50%, not 75%, not 100%, but 150%. Remember, Frank has given us an opportunity to earn not just our salaries, benefits, raises, ESPP, but substantially over and above if we make 6.46. So it is our moral obligation to give well above our 100% and if we do this, we would have made tangible contributions to Frank’s goals (emphasis in original).
The Stern authors of both Guaranteed to Fail and “Manufacturing Tail Risk” do not discuss accounting, the prime fraud “weapon” in the schemes they describe. Their “tail” risk logic, however, requires them to make powerful implicit assumptions that the SDIs engaged in endemic accounting and securities fraud. Indeed, their entire thesis – Guaranteed to Fail – is only true if the CEOs and CFOs of the fraudulent SDIs ensure that they provide grossly inadequate allowances for loan and lease loss (ALLL) reserves. If the reserves were adequate as required by generally accepted accounting principles (GAAP), then the SDIs making and purchasing liar’s loans would have reported currently (and truthfully) that they were losing money when they purchased or originated liar’s loans because very large losses were guaranteed to occur down the road. The loans guaranteed the creation of intense adverse selection and the “expected value” of mortgage lending under adverse selection is sharply negative – in plain English, the entity making or purchasing the liar’s loans will lose large amounts of money. (The reader may wonder why the seller of fraudulent liar’s loans would lose money. The answer is that sales of liar’s loans were overwhelmingly made under “representations and warranties” (reps & warranties) that would (and are) allowing the purchaser to “put” the fraudulent loans back to the originator.)