The recent financial crisis had several causes: poor regulatory oversight, inadequate modelling of risks, and inefficient institutions for trading and managing complex financial instruments. An important aspect of the institutional deficiencies was how rewards and risks were shared. Financial traders, salespeople and portfolio managers have a bias towards risk-taking because of their compensation schemes, which provide large rewards for large successes, but no commensurate penalties for large failures. The linking of rewards to relative performance exacerbates this problem. Another factor that makes things worse is the possibility of timing exit from a high-reward, high-risk situation so that someone else is left “holding the bag”. Financial sector compensation schemes have certainly been receiving much attention, because of their disconnect from performance.
The above explanations are based on conventional models of rational behaviour. However, as we now know from a large body of work in behavioural economics, finance and psychology, people exhibit systematic deviations from standard models of rationality. These include compartmentalization, over- or under-estimation of probabilities, myopia, and so on. In the context of systemic risk and propensity to crises, hubris, or “overbearing pride or presumption”, could be an important issue to think about for regulatory redesign efforts. Hubris was used in academic work in the 1980s to explain puzzles in corporate merger and acquisitions, referring to managers’ confidence in their own valuations of firms, over those of the market. There was no specific model of hubris—it was simply a residual explanation once all others were eliminated by analysis of the data.
A formal behavioural analysis of hubris might be used to guide regulatory design, since it can be a critical factor in driving market dynamics to situations where systemic risks are unacceptably high. Hubris is the centre of Tom Wolfe’s “Masters of the Universe” in The Bonfire of the Vanities, and Michael Lewis’ “Big Swinging Dicks” in Liar’s Poker, two popular accounts of the culture and conduct of Wall Street. Indeed, hubris might be the generator and driver of “irrational exuberance”. Two components of any analysis are straightforward, and more general than the concept of hubris: distortion of probability estimates by decision-makers in financial markets, and biases in the incorporation of information by those decision makers. The third component draws on my work with Daniel Friedman, which applies the idea of emotional state dependent utility components (Esducs), where certain types of situations may trigger changes in the utility function, which in turn have short-and long-run implications for market or social equilibrium. Our analysis considered a simple game of trust, where defection from the cooperative action by one person (violating the trust placed by the other) triggers a change in the other’s utility function: The aggrieved person now gets satisfaction from taking vengeance on the violator, even if at a personal cost.
Similarly, certain kinds of market dynamics inherently lead to hubris—Esducs that distort the probability estimates and/ or information weighting of successful market participants. In turn, behaviour driven by hubris further increases systemic risk, precisely in those market situations in which that risk is greatest. In this case, regulations that place limits on certain types of market behaviour, or that require disclosure, which would be sufficient in normal times, may not be adequate. Instead, regulations may need to be state-dependent or contingent as well.
The contingency idea is not new. For example, Raghuram Rajan (The Economist, 8 April) discusses the problems with counter-cyclical bank capital requirements—one idea for contingent regulations. He argues that banks may shift activities off-balance sheet (he refers to “euphoria” as a driver, where I suggest hubris may be more fundamental). His own idea is for banks to have contingent capital—debt that converts to equity when indicators for the individual institution or the system suggest that trouble may be coming.
Modelling behavioural factors such as hubris may allow an additional line of approach to contingent regulation, using market activity indicators to trigger certain kinds of brakes, or additional depth of monitoring. It is easier to observe and quantify indicators of activity in specific markets (open interest for futures, short interest for stocks, volatility indices, and so on) than it is to detect a crisis that triggers special debt-to-equity conversions. Theoretical analyses of human behaviour cannot be immediate guides to determining regulatory triggers, but they can help in figuring out what those triggers might be, so that looking at data is more than just reading tea leaves.
Clearly, there is no unique solution to the problem of regulation of complex financial systems that are subject to large endogenous risks. But understanding the behaviour of market participants, and systematic biases that are triggered in certain risk situations, as well as exacerbating those risks, can be important for designing robust, proactive regulatory schemes.
Nirvikar Singh is professor of economics at the University of California, Santa Cruz. Your comments are welcome at firstname.lastname@example.org