Moral hazard and the financial crisis.

• Author: Dowd, Kevin
• Position: Report

There is no denying that the current financial crisis has delivered a major seismic shock to the policy landscape. In country after country, we see governments panicked into knee-jerk responses and throwing their policy manuals overboard: bailouts and nationalizations on an unprecedented scale, fiscal prudence thrown to the winds, and the return of no-holds-barred Keynesianism. Lurid stories of the excesses of "free" competition--of greedy bankers walking away with hundreds of millions whilst taxpayers bail their institutions out, of competitive pressure to pay stratospheric bonuses and the like--are grist to the mill of those who tell us that "free markets have failed" and that what we need now is bigger government. To quote just one writer out of many others saying much the same, "the pendulum will swing--and should swing--towards an enhanced role for government in saving the market system from its excesses and inadequacies" (Summers 2008). Free markets have been tried and failed, so the argument goes, now we need more regulation and more active macroeconomic management. (1)

Associated with such arguments is the claim that the problem of moral hazard is overrated. A prominent case in point is Lawrence H. Summers himself. In a widely cited column, he exhorted his readers to beware of a "moral hazard fundamentalism" which, he argued, was "as dangerous as moral hazard itself" (Summers 2007). His use of the disparaging term "fundamentalism" suggests that he did not intend it as a compliment. But whatever his intent, the issue identified by Summers--the role of moral hazard--is central to the controversy over the causes of the present crisis and the lessons that should be drawn from it. Unlike him, however, I believe that moral hazard is a (much) underrated problem: moral hazard played a central role in the events leading up to the crisis, and we need to appreciate this role if future reforms are to be well designed and prevent further disasters down the line. Understanding moral hazard is fundamental to understanding how the economy works--and if this is "moral hazard fundamentalism," so be it.

The Nature of Moral Hazard

A moral hazard is where one party is responsible for the interests of another, but has an incentive to put his or her own interests first: the standard example is a worker with an incentive to shirk on the job. Financial examples include the following:

* I might sell you a financial product (e.g., a mortgage) knowing that it is not in your interests to buy it.

* I might pay myself excessive bonuses out of funds that I am managing on your behalf; or

* I might take risks that you then have to bear.

Moral hazards such as these are a pervasive and inevitable feature of the financial system and of the economy more generally. Dealing with them--by which I mean, keeping them under reasonable control--is one of the principal tasks of institutional design. In fact, it is no exaggeration to say that the fundamental institutional structure of the economy--the types of contracts we use, and the ways that firms and markets are organized--has developed to be the way it is in no small part in response to these pervasive moral hazards.

Subsidized Risk-Taking: Heads I Win, Tails You Lose

Many of these moral hazards involve increased risk-taking: if I can take risks that you have to bear, then I 'nay as well take them; but if I have to bear the consequences of my own risky actions, I will act more responsibly. Thus, inadequate control of moral hazards often leads to socially excessive risk-taking--and excessive risk-taking is certainly a recurring theme in the current financial crisis.

A topical example is the subprime scandal. In the old days, a bank would grant a mortgage with a view to holding it to maturity. If the mortgage holder defaulted, then the bank would usually make a loss. It therefore had an incentive to be careful who it lent to and prospective borrowers would be screened carefully: a subprime would-be borrower didn't have much chance of getting a mortgage. However, if a bank originates a mortgage with a view to selling it on (i.e., securitizing it), this incentive is seriously weakened. In fact, if it sells on the mortgage to another party it has no interest in whether the mortgage defaults or not, and is only concerned with the payment it gets for originating the loan. The originating bank is now happy to lend to almost anyone, and we end up in the patently unsound situation where mortgages are being granted with little or no concern about the risks involved. On this basis,

even the doziest mortgage broker can originate subprime mortgages for even the least creditworthy borrowers. The fact that the borrowers are incapable of making payments on the mortgage will magically be priced into the mortgage by the securitization process, which will bundle the mortgage with other mortgages originated by a similarly lax process and sell the lot to an unsuspecting German Landesbank attracted by the high initial yield. Everyone will make fees on the deal, everyone will be happy [Hutchinson 2008a].

Unfortunately, this giant Ponzi scheme could keep going for only as long as house prices continued to rise and new entrants continued to come into the market. Once interest rates started to rise and house prices started to fall, then the supply of suckers inevitably, dried up and the whole edifice began to fall in on itself.

A second example is what the BBC's Robert Peston christens the "greed game." The partners of private equity and hedge funds would invest their backers' funds on a compensation arrangement that typically gave them 20 percent of gains made (plus a 2 percent annual management charge); any losses, however, were shouldered by the backers alone. Investments were then leveraged by enormous borrowing. As Peston (2008) explains,

Thus, if a private-equity firm or hedge fund generates a capital gain of 1bn £--and in the boom conditions of the past few years, that wasn't unusual--the partners in the relevant fund would trouser 20 percent, or 200m £. But if there was a loss of 1bn £, well only the backers would lose. Backers were willing to go along with these generous terms because the funds had generated good returns for many years. These remuneration packages prompted an exodus of (real or imaginary) talent from the banks into the funds, and the banks responded by adopting similar practices themselves. Fund managers and bankers then took much greater risks than they would had their own money been at stake. "But with none of their own money on the line and the potential to generate colossal bonuses," Peston goes on to explain, many were seemingly seduced by their own propaganda: they apparently

believed that structured finance was revolutionary financial technology for transforming poor quality loans into high quality investments. There was an epidemic of Nelsonian Eye Syndrome on Wall Street and London. And bankers, privateequity partners and hedge-fund partners acknowledge--or at least some do--that the cause was good, old-fashioned greed induced by a turbocharged remuneration system that promised riches in return for minimal personal risk [Peston 2008].

Note too that, once paid, bonuses are typically not recoverable later. This absence of any deferred compensation gives fund managers an incentive to focus only on the period to their next bonus. If the fund makes losses later, then that is not their concern (or, of course, their fault). The absence of deferred remuneration thus institutionalizes short-termism and undermines the incentive to take a more responsible longer-term view.

Yet the subprime scandal and the greed game are merely illustrative of a much broader and deeper problem--namely, that moral hazard in the financial sector has simply been out of control. As Martin Wolf (2008a) aptly put it, no other industry but finance "has a comparable talent for privatising gains and socialising losses." Instead of "creating value," as we were repeatedly assured, the practices of financial engineering (including structured finance and alternative risk transfer), huge leverage, aggressive accounting' and dodgy credit rating' have enabled their practitioners to extract value on a massive scale--to walk away with the loot, not to put too fine a point on it--while being unconstrained by risk management, corporate governance, and financial regulation, all of which have proven to be virtually useless. We therefore need to ask why the various "control systems" failed so dismally.

The Failure of Financial Risk Management

The first question is what went wrong with financial risk management. The answer is a complex and multi-layered one. At the most superficial level, practitioners of modern quantitative risk management all too often make a range of inadequate assumptions: they assume that financial risks follow Gaussian distributions (and so ignore the "fat tails" which really matter); they assume that correlations are constant (and ignore the fact that correlations tend to radicalize in crises and so destroy the portfolio diversification on which a risk management strategy might be predicated); and they make assumptions about market liquidity that break down when they are most needed. Many risk models and risk management strategies also ignore strategic or systemic interaction: this is comparable to a cinema-goer who thinks he can easily get to the exit in the event of a fire, ignoring the likelihood that everyone else will be running for it as well. These and other common modelling errors lead to risk models that are focused far too much on the "normal" market conditions that do not matter at the expense of ignoring the abnormal market conditions that do. This leads to the somewhat worrying conclusion that the practice of what passes for risk management might actually be counterproductive and leave the financial system more rather than less exposed in a crisis

Then there are the problems with the...