Monday, May 20, 2013
What Caused The Crash: An Autopsy
Certainly, one of the crucial questions for an understanding of our financial system and its vulnerabilities is: Who was responsible for the financial crisis?
The immediate cause of the financial crisis was that about a trillion dollars worth of subprime mortgage-related securities, which had been carried at par, became unsellable and unpriceable almost overnight. The entire global financial system froze up, because every major bank was exposed and no bank’s true solvency could be known (not by anyone). The financial system shut down, credit growth went into reverse, and the household debt bubble began to deflate. Nominal GDP declined for the first time since the Great Depression.
The crisis began in mid-2007, when delinquencies on recent-vintage subprime mortgages began to rise to frightening levels. Overnight, these securities went from liquid commodities to unsellable. Coincidentally, at this very moment the accounting profession was implementing a stricter version of fair-value accounting (FAS-133), which made it more difficult to carry illiquid securities at their “intrinsic value” (i.e., at par). Instead of simply marking their subprime securities to their internal models, as banks had been doing, they were forced to mark to a distressed market where there were no buyers. Accounting losses began to be realized in 2007-08, fitfully and unequally. No one wanted to believe the marks they were getting, and so some of them fudged them to one degree or another.
These problematic subprime securities took two forms: subprime Residential Mortgage-Backed Securities (RMBS), and subprime Collateralized Debt Obligations (CDOs). Prices for RMBS fell rapidly, but the securities were still somewhat analyzable, and were not a complete black box (although no one could predict where the delinquencies were going). However, the subprime CDOs were repackagings of RMBS that were consequently complex and opaque to potential bidders, and thus impossible to value or to sell. Most of these securities were unregistered, and in any case lacked sufficient disclosure for an arms-length investor to make an informed valuation. There was no functioning market for the subprime CDOs.
There were two principal engines at work during the subprime era: (1) purchase and warehousing of mortgages for securitization into RMBS; and (2) repackaging of RMBS into CDOs. The CDOs were ostensibly underwritten to a high credit standard, and were also wrapped by AAA-rated guarantors (e.g., AIG Financial Products). These two motors ran smoothly and very profitably for a few years, but then the pipes began to back up in late 2006 and early 2007, as the securities began to perform badly.
The combination of filled-up concentration limits at the guarantors coupled with increasingly disturbing mortgage delinquencies meant that subprime RMBS and CDOs could no longer be sold to investors, even the unsophisticated “real money” investors. This would have been a good time for Wall Street to turn off the securitization motor and seek greener pastures. However, these two businesses were very profitable, and a lot of people on the Street depended on this business for their livelihoods. They couldn’t turn off the motor without putting themselves out of a job (and forfeiting more “profits”). So, no longer able to sell the products of their mortgage mills, some firms decided to take them into inventory “temporarily”. They judged them to be of low risk and requiring a minimal capital coefficient. This was the original sin that caused the crisis.
These firms willy-nilly began to expose themselves, their shareholders and their creditors to tens of billions of subprime mortgages and their toxic derivatives. This was the equivalent of “full speed ahead” when the Titanic entered the iceberg field. These CEOs exposed their firms to billions in economic losses in order to make millions in ephemeral accounting profits. This was a catastrophic management failure.
Over time, the markets were able to decide which banks were the most exposed to subprime (net of effective hedging), and began to reduce their counterparty limits and to raise their collateral standards. The most suspect firms (Bear Stearns, Lehman Brothers, and Merrill Lynch) were increasingly squeezed. The pressures from the accountants were unabated, which began to force these firms to assess the size of their exposures and the marks at which they were carrying them.
Some Wall Street executives were dumbfounded when they they were ultimately confronted with their firm’s subprime exposures. They were surprised by both the size of the positions and the potential harshness of the marks. They went through the five stages of grief, but they could not change the facts. The credibility of their financial reporting began to suffer as their reported losses seemed puny in relation to the rumored size of their positions. In view of subsequent developments, those reported losses appear in some cases to have been heroic in their wildly optimistic assumptions. But the markets didn’t believe the financial reporting, and the withdrawal of credit and liquidity to the contaminated names continued.
The first to fall was Bear Stearns, which lost access to the credit markets in early 2008. It’s saga of mistakes has been well-documented; it appears that no one was fully in charge. The next victim was Lehman. The markets knew that, in addition to having a large inventory of subprime securities, Lehman also had a large and illiquid portfolio of commercial real estate (just as it had in 1998, when it almost failed). Other firms did the math and concluded that Lehman’s solvency was problematic. But what happened next was completely unexpected: the secured repo market closed to Lehman regardless of the quality of the offered collateral or the margin ratio. No one wanted any exposure to Lehman’s name, no matter how minimal the risk. This was never supposed to happen.
Prior to that event, the brokers’ liquidity stress-scenarios had them move out of the confidence-sensitive unsecured markets and into the secured markets with quality, pledgeable collateral. That was considered to be an open, liquid market for any name with the collateral. However, in 2008 the secured markets (their risk committees) decided that no collateral was good enough. What this meant was that the Wall Street broker/dealer business model became obsolete overnight. A standalone broker/dealer financing itself in the secured markets without access to the Fed discount window was not longer a viable business. This caused the run on Merrill, Goldman and Morgan Stanley, which forced the “Wall Street Bailout” (and the conversion of these firms into regulated bank holding companies). “Wall Street”, as we knew the securities industry, had ended.
Who was responsible for this unfortunate series of events? Who could have prevented it by acting differently? I have come to the conclusion that the culpable parties are the CEOs of the Wall Street firms that overexposed themselves to large risks (RMBS, CDOs, commerical real estate) and lost the ability to finance themselves in the market.
Not only did these firms originate many billions of securities that did notpeform as predicted, but they invested their own money in them, in the tens of billions of dollars. The whole point of the “originate-to-sell” model is to get rid of the product long before it becomes shopworn. They were so eager for imaginary short-term profits that, not only did they originate bad securities, but they also agreed to hold onto them so that their mortgage securitization machines could keep making “profits”.
The literature of the crisis is rich with other suspects, such as the mortgage originators, the rating agencies, Barney Frank (Fannie & Freddie), Bob Rubin (repeal of Glass-Steagall), the GOP (failure to regulate derivatives), the SEC (which eliminated traditional leverage limits), not to mention the millions of borrowers who falsified the financial information on their loan applications and the mortgage brokers who enabled them. But these suspects didn’t cause the financial system to collapse, and could not have prevented it. That responsibility rests with the firms that required rescue. (The managements of the GSEs made collosol mistakes, but they were partly acting as agents for the government and its imprudent housing policies.) Evidence that these firms were not fated by history to fail is the fact that similar firms survived and not just because of TARP; they survived because they controlled their risks better.
What were these CEOs’ sins? I think that their sins can be summarized in three words: poor risk management. A Wall Street firm is a dynamic cauldron of risks of every imaginable description. Every Wall Street firm depends upon the robustness of its risk management systems, risk discipline and risk culture. Wall Street risk management departments are locked in perpetual war with their firms’ profit-making businesses, which could make much more money if they weren’t handcuffed by stupid low-paid risk managers who don’t understand the business.
There has never been a Wall Street firm which did not claim to have “world-class” risk management and a lot of PowerPoint slides to prove it.
But in point of fact, not all risk managements are created equal. Some are very good and some not so good. It is easy to tell the difference after the fact: the risk management at firms that do well is good; the risk management at firms that fail is not so good. A Street firm has scores of businesses taking risks every day, and not one of those risks is supposed to be able to crater the firm. If that happens, then mistakes were made. And don’t refer to the “hundred year storm” to excuse your mistake, if other firms with better risk-management made it through the storm.
The big mistake, the original sin at the firms that crashed, was allowing their securitization operation to warehouse subprime mortgage-related assets that they couldn’t sell. This did not happen at every firm. But it did happen at a handful of the biggest firms, causing the financial crisis. Because they judged these exposures as “low risk”, they allowed themselves to hold tens of billions worth, financed with unsecured short-term debt. At the firms that actually did have world-class risk management, their (non-negotiable) risk concentration limits forced them to either stop originating, or to sell or hedge their positions.
It is possible that the decision to fill up on subprime assets was made in good faith, on the assumption that they really were low-risk, liquid assets requiring a minimal capital coefficient. If so, then the verdict is incompetence. But in fact, I think that the basic issue was cultural: the weak power position of the risk management department in the firms’ management and culture.
In the old days, before these firms went public, they were playing with their own money and had an incentive structure that demanded competent and powerful risk management. In fact, the top executives were the most aggressive risk managers, walking among the bond traders, looking at their positions, and ordering them to sell some or all of their ugliest positions. A bond was carried at what somebody else would pay for it (or less), not what the desk said it was worth.
That culture faded away at many firms after they went public. The top executives were on the executive floor, engaging in corporate-level activities, instead of walking the trading floor. In some cases, the top people were not ex-traders, and only knew the rudiments of the bond business.
I think it is fair to say that the common denominators at the firms that blew up were (1) the failure by top management to know and understand the scale of the firm’s bet on subprime mortgages, (2) over-reliance on the business heads to police their positions, and (3) failure to give the professional risk managers sufficient power in the firm such that they could impose limits on profitable business heads.
By contrast, at the firms that did not crater, the CEOs were more familiar with their firms’ positions, risk management was central to the culture, and large concentrations were actively managed. I remember a Wall Street executive saying that his firm’s business was “long volatility”, meaning that the firm would make money in bad weather and lose money in good weather. That is a good description of the proper risk management philosophy: don’t make big directional bets, just position to make money when things get bumpy. Easier said than done; but a good way of explaining how a risk-taking firm can stay in business with finite capital resources.
Only one of these mismanaged firms went bankrupt; only one of the CEOs had to pay the price of seeing his company disappear, and to be pilloried in public. The others were rescued, and never had to stand trial before the court of public opinion. In other words, if the world were fair they would have gone bankrupt (and we can thank God that the world is not fair).
The villains of the books and movies about the crisis are those who were actually heroic: Tim Geithner, Hank Paulson, Ben Bernanke, George Bush, Barack Obama, and the congressmen and senators who voted for the TARP to save the financial system.
What are some of the lesson to be learned from this story? One lesson is that if you don’t understand the bond business, you shouldn’t be running a bond firm. Another is that if you have a large position which would wipe you out if it went bad, sell it. These are lessons about stewardship. The shareholders didn’t hire you to maximize profits, they hired you to create long-term shareholder value. Only three Wall Street CEOs sailed into and out of this crisis with their jobs, companies and reputations somewhat intact: Jamie Dimon, Lloyd Blankfein and John Mack. That is not an accident of history.
A Defensive Note on the Rating Agencies
There has never been and there never will be an institutional investor who says that he relies on bond ratings to make his investment decisions. Not only are institutional investors required by the SEC to do their own analysis, but they represent to their clients that they do their own analysis. “The rating agencies made me do it” is not a credible excuse for an unsuccessful institutional portfolio manager.
It is true that bond ratings can provide a lazy investor with a false sense of security, and there have always been investors whose “strategy” was to buy the highest-yielding bond in a specified rating category. But fiduciaries are expected to act like fiduciaries, and bond ratings are only a market opinion to be considered along with other sources of information (such as reading the offering documents). During the bubble, the Street would demand a purchase decision from investors before any disclosure was ever provided, even though selling a bond without adequate disclosure is what the Securities Acts were supposed to prevent. The fact that most securitized product was unregistered reflects poorly on those managers who bought them, on their clients who let them buy them, and on the SEC for letting them be sold. To my knowledge, this Grand Canyon-sized loophole has not been closed.
But, here is the most devastating defense of the rating agencies in the context of the Wall Street crisis: These firms didn’t buy rated securities that went bad; they originated the securities that killed them. They had the complete documentation, the loan-by-loan data, the due diligence reports. They became sick from eating their own cooking.
Further, Wall Street sold RMBS and CDOs to gullible institutional investors on the basis of reps and warranties that did not reflect the underlying collateral. Disclosure was minimal, inaccurate and too late. These institutional investors performed no analysis on their own, despite their representations to investors that they did perform their own analysis. Rating agencies relied on these same misleading representations in forming their opinions about the quality of the securities.
If Congress wants to fix this problem, it should require that: (1) all securities sold in the US to more than one buyer must be registered with the SEC; (2) registration and disclosure must precede the selling process; and (3) fiduciaries must warrant and be able to document that they have performed an independent analysis of such disclosures.
Reliance on third-party opinion such as ratings or analysis provided by the underwriter is no substitute for independent credit analysis performed by the fiduciary itself.
Public Policy Lessons
What are the public policy lessons from the crisis in general, and these firms in particular?
Clearly one lesson is that any financial institution that is too big to fail (Systemically Important Financial Institution, or SIFI) should be regulated like a bank holding company. This was included in Dodd-Frank and is essential. No more unregulated AIGs or GE Capitals.
A second lesson is that bank regulators should not outsource capital adequacy regulation to the regulated firms. Models should be imposed on the firms, not matter how blunt and how “inappropriate”. There should no longer be any discussion about how competent financial regulation hurts American competitiveness and puts American firms at a disadvantage. If you come under the safety net, then you forfeit your free-market rights. SIFIs are a contingent liability of the government, and as such they have no rights, and their “competitiveness” does not even enter into the equation. Failed banks are not “competitive”.
A third lesson is that evaluation of management competence must be a component of financial regulation. The CEO and the chairman of the audit committee of the board must be able to explain the firm’s positions and exposures in detail, and to be familiar with the work of the risk management department.
A fourth lesson was supposed to have been learned with Enron/WorldCom: boards must be held accountable for the failure of their stewardship. This wasn’t true before Sarbanes-Oxley, it wasn’t true after Sarbanes-Oxley, and it’s not true now. Board incompetence at SIFIs endangers financial stability and the taxpayer, and should be regulated by the Fed. Failures of oversight resulting in catastrophe must have sanctions, such as “disbarment” as a director of a SIFI, or fines. Enron was a private enterprise that was not too big to fail: it’s failure had minimal domino effects and it wasn’t systemically significant. That is qualitatively different from Citigroup, which comes under the safety net, and whose directors play a quasi-public role.