What
Happened, the Financial Crisis of 2008
|
The profound events
of 2007 and 2008 were neither bumps in the road nor an accentuated dip in the
financial and business cycles we have come to expect in a free market
economic system. This was a fundamental disruption-a financial upheaval, if
you will-that wrecked havoc in communities and neighborhoods across the
country. · We conclude the failures of credit rating agencies were essential cogs in the wheel of financial destruction. · We conclude dramatic failures of corporate governance and risk management at many systematically important financial institutions were a key cause of this crisis. · We conclude widespread failures in financial regulation and supervision proved devastating to the stability of the nation’s financial markets.
Financial Crisis Inquiry Commission Report |
Monetary policy mistakes were the necessary but not sufficient causes of this financial crisis. If social priorities were different, low interest rates set by the Fed after the recession of 2002 could just as easily have lowered the cost of capital for the entire U.S. economy, enabling long-term investments in municipal infrastructure, education, scientific research, plant and equipment, and health. In the words of a former investment banker, "People say 'Well, the Fed is to blame because there was all this loose money.' But guys who run banks are paid to be cautious when there's loose money around." 1 An analysis published in the 1/25/09 IHT, by Alan Blinder, former vice-Chairman of the Fed argues the historical case. He contends that the Crisis of 2008 did not need to happen. He cites: 1) SEC allowing sky-high leverage at securities firms 2) A subprime surge 3) Fiddling on foreclosures 4) Wild derivatives 5) Letting Lehman go 6) Inconsistency in the Troubled Asset Relief Program.
We do not think this crisis was inevitable. Mismanaged and ill-understood securitizations ("wild derivatives") enabled the globalization of investment mistakes, spreading the contagion from the originating investment banks onto the balance sheets of the world's credit-generating commercial banks. The key problems behind those unsound securitizations were problems of investment analysis, promising the precise management of risk in an uncertain world; the problems of agency, where all those along the chain of origination were dealing OPM with other people's money and not their own; and absent government. We discuss these in detail and conclude with some general observations about this crisis.
Faulty Investment
Analysis
The ancient Greeks assumed that there is an inherent order in nature. This assumption paid off spectacularly in the scientific world that we now inhabit. But, does this assumption of geometric and then mathematical order apply to human nature? Isaiah Berlin wrote extensively about the failed attempts to create just that in the totalitarianisms of the 20th century. He preferred liberal societies in a state of “uneasy equilibrium,” 2 i.e. error correction.
The idea of an “uneasy equilibrium” is antithetical to the ideal of financial economics that assumes all investors have the same information, most important - interpret it in the same way, and then act rationally. That world is naturally in a rather placid state of Gaussian equilibrium. Taleb (2007), Soros (2008), and Bernstein (1996) have written extensively to refute this key assumption, and if you just look around today you will see that they are right. What interests us here is how this Gaussian assumption became the basis of the structured financings that seized up the world’s financial system.
To start with the basics: the subprime loans, the constituents of many structured financings, were designed to fail. Collateralized term loans ordinarily consider the first source of repayment the obligor’s cash flow, with collateral protection then secondary. At a Fed conference, Gorton (2008) 3 described the national average rate of the problematic 2006 vintage of subprime mortgages. The average fixed rate of 8.5% (sic) would reset after two years to a floating rate of 6.1% (sic) over LIBOR.
No rational borrower would have agreed to pay this level of rates. Borrowers were likely simply ignorant of these terms, or they were gulled by their less than candid mortgage bankers into these contracts, with the expectation that they could refinance in two or three years at a lower rate after they had “built up” equity in the houses due to rapid price appreciation. Financings built upon such a shaky foundation would later on crumble.
The subprime were placed by the former investment banks into structured financings, whose supposed risks and returns were tiered according to the essentially Gaussian binomial formula. A crucial input to this formula was the probability (p) of a single mortgage default, the idea that you could specify the probability of a single mortgage default from two or three prior years of data when housing prices had only increased. A second crucial assumption of the binomial formula was that subsequent mortgage defaults in a large pool were independent of each other, belied by the reality that the whole subprime business was viable only if U.S. housing prices continued to increase.
Of course, they did not. House prices began to decrease in 2006; the 2007 failure of two Bear Sterns hedge funds invested in structured subprime financings began to call into question the valuation of all complicated and opaque structured financings, including those that contained derivatives and – most significant – corporate credits.
An inapplicable Gaussian (binomial) model was used to evaluate the risks of a portfolio of homogeneous subprime assets. What about the collective behavior of a large number of different market-traded assets like bonds? In 2000, David Li applied Gaussian Copulas to model the risk behavior of a portfolio of corporate bonds. The model conveniently required easily obtainable market data about corporate bonds spreads over risk-free treasuries. The model assumed that these bond spreads have statistical distributions, and that they could be characterized by some Gaussian random process (copula), related to the whole economy that could measure total portfolio risk. Using copulas it was then possible to calculate the default correlation of a portfolio, a measure of correlation among credit survival times. The lower the default correlation of the portfolio, the higher its credit rating due to diversification.
This analysis is an improvement over the simple Gaussian (binomial) case, but it still assumes that the bond yield premia over treasuries must have some statistical distribution. But we note, summarizing Kindleberger (2000 ed.), that world-wide financial markets are subject to manias, panics, and crashes on average every ten years.
In the language of financial economics, these models are not robust to regime changes, when the rules of the market change suddenly after large market drops. The fact that the government’s Troubled Asset Relief Program (TARP) is unable to value the structured financings held throughout the system hints that this problem is massive.
This is why all financial models require the use of judgment, a sense about when they are applicable and how they should be modified to actual circumstances. Since we don’t have a PhD in statistics, we do not comment further about them beyond noting that they are like highly tuned Ferraris, rather than SUVs, and should be treated as such. The fact that there are hedge funds that trade according to these models suggests, with the above qualification, that they can be useful. However, they should not be at the core of the world’s credit generating system that has to be at a minimum understandable in order to be credible.
Problems of Agency
The reader can imagine what happened when these complicated and nuanced financial tools hit the production environment of Wall Street. How did Wall Street manage to turn billions of dollars of low-rated securities into securities containing a high proportion of AAA credits, a gold standard that enabled their inclusion into the portfolios of the credit-generating banks?
Structured finance pricing is extremely complicated; a simple change in one of many assumptions or a misprogramming of a single equation can change the characterization of a portfolio entirely. A major assumption that enabled the manufacture of gold from dross was the assumption that diversified portfolios of junk with low historical statistical default correlations could be awarded high ratings. Furthermore, as we described, the actual structured financing ratings process had broken down at the deal motivated credit ratings agencies.
The entire credit process from subprime borrower to investment bank marketer was afflicted with the general problem of agency, no one along the chain of origination had any financial incentive to direct capital to its best uses except the ultimate investors in the products of structured finance. No investor, however, could possibly understand the hundred page disclosures behind every structured finance transaction, that further contained additional assets that contained additional hundred page disclosures (and misapplied financial models).
What in the world could have motivated the otherwise sophisticated investment banks to engage in behavior that would ultimately take down the entire financial system? In any financial organization, the CEO is the chief marketer and risk control officer. Reports about what happened at Merrill Lynch and Lehman relate that over the last few years, their CEOs dismantled risk controls to take advantage of perceived profit opportunities. In the mania for overleveraged lending, house prices could only increase and the self-correcting nature of markets would ensure future moderation in the world economy and the absence of severe economic downturns.
For example, to quote the 2006 Merrill Lynch Annual Report:
The risk management and control process ensures that our
risk tolerance is well-defined and understood by our businesses as by our
executive management. Independent risk and control groups interact with the
core businesses to establish and maintain this overall risk management
control process. While no risk management system can ever be absolutely
complete, the goal of these independent risk and control groups is to
mitigate risk-related losses so they fall within acceptable, predefined levels,
under foreseeable scenarios. 4
The 11/9/08 NYT, however, reports what really happened at Merrill:
(2006) was a moment to savor for E. Stanley O’Neal,
Merrill’s autocratic leader, and a group of trusted lieutenant who had
helped orchestrate the firm’s profitable but belated mortgage push. Two
indispensable members of Mr. O’Neal’s clique were Osman Semerci, who among
other things, ran Merrill’s bond unit and Ahmass L. Fakahany, the firm’s
vice chairman and chief administrative officer…. Former executives say Mr. Fakahany had weakened
Merrill’s risk management unit by removing longstanding employees who
“walked the floor” talking with traders and other workers to figure out
what kinds of risks the firm was taking on. Former Merrill executives say that the people chosen to
replace those employees were loyal to Mr. O’Neal and his top lieutenants.
That made them more concerned about achieving their superiors’ profit
goals, they say, than about monitoring the firm’s risks.
It would have been too much to ask for a fiduciary standard; but it was reasonable to expect that the investment banks have a commercial standard, not selling bad merchandise to their customers and thus jeopardize their futures.
So why did the investment banks spread their complex toxic tranches around the world? The simplest explanation is that they initially didn’t think their creations were toxic. There was a big demand for high and safe returns (sound familiar?); structured finance calculated that these investments could produce a precise tradeoff between minimal risk and higher returns, all those formulas and Greek letters proved it; and most important, management was rewarded billions of dollars in immediate compensation for doing this business 5. These were probably the reasons management did not inquire further.
Absent Government
This leads to a discussion of the third factor that could have braked this mania, the government.
The U.S. conservative view that people ought to be able to act freely in markets under conditions of minimal government proceeds from the major liberal Enlightenment assumption of a prudential human nature. Reinhold Niebuhr wrote:
There is a specially ironic element in the effort of the
seventeenth century to confine man to the limits of a harmless “nature” or
to bring all his actions under the discipline of a cool prudence. For while
democratic social philosophy was elaborating the picture of a harmless
individual, moved by no more than a survival impulse, living in a social
peace guaranteed by a pre-established harmony of nature, the advancing
natural sciences were enabling man to harness the powers of nature, and to
give his desires and ambitions a more limitless scope than they previously
had. 6
The U.S. was not the world’s first modern democracy, but it was the world’s first large-scale democracy. Although motivated by the ideals of the Enlightenment, the designers of the Constitution had a healthy appreciation of human realities and designed the State according to a system of checks and balances. In Federalist Paper No. 51, James Madison wrote:
…what is government itself, but the greatest of all
reflections on human nature? If men were angels, no government would be
necessary. If angels were to govern men, neither external nor internal
controls on government would be necessary. In framing a government which is
to be administered by men over men, the great difficulty lies in this: you
must first enable the government to control the governed; and in the next
place oblige it to control itself. A dependence on people is, no doubt, the
primary control on the government; but experience has taught mankind the
necessity of auxiliary precautions. This policy of supplying, by opposite and rival
interests, the defect of better motives, might be traced throughout the
whole system of human affairs, private as well as public. 7
Large-scale democracies and large-scale markets are almost perfect analogs.What was the role of government in this crisis? In testimony before congress on 10/23/08, former Federal Reserve chairman, Alan Greenspan, said:
I made a mistake in presuming that the self-interest of
organizations, specifically banks and others, was such that they were best
capable of protecting their shareholders. 8
Government loosening its regulation of the commercial banks, the SEC’s approval of much higher securities firm leverage in 2004, and the attempts of government sponsored FNMA and GNMA to “remain relevant,” by purchasing subprime loans all resulted in the encouragement of excess in the private sector, rather than in its regulation.
All manias, panics, and crashes have unique features. We are now living with the consequences of three major mistakes: faulty investment analysis, problems of agency, and absent government. We have mainly discussed these issues at a level of detail because, as Pimco’s Mohamed-El-Erian said “this is a crisis of the system,” many of whose components need repair.
In previous economic downturns, the Fed had merely to pursue a counter-cyclical monetary policy that would lower interest rates and thus increase economic activity. Monetary policy worked through a functioning financial system. Now, significant problems mean that rebuilding will take time:
1) Modern economies are not possible without functioning financial systems. On 12/07 the book net worth of U.S. financial companies was around $1.7 trillion. The IMF currently estimates U.S. company writeoffs around $1.1 trillion, about 50% of the total. The necessary level of replacement capital certainly exists in the U.S. economy, but the incentives for new private investment have to exist. New credit will have to flow from government involvement in both the commercial banks and simpler market securitizations.
2) On 2/10/09 the Treasury announced a plan to get toxic assets off bank balance sheets by sponsoring a joint public-private sector fund, giving the private sector incentives for price discovery and investment. We have mentioned a theoretical problem with many structured finance valuations; individual bond spreads must have actual statistical distributions. What if they don't, and those financings therefore can't be priced? Those toxic assets have to be removed now from the banking system, one way or another, to be valued if possible in a more normal economy.
3) Traditional banking matched savings with borrowing by directing capital to the growing parts of the economy, a role that was increasingly taken over by market securitizations. But banks are necessary as a funding source because, as the funding manager for a N.Y. money center bank said of the retail branches, “They are always there.” But the credit generating banks have to be simplified. The systems tradeoff is apparently between the potential reliability of banks and the lower immediate costs (and usually creativity) of markets.
4) Government has to rebuild a lapsed regulatory system.
In every mania, the object of speculation differs. In 1636 it was tulip bulbs; in 1929 it was stocks; and this time it was risk itself - the idea that risk could be managed with precision. The 2/2/09 New Yorker Magazine wrote, “For reasons to do with a lack of historical awareness, overconfidence, and faulty mathematical modeling, the entire global financial system was built on mistaken calculations about probability.” To also quote a director of the Bank of England, “Enthusiasm about return gave way to hubris and a collective blind eye was turned to the (real) resulting risk.” 9
We have described the present condition of the financial system. Rebuilding will take a great deal of skill, time, and money, reforming a system that had profitably grown for decades and then collapsed by its own unregulated excesses. Markets are like democracies; they are the best of imperfect systems.
__
We have described conditions in the supply chain that manufactured toxic securities. Global trade imbalances led to the demand for these securities, as this article well describes. Imbalances cause recessions.
It is said that regulators cannot foresee bubbles and deal with them. The former may be true, but the latter not. When asset classes, unlike trees, start to grow the sky; it is assured that people are reaching unduly for profit. In spite of the enthusiasm, a very detailed and critical analysis of that asset class is then in order. For example, when a bank starts to expand its loans by 14%+ per year (after including all items) that is a sure sign of future trouble.