Moral Hazard: A Primer

“If you give someone the opportunity to manipulate, they’ll do it. […] If you’re not being detected you will continue – it’s human nature.” Nick Leeson, cited in Mirror, 24 February 2015.

The term moral hazard originated in the insurance business. It was a reference to the need for insurers to assess the integrity of their customers. When modern economists got ahold of the term, the meaning changed. Instead of making judgments about a person’s character, the focus shifted to incentives. For example, a fire insurance policy might limit the motivation to install sprinklers while a generous automobile insurance policy might encourage reckless driving. Then there is Kenneth Arrow’s original example of moral hazard: health insurance fosters overtreatment by doctors. Employment arrangements suffer from moral hazard, too: will you shirk unpleasant tasks at work if you’re sure to receive your paycheck anyway?

Moral hazard arises when we cannot costlessly observe people’s actions and so cannot judge (without costly monitoring) whether a poor outcome reflects poor fortune or poor effort. Like its close relative, adverse selection, moral hazard arises because two parties to a transaction have different information. This information asymmetry manifests itself in two ways. Where adverse selection is about hidden attributes, affecting a transaction before it occurs, moral hazard is about hidden actions that have an impact after making an arrangement.

In this post, we provide a brief introduction to the concept of moral hazard, focusing on how various aspects of the financial system are designed to mitigate the challenges it causes.

We start by stating the obvious: as an investor or a lender, if you give someone some money, you expect compensation. In assessing the prospective reward, you have to evaluate the quality of the project you consider funding as well as the likelihood that the recipient of the resources will do what is necessary to be able to repay. The first of these, your need to distinguish the quality of various undertakings or the creditworthiness of potential borrowers, occurs before the transaction and raises the problem of adverse selection.

Moral hazard is about the second: If you purchase an equity stake in a venture or make a loan to a firm, how do you know the person receiving the resources will in fact look out for your interest? If you cannot distinguish fortune from performance, the managers to whom you have given your funds will tend to run their company in the way most advantageous to them. The sep­aration of your ownership from their control creates a principal–agent problem (or, simply, an agency problem) that can cost investors dearly. Luxurious offices, corporate jets, chauffeured limousines, and generous self-awarded compensation are all testament to how unmonitored managers run firms for their personal benefit.

The most dramatic manifestation of the agency problem is outright fraud. Nick Leeson, whom we quote above, became one of the more famous rogue traders when his unauthorized derivatives trading precipitated the collapse of Barings Bank in 1995. More recently, Jérôme Kerviel was in the news when he engaged in nearly €50 billion of unauthorized trades that generated €4.9 billion of losses at Société Générale in 2008.

Then there is the spectacular case of Bernie Madoff, known as The Wizard of Lies. The dramatizations of the story reveal that Madoff had no genuine record-keeping system, but was able to sustain the fraud for decades despite a variety of fairly obvious red flags. Put simply, the principals (the investors, including some sophisticated intermediaries) failed to monitor their agent (Madoff), and the result was catastrophic. Simple mechanisms to detect such Ponzi schemes are widely available: for example, under the Investment Company Act of 1940, mutual funds must use an independent custodian to safeguard financial assets. Today, four such custodians hold a combined total of roughly $87 trillion worth of securities. Allowing Madoff’s firm to serve as its own custodian may have seemed frugal, but in reality, it helped conceal his fraud. In the end, monitoring is far less costly than catastrophe.

Having read this far, you might think that criminality is the prime source of costs arising from moral hazard. In advanced economies, however, well-defined property rights and rigorous enforcement limit these costs. Indeed, Leeson spent years behind bars, while Madoff’s sentence is 150 years in prison and $170 billion of restitution!

Instead, the key challenge facing providers of funds is how to motivate the manager of a perfectly legitimate firm to act in a manner fully consistent with the financier’s interest. Keep in mind that moral hazard is not about the expected performance of the firm or about the manager’s skills. Those information concerns occur before a transaction, so they give rise to adverse selection. As we discuss in our primer on the subject, there the typical solutions have to do with information collection, disclosure and collateral.

So, how does modern finance overcome moral hazard? Many people are familiar with the tools that insurers employ to address the moral hazard problems they face in writing fire, auto or health insurance. These include forms of self-insurance (such as deductibles and co-pays), as well as reporting requirements, to limit poor incentives.

What about overcoming moral hazard in business finance? Let’s start with equity. Suppose that a firm needs $100,000 to start operations, and that of this 99 percent comes from outside equity investors, with the remaining 1 percent from the firm’s manager. Suppose also that, if the manager succeeds, the value of the firm will double, otherwise the firm fails. Finally, suppose that there is no way to tell whether failure occurs because of poor fortune or due to insufficient managerial effort. In this example, the manager has little incentive to work hard (success leads to a gain of $1,000, while the loss is also limited to $1,000). Knowing this, potential suppliers of funds will be reluctant to invest in the first place.

Solving this incentive problem turns out to be very difficult. One approach is to tie managerial compensation to firm performance. However, it has proven too easy for managers to manipulate measures of profitability and to conceal catastrophic losses (think Enron and WorldCom). An alternative is to put the principal (the investor) inside the firm, so that they can more easily monitor the agent (the manager). For example, prior to the advent of securities regulation, corporate disclosure rules, and the SEC, financiers like J.P. Morgan would provide resources under the condition that they could place their lieutenants on the board of directors of the company. As Brad De Long describes, this approach can raise the value of a firm.

In today’s economy, the market for corporate control—in which an outside management team purchases an inefficiently managed firm—works in an analogous way. Private equity firms (such as the Carlyle Group or KKR) have systematized this approach: they purchase a firm and then install their own board members or managers. Activist investment funds that purchase big equity stakes in publicly traded firms also may engage in extensive monitoring and seek to influence management behavior.

The difficulties and complexities associated with external monitoring are an important reason that equity finance seems so costly to firms. In the case of debt finance, the moral hazard problem is easier to solve. Since repayment is fixed, equal to the loan amount plus interest, any gain or loss in the firm’s value accrues to the owner who borrowed the money. That is, there is no separation of ownership from control, limiting the conventional principal-agent problem. However, solving one problem creates another. Limited liability means that the owner cannot lose more than their initial investment, providing an incentive to borrow too much and (more generally) to take too much risk. The resulting debt overhang—in which firms with minimal equity have an incentive to gamble for redemption, rather than to recapitalize—can lead to underinvestment.

Fortunately, this form of moral hazard—the incentive for a borrower to take risks that are not in the interest of the lender—has well-known solutions. Lenders can require collateral that they can seize in the event of default. They can impose covenants that restrict behavior. And, they can require frequent performance reporting. Compared to the mechanisms for monitoring equity, these seem more effective, helping to account for a system of business finance with more debt than equity.

Moral hazard plagues public policy as well as private finance. One important and pernicious illustration is the too big to fail problem. To see this, suppose that the failure of a financial behemoth would wreck the system as a whole. Knowing that a government cannot credibly commit to let their firm fail, the managers take on risk that increases their likelihood of failure. Doing so provides upside to the owners at the potential expense of future taxpayers (who would fund the bailout). The current, still-untested solution to the moral hazard embodied in too big to fail is a complex combination of higher capital requirements, requirements for the issuance of loss-absorbing long-term debt, stress tests, living wills and other forms of enhanced disclosure. These are analogous to the self-insurance and information requirements that insurers impose on their policyholders.

How will the challenge of containing moral hazard evolve? Since the cost of monitoring is the source of the problem, it is natural to think that technology will ease it. Indeed, the ongoing plunge in information costs is fueling the potential for data analytics that can expose (and impose greater discipline) on risk-taking. Perhaps some of the students we teach will help design these better monitoring systems (and profit from doing so).

Take the apparently simple example of measuring leverage. You might think that computing the ratio of assets to equity is a simple task. For on-balance-sheet exposures, in normal times, it usually is (aside from notable exceptions, like goodwill and deferred-tax assets, where the value depends on the solvency of the firm). But off-balance-sheet exposures create leverage as well—think of AIG’s sale of $446 billion worth of credit risk protection in the mid-2000s. In the past, this has made measurement difficult. Will some combination of enhanced disclosure and better analytics help counterparties improve their ability to observe such risk-taking, allowing them to charge a commensurate risk premium that disciplines it? We will have to wait and see.

At the same time, it is easy to imagine an arms race in the use of information technology. That is, those who benefit from the upside of concealed leverage have an incentive to find new ways to camouflage it. While technology could help us expose certain actions, people who know how the monitoring mechanisms work might also figure out how to keep some critical information out of view—at least, for a while. Indeed, both Nick Leeson and Jérôme Kerviel did back-office stints that gave them insight into how compliance and risk management worked. So long as poor behavior goes without detection, as Leeson’s quote suggested, moral hazard will continue throwing sand into the gears of the financial system.