Financial Markets

On 29 April CEPR continued its series of workshops on the performance of financial markets organized by ATE Programme Director Colin Mayer (City University Business School (CUBS)), with financial support from the Ford and Alfred P Sloan Foundations. The January meeting, reported in Bulletin No. 19, suggested three areas of research: the performance of markets, financial regulation and the provision of corporate finance. The last of these is in progress under the direction of Margaret Bray (LSE and CEPR), Jeremy Edwards (Cambridge University and CEPR) and Colin Mayer. Financial regulation was discussed at workshops in March and April, reported above. The April 29 workshop concentrated on the the performance of markets: the emphasis was on exploring three promising directions for new work rather than presentations of completed research.

New Financial Instruments
Richard Stapelton (Churchil College, Cambridge) opened the workshop with a discussion of new financial products. He contrasted three types of options. 'Simple' options represent a straight forward gamble on the final value of the asset on which the option is written. 'Path-dependent' options have pay outs that relate not only to the final value of the asset but also to the path that the asset price takes over the life of the option. 'Multivariate' options" have values that depend on more than one variable, for example the dollar and the yen. Options are now sold in packages that combine features of these three varieties, such as combinations of puts and calls and options in which the gains are shared between the seller and the purchaser of the option. Was it possible to describe these new instruments by means of only a few parameters, such as their forward prices, put prices and percentage paybacks? Such a characterization could help evaluate the functions performed by the new instruments.

Any evaluation of new instruments requires some assessment of their contribution to risk hedging. Roy Batchelor (CUBS) suggested that theoretical tools already exist to assess the usefulness of financial instruments: these tools characterise instruments in terms of their effects on risks and returns, rather than more complex welfare measures. A separate but related issue concerns the way in which the users of financial instruments establish whether a hedge is performing a valuable function. Can performance measures be derived? Gordon Hughes has suggested that these measures can be extended to institutions to evaluate, for example, the performance of unit and investment trusts. This is a research topic that would be of obvious interest to practitioners as well as academics. There was also discussion about the identity of the users of new financial instruments and, in particular, the extent to which these instruments are employed outside the financial sector. Little is known about the ultimate use of financial instruments: this too is an important area of future research.

Market-Making
Gordon Gemmill (CUBS) then introduced the second theme of the workshop: market-making. He described two areas of research that he thought should be undertaken. First, an in-depth study of traders was needed. This had already been done in the United States and had revealed, for example, the remarkably short holding periods which certain classes of traders required in order to make profits. Gemmill thought that an attempt should be made to describe the activities of traders and the way in which information is incorporated in trades. Second, he noted that there were interesting differences between markets which theory found difficult to explain. Why are some markets over-the- counter and others dealer-driven? Why is the housing market essentially an 'agency' one? He noted intriguing cases in which markets (for example, wood pulp) had switched from one for of organization to another and others where particular activities were organized in different ways. Shipping, for example, is an agency market, but freight futures are traded through dealers.

Ailsa Roell (LSE) noted that the analysis of market-making presents some difficult theoretical problems. Theory predicts that under some circumstances competition reduces short run profits to zero and firms fail to cover their fixed costs. Differences in information, however, may permit some traders to earn positive profits. These theoretical problems raise several questions. What happens when market makers exchange information? Would one expect to see markets dominated by single market- makers? How does the analysis change if, as seems quite plausible, market-makers are better informed than investors? A second set of issues concerned the response of markets to the imposition of regulatory rules. For example, stock exchange rules stipulate a minimum size of trades. Such rules tend to reduce the number of competitors in a market and therefore to create monopolistic distortions. But such rules may be beneficial if they raise the size of trades by monopolists who would otherwise choose to trade in smaller volumes.

Trading and Pricing of New Instruments
A number of analyses suggest that an extension of participation in markets is not necessarily desirable. 'Good' investors do not necessarily drive out 'bad'; imperfectly informed investors may increase market volatility; and speculation is not necessarily stabilizing. Marco Pagano (University of Naples and CEPR) argued that any analysis of the desirability of trades must rest on a description of the motives for trade. He outlined six models of trading that gave rise to different predictions and welfare implications. Trade could be motivated, he argued by a change in market fundamentals, by changes in macro policy variables or for consumption reasons.

This analysis suggested a possible research strategy, which would involve an assessment of whether trade is beneficial in the context of different models of the motives for trade, an examination of what information is used in trading and how it affects prices, and a comparison of different market structures.

Several participants noted a number of important aspects of market structures that needed investigation. Why, asked Gordon Gemmill, does one market become dominant? Why did an oil futures market fail to become established in London? Why are some commodity markets concentrated in one country while others, such as those for cocoa, coffee and copper, are about the same size in the United Kingdom and United States? Pagano expressed interest in the factors that determine the growth of capital markets in countries and the ways in which countries with poorly developed markets can promote them.

Workshop participants were able to define a fairly precise theoretical agenda. The appropriate set of empirical tests may prove harder to specify. Elisabetta Bertero (LBS and CEPR) thought that non-parametric tests should be used to assess asset price disturbance models, which would help avoid relying on specific descriptions of the fundamental determinants of asset prices. Christian Wolff (LBS and CEPR) advocated the use of signal extraction techniques borrowed from engineering to identify unobservable components in asset prices. He had successfully employed Kalman filtering in evaluations of risk premia associated with forward foreign exchange markets: these techniques can be easily modified to consider such phenomena as asset bubbles.

The development of a research programme arising from these discussions will be the theme of future workshops.