When identifiable actions of private firms impose negative externalities, it is natural to ask about the feasibility and desirability of government intervention. Of course, we should mean the feasibility and desirability of interventions that will be undertaken by actual governments in specific institutional settings, rather than invoking ideal governments or ideal institutions.
Eatwell is concerned in this paper about a particular kind of externality. He is concerned about the actions of firms in international financial markets, especially firms that take very large positions. The externality that concerns him is the contribution to systemic risk of the profit seeking activities of these firms.
Eatwell takes the desirability of government action for granted, and he explores how most effectively to regulate international financial markets when the goal is to eliminate systemic risk. As a side constraint, he desires the process to be participatory enough that developing countries do not have a regulatory structure imposed upon them by developed countries.
Eatwell's focus on an externality in the determination of systemic risk allows him to treat the issue of financial market regulation as fundamentally one of efficiency. He finds certain events, such as the Long Term Capital Management debacle, to render the potential economywide inefficiency of liberalized financial markets ``indisputable.'' Other examples offered include Herstatt Bank in 1974, and the Asian financial crisis. He does not take up the question whether low frequency bad outcomes might be justified by higher average growth, except to suggest that higher average growth has not accompanied financial liberalization.
Indisputable or not, it would be nice to know precisely what Eatwell has in mind when he refers to systemic risk and the externality in its determination. This is left largely implicit in the paper, and we guess from his examples and discussion that his focus is the following: the international asset positions taken by private firms can increase the likelihood of particular types of macroeconomic disruptions. Specifically, it increases the likelihood that ordinary production and distribution activities will be disrupted by sudden changes in interest rates, exchange rates, or the availability of finance.
Eatwell makes it clear that he is irritated by (p.1) ``overblown claims for the efficiency of financial liberalisation''. But fundamentally he is not challenging the desirability of international capital flows, at least in this paper. Instead he appears to be noting that financial liberalization comes with costs as well as benefits, and appropriate regulation can mitigate the costs. However, he is also willing to offer provocative observations such as (p.1): ``liberalisation of international financial markets has coincided not only with increased financial instability, but also with a worldwide slowdown in the rate of growth.'' In a paper such as this one, I think he should be less coy in describing the full range of costs and benefits of international financial liberalization as he sees them.
Eatwell believes that (p.1) ``for efficient regulation the domain of the regulator should be the same as the domain of the market that is regulated.'' I'll call this the ``weak-link hypothesis,'' since I believe the core idea is that lack of coordination among national regulatory regimes will tend to produce local inadequacies that contribute to systemic risk. (For example (p.9), ``If, in a liberal international financial environment, high standards are not uniformly maintained then firms authorised in a less demanding jurisdiction can impose unwarranted risks on others, undermining high standards of authorisation elsewhere.'') I am uncertain of the status of this hypothesis. I take it to be a pragmatic point rather than an abstract theoretical point. For example, at the abstract level it seems to me that accord among national governments---even accidental accord---would achieve whatever could be achieved by an international body. The pragmatic response is perhaps that nations will not converge to efficient rules without an international regulatory body. Any such response, it seems to me, deserves an extensive political economic justification. Furthermore, once the discussion enters the pragmatic realm, we have to take up the problems of extending the domain of any regulatory body. We do not want to slip into the assumption that a new international regulatory body will be more skillful or less corruptible than existing bodies.
Anyway, suppose we accept the weak-link hypothesis. Since financial markets are increasingly global in scope, this suggests that a global regulatory body is needed. As a point of reference, Eatwell returns to the hypothetical World Financial Authority (WFA) introduced in earlier work with Lance Taylor. Adoption of this point of reference has benefits and costs. The benefit of having an ideal point of reference is that it aids assessment of the performance of the existing institutional structure, and it suggests paths for the evolution of existing institutions. The disadvantage is that if one does not believe that a properly functioning WFA is an achievable institution---and I believe this is no more likely than that the IMF will become an embodiment only of the virtues implicit in its charter---then comparisons to the WFA may distort the assessment of existing institutions. Nevertheless, I think that Eatwell has rightly judged that the benefits of taking this approach in this paper outweighs the costs.
In this paper, Eatwell wants to (p.2) ``identify exactly what the key regulatory tasks might be and thereby to develop proposals as to how they might best be performed.'' His view of financial markets is explicitly Keynesian (p.3), which means he focuses on the implications of instabilities in investor attitudes for outcomes in financial markets (Keynes, 1936). On the one hand, this is a welcome contrast to silly rational expectations analyses of the performance of financial markets, which take place in fantastic worlds where market participants possess godlike understanding of the economic structure in which they act. On the other hand, it means that some of the analysis is hard to assess. For example (p.3), ``Average opinion is typically stable for long periods, dominated by convention. In these circumstances it may be comparatively easy to identify the behavioural relationships that characterise such periods of tranquillity, to believe that they are stable and enduring, and use them to assess values at risk. Yet these seemingly `true models' of the marketplace can be completely overwhelmed by a sharp shift in average opinion, driving markets in previously unimaginable directions, and producing potentially catastrophic disruption in the operation of the real economy.'' I confess again that I simply do not know how to assess such a statement. What is a ``long period''? Which beliefs are dominated by ``convention'', which I presume is intended to be a contrast to beliefs dominated by market fundamentals? (Explicit reference to the actual experience of, say, bond traders would be helpful in pinning this down.) Are we to treat the changes in ``average opinion'' as exogenous, or are these institutionally conditioned? What is the meaning of ``catastrophic''? E.g., were the actual outcomes of the Asian currency crisis catastrophic (and against what counterfactual)? Finally, once we clear up all these questions, is the statement supposed to be self-evident, or is it supposed to be confirmed already by the specific events referenced in the paper, or is it supposed to be a testable empirical claim about the functioning of current financial markets?
I speculate that the choice of the word ``catastrophic'' is a signal that catastrophe theory is a background framework for the interpretation of the economy. In particular, the idea may be that small changes in parameters (e.g., investor sentiment) can lead to large changes in the behavior of the system (e.g., international financial markets). If so, I would like to be offered more of a sense of what evidence drives the adoption of such a framework for interpretation. After all, one might also claim that---in comparison with the overall functioning of the international macro economy---even the crises experienced in the postwar era are less impressive than the tendency of the system to stabilize itself and ``muddle on'' much as before. How would one assess the relative importance of these two contrasting interpretations of the postwar experience?
However these questions are resolved, we can draw at least two interesting suggestions out of them. The first is perhaps seen as a Post Keynesian twist on the Lucas critique: the structure captured in macroeconometric models may prove irrelevant in times of crisis. The second is an unusual characterization of one of the externalities in financial markets: the actions of private firms may influence the aggregate state of expectations (not just by being informative about fundamentals). For example (p.4), ``the failure of a single firm can, by influencing expectations, have an influence not only on its immediate counterparties, or even firms dealing in similar products, but also, through its impact on expectations, on financial markets as a whole, and, via macro variables, on the real economy at home and abroad.''
I like this paper's pragmatic, policy oriented perspective and its eschewal of oversimplified abstract theory in favor of my subtle and informed political economic analysis. Another thing I like about this paper is that it focuses our attention on the the importance of relating microeconomically motivated regulatory rules to macroeconomic outcomes, particular systemic risk. Eatwell seems particularly sympathetic to macro-regulatory attempts to ``price in'' this systemic risk (p.5), especially in ways that recognize that identical risk taking in different kinds of firms can have different implications for systemic risk. I also like the sensitivity to the inadequacy of aggregate data to the problem of systemic risk. For example Eatwell notes (p.5), ``aggregate capital adequacy ratio of the financial sector, one of the indicators collected, could easily conceal major risks.'' I also like the Minskyan emphasis on positive short-run feedback loops in the financial markets, and the suggestions that regulatory practices should be modified to dampen rather than exacerbate these tendencies.
Finally, I like Eatwell's discussion of the ongoing evolution of new and existing institutions to the needs of evolving financial markets. For example, he discusses how the IMF is evolving a surveillance role related to that envisioned for a WFA. I think he is right that (p.10) ``There is, in effect, a creeping internationalisation of the regulatory function in confederalist international financial markets.'' He observes further that (p.17), ``10 years ago most of the regulatory functions, with the exception of the policy function, would lack any international dimension. Today in all areas other than authorisation, international bodies are taking up some of the regulatory tasks.'' This point is undeniably important, and Eatwell's paper offers useful points of reference for assessing the usefulness and adequacy of the evolving institutional response to modern financial markets.