Streetwise Professor

December 8, 2008

Are Financial Crises Optimal?

Filed under: Derivatives,Economics,Politics — The Professor @ 11:23 pm

It is almost axiomatic that financial crises in general, and this financial crisis in particular, represent market failures that require government regulation to correct.   But this conclusion is logically defective for two reasons.   First, as hard as it might be to believe, financial crises can occur in a first-best, or constrained first best equilibrium.   Second, because there is no guarantee that governments that are subject to the same information constraints that likely make financial crises possible, and which suffer other defects in addition, will make things better rather than worse.

These points are made quite well in a book providentially published in 2007, Understanding Financial Crises, by Franklin Allen and Douglas Gale.   (I don’t know Douglas Gale, but I can say that Franklin Allen is not only a great economist, he is one of the true gentlemen of the profession–a truly nice man.)   I imagine the book would be rough sledding for non-economists (but, not as hard as the even more technical papers upon which it is based), but it does lay out in a systematic and persuasive way some relatively straightforward models of financial crises and contagion.

In several of the models, financial crises–such as widespread bank failures–occur even with optimal contracts.   A couple of extended quotes are worth reading:

[W]e want to challenge the conventional wisdom that financial crises are always and everywhere a ‘bad thing.’   It may well be true that financial crises impose substantial costs on the economy . . . . At the same time any regulation of the financial system involves costs.   The most important of these costs are the distortions imposed on the financial system by a regulatory regime that restricts what financial institutions may and may not do.   To measure the costs and benefits of any policy, we need to have a clear understanding of the conditions for efficiency in the financial system, including the conditions for efficient financial crises. (p. 153)

As long as we have complete markets for hedging aggregate risk and intermediaries can use complete contingent risk-sharing contracts, the equilibrium in a laisser-faire economy will be incentive efficient.   If intermediaries are forced by transaction costs to use incomplete contracts, the equilibrium will be constrained efficient.   In either case, it is wrong to suggest that financial crises constitute a source of a market failure.   A central planner subject to the same information constraints or the same transaction costs could not do better than the market. . . . Incomplete contracts . . . distort the choices that an intermediary would otherwise make; relaxing these constraints by defaulting in some states of nature allows the intermediary to provide the depositor with superior risk sharing and/or higher returns.   (pp. 188-189)

To provide a justification for regulation of the financial system, we first need to identify a source of market failure (constrained inefficiency).   Then we need to identify a practical policy that can remedy or at least ameliorate that failure. . . . Our view is that it is not enough merely to show that there exists a welfare-improving policy.   We also need to characterize the policy and show that it can be implemented.   A badly designed intervention could make things worse.     If the welfare-improving policy is too complicated or depends on information that is unlikely to be available to the policymaker, such mistakes are likely.   (p. 191)

[T]he optimal allocation [of resources] can involve financial crises.   So it is not the case that eliminating systemic risk is always optimal.   A careful analysis of the costs and benefits of crises is necessary to understand when intervention is necessary.   This analysis is typically missing from proposals for capital adequacy regulations such as the Basel Accords.   Incomplete financial markets provide one plausible source of market failure and a possible justification for capital regulation. (p. 214)

These quotes just give a flavor of the conclusions, without shedding light on the careful analysis that underlies them.   But the analysis is there.   The models are very bare-bones, and I imagine that some will conclude as a result that the analysis is therefore unrealistic.   But I derive a very different conclusion.   Things are very devilishly complicated in even the simple models, and even in the simple models seemingly “obvious” policies like increased capital requirements or regulations on bank balance sheets can have beneficial or baleful effects depending on the values of parameters that are impossible to measure in practice (e.g., the coefficient of relative risk aversion).   Thus, the models show that in even simple economies “obvious” solutions can have perverse, unintended consequences.   The potential for such policy failures is even greater in the more complex real world.

The chapter on financial regulation (Chapter 7) is particularly provocative and insightful.   Allen and Gale present a model in which markets are incomplete and there are random liquidity shocks.   In such an environment, bank capital can improve risk bearing; bank equity investors can help insure depositors against liquidity shocks.   But bank capital is costly.   Evaluation of the optimal amount of bank capital requires a comparison of the benefits of the cost of equity capital with the benefits of insurance.   This is not a trivial task even in the simple model.   Moreover, the answer is not always that forcing banks to hold more capital than they would under choose to hold absent any regulation improves welfare.   In some cases, forcing banks to hold less capital can increase welfare (p. 200).   The authors’ warning is well warranted:

It is not known how robust these results are when the model specification is altered or generalized, but even if the results turn out to be special they reinforce the lesson that, when general-equilibrium effects are involved, it is very difficult to predict the macroeconomic effect of changes in capital structure across the financial system.   Until we have a general theory to guide us, caution in policy making seems to be advisable.

In the current environment, this discussion of the response to the Great Depression is particularly worth chewing over:

Looking back, there is no sign of formal theory guiding these [Great Depression-era policy] changes.   Everyone seems to have agreed the experienceo f the great Depression was terrible; so terrible that it mut never be allowed to happen again.   According to this mind set, the financial system is fragile, adn the purpose of prudential regulation is to prevent financial crisis at all costs.   Why does the mindset of the 1930’s continue to influence thinking about policy? [Why] does policy making continue to be an empirical exercise, with little attention   to the role of theory?   This empirical procedure is unusual.   Indeed, the area of financial regulation is somewhat unique n the extent to which the empirical developments have so far outstripped theory.   In most areas of economics, when regulation becomes an issue, economists have tried to identify some market failure that justifies the proposed intervention.   Sometimes they have gone further and derived the optimal form of regulation.   This has not been the usual procedure with financial regulation, however.

Re that last comment: if that was true in last year, it is infinitely more true today   There are more proposals to “reform” the banking and financial systems than you can shake a stick at.   Most of these are based on casual empirical observations of “what went wrong” that are, at best, journalistic and impressionistic.   The proposals do not reliably identify what the market failures are, or how proposed regulatory changes would remedy those market failures.   They don’t trace through the likely responses of market participants to the proposed regulations, or the costs and benefits associated with these responses.

I’m not trying to be Panglossian here, suggesting that we are in the best of all (financial) worlds.   But I am suggesting that there is a major outbreak of the nirvana fallacy going on–i.e., there is a naive belief that the market has failed and that government regulation can only make it better.   Both of these beliefs are of a religious variety, and only weakly grounded on rigorous logic or empirical evidence–if they are grounded at all.

Allen and Gale don’t have all the answers–and they admit as much–but they do pose some very provocative questions and challenges.   These questions and challenges are especially apposite today.   Anyone giving the book a fair read should come away with a healthy skepticism of any proposed regulatory fix to prevent the occurrence of the next crisis.     Especially if that fix seems intuitive or obvious.

Print Friendly, PDF & Email

No Comments »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress