Arrow on the Financial Crisis

I'm currently teaching Arrow's work on risk and uncertainty in a graduate course. Here's an editorial from the Guardian by Kenneth Arrow, winner of the 1972 Nobel prize in economics and truly one of the ground-breakers in modern economics. Here's he provides his insights on the current financial crisis.

The current financial crisis, the loss of asset values, the refusal to extend normally-given credit and the great increase in defaults on obligations ranging from individual mortgages to the debts of great investment banks presents, of course, a pressing challenge to the fiscal authorities and central banks to take measures to minimise the consequences. But they also present a challenge to standard economic theory, a challenge all the more important since the development of policies to prevent future financial crises will depend on a deeper understanding of the processes at work.

That economic decisions are made without certain knowledge of the consequences is pretty self-evident. But, although many economists were aware of this elementary fact, there was no systematic analysis of economic uncertainty until about 1950. There have been two developments in the economic theory of uncertainty in the last 60 years, which have had opposite implications for the radical changes in the financial system. One has made explicit and understandable a long tradition that spreading risks among many bearers improves the functioning of the economy. The second is that there are large differences of information among market participants and that these differences are not well handled by market forces. The first point of view tends to argue for the expansion of markets, the second for recognising that they may fail to exist and, if they do come into being, may fail to work for the benefit of the general economic situation.

The value of spreading risks has, of course, been recognized as the basis of conventional insurance as well as the issue of company shares that spread corporate risks widely. The central element of standard economic analysis since the 1870s has been the concept of general economic equilibrium, which, under competitive conditions, leads to an optimal allocation of resources. In the 1950s, it was shown how to incorporate uncertainty into general equilibrium, which suggests, at least, that increasing the number and coverage of risk-bearing instruments would improve the running of the economy. Not only would risks be more efficiently borne, but, more importantly, additional socially valuable risky enterprises would be undertaken. Research showed how derivative securities should be priced, how individuals should choose portfolios to minimise their variability, and how individual contracts, such as mortgages, could be bundled so as to distribute the risks for different parts of the market with different risk tolerances.

The second strand of analysis was a growing recognition of the importance of information in governing reactions to uncertainties. If individuals in the market have different degrees of information, the ability to create securities or engage in other forms of contracts becomes limited; the less informed understand that the more informed will take advantage and react accordingly. This situation was long recognized by insurance companies under such terms as, "moral hazard" (when the insurer cannot tell how well the insured is avoiding risks) and "adverse selection" (when the insurer cannot distinguish among differentially risky insured, so that, at any given premium, the more risky insure themselves most extensively). Economists began to realise that "asymmetric" information was the key to understanding the limits of health insurance and the incentive problems of socialism and then that these concepts found their most important application in financial markets, precisely in the complex securities that the first strand of analysis had called for.

There is obviously much more to the full understanding of the current financial crisis, but the root is this conflict between the genuine social value of increased variety and spread of risk-bearing securities and the limits imposed by the growing difficulty of understanding the underlying risks imposed by growing complexity.