Printer Friendly

The Stock Market: Bubbles, Volatility, and Chaos.

This volume presents a collection of papers presented at the Federal Reserve Bank of St. Louis's thirteenth annual conference on economic policy, held during October 1988. As the title suggests, the papers focus on the behavior of asset prices and the workings of the markets where stocks are traded. The general theme connecting the five papers is the 1987 world crash. A commentary follows each paper. Since the papers approach the issue of asset pricing from several vantages it is worthwhile to discuss each paper and commentary separately.

The first paper, by Diba, reviews some possible theoretical models which appear relevant for assessing the existence of stock price bubbles in a rational expectations setting. Referring to the voluminous work on stock price volatility, Diba concludes that while volatility probably does not move in lock step with changes in fundamentals, bubbles cannot be identified as the missing link in the volatility puzzle. In his discussion of the empirical work he confirms a well-known dilemma found in all bubble literature--the inability of the tests to discriminate between economic behavior motivated by fundamentals and that due to bubbles. He concludes that the presence of rational pricing bubbles, especially in long-run U.S. data, is improbable.

Regarding the modeling environment implied by the behavior of asset prices, Robert Flood's comment on the Diba paper and related literature is more condemning. He states that when it comes to identifying the existence or nonexistence of bubbles, "we have no idea what is going on." He is correct; for this avenue of financial research to make a meaningful contribution it will be necessary to more adequately relate the theory to the available data.

Editors Dwyer and Hafer using international data from seven globally active stock exchanges ask whether bubbles and/or fundamentals determine stock prices. They also address the "contagious bubble" phenomenon associated with the 1987 world crash. Using various cointegration and regression models they find little support for either a bubbles or fundamentals-cum-bubbles explanation of cross-country stock index movement. Regarding October 1987, their tests do little to suggest that the world crash was evidence of the bursting of a global bubble. The critique by von Furstenberg and Jeon, however, questions the latter conclusion. They feel that the editors' unit root and cointegration methodology, by failing to model cross-market information flow, is biased towards rejecting the hypothesis of contagion. In general, their critique returns to the same questions raised with the first paper: "Are pricing models correctly specified so that reliable conclusions can be drawn?"

The paper by Ramsey contends that since noise so dominates the time variation is economic and financial data, chaos theory and qualitative nonlinear dynamic analysis can provide useful alternatives to traditional structural and vector autoregressive modeling methods. Chaos theory is borrowed from the physical sciences. Applied to economics or finance, the paradigm is based on the idea that relationships, which are inherently nonlinear and deterministic in nature, still give rise to observations that follow white noise processes. Thus, Ramsey contends, chaos methods can outperform more conventional techniques in identifying the true underlying model.

The commentary by Dechert expresses severe reservations about the usefulness of chaos techniques in economics/financial applications. First, any improvements in forecasting using nonlinear models are limited to the very short-run. Second, data requirements are prohibitive, requiring upwards of 50,000 observations. I add that the mathematics requirements are prohibitive as well; those not already well grounded in the qualitative dynamics literature will find chaos applications to financial and economic data rough going.

Under normal market conditions when trading in securities is continuous, the S&P 500 basis, the difference between the cash market and the contemporaneous index futures prices, should be zero or slightly positive. As a result any nontrading bias in the basis due to stale prices should be small. The fact that during the October 1987 stock market crash a basis of (negative) 10% and more was systematically observed has lead some researchers to investigate the continuity of trading during this time of drastically declining stock prices. When prices declines are large any nontrading effects will be magnified such that the index will be biased upward and the basis biased downward. Thus, the relevant research question is what part of the negative basis observed over the days preceding the crash was due to a nontrading effect and what part was real.

The inventive paper by Moriarty, Gordon, Kuserk, and Wang answers this question. They do so by constructing a proxy S&P 500 index, one that includes only those stocks that were actively traded between October 14 and October 26. Replacing the normal S&P 500 index with their proxy they find the lead time between futures and cash prices declining from fourteen to four minutes during the week of October 12, and to zero during the week of October 19. Thus, contrary to what is observed using recorded data, they conclude that there was simultaneous price discovery in both markets, a result undoubtedly attributable to the chaotic events of the episode and the effects these events had on traders.

In (separate) commentary by Harris and Cornell, the findings of the Moriarty et. al. study are used to make suggestions to improve the mechanics and interaction of the spot and futures markets during periods of turbulence. These suggestion center around improved information processing.

The final paper, by David Haddock, examines the Brady Report and the possible social efficiencies which might accrue from increased regulation of securities markets. Haddock's comments, while insightful, just reiterate what is the standard scenario for situations like the crash: (1) a terrible market disruption occurs; (2) legislators immediately call for a commission to study the catastrophe; (3) a report is issued which points out market excesses; (4) legislation is proposed which is supposed to correct or reduce the problem(s); (5) economists under the banner of free markets, decry the proposed legislation; (6) the whole flap goes away until the next catastrophe occurs, when the scenario repeats itself.

However, Haddock's comments do offer something very constructive in that they suggest modernization of securities markets in place of regulation. Like others, he implies that the crash highlighted what has long been suspected; the ability of securities markets to received information has far outpaced their ability to process and disseminate that information. But, he argues, modernization is best accomplished via the profit motive; regulators are doomed to failure since they lack this motive.

While Kenneth M. Lehn's commentary on the Haddock analysis is generally supportive, he does believe the Brady study makes some positive suggestions regarding regulation, especially as they concern the importance of maintaining the linkage between the futures and stock markets during periods of financial turbulence.

From speculative bubbles to excessive stock price volatility to market regulation, this volume presents an interesting collection of essays. Questions regarding the crash (Was a bubble involved?), and asset price volatility (Why are prices so volatile if fundamentals aren't?) remain unanswered. Questions regarding the crash-related basis (Why was the negative basis observed?), and regulation (Can additional equity market legislation provide positive social welfare?) have been reasonably answered.

To this reviewer the book suggests a sequel; one specifically aimed at the on-going research concerning the modeling of linkage connections tying national stock markets. This topic has taken on added importance since the crash when cross-market volatility was at a peak. The editors' own work has provided interesting results in this regard |2~, as has the research of others |1; 3~.

In conclusion, the book should provide an important contribution to the libraries of those interested in crash-related research on asset pricing and financial markets.


1. Brocato, Joe. "Evidence on Adjustments in Major National Stock Market Linkages over the 1980s." Journal of Business Finance and Accounting. Forthcoming.

2. Dwyer, Gerald P., Jr. and R. W. Hafer, "Are National Stock Markets Linked?" Federal Reserve Banks of St. Louis Review, November/December 1988, 3-14.

3. Koch, P. and T. Koch, "Evolution in Dynamic Linkages Across Daily National Stock Indexes." Journal of International Money and Finance, 10, 1991, 231-51.
COPYRIGHT 1993 Southern Economic Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brocato, Joe
Publication:Southern Economic Journal
Article Type:Book Review
Date:Jan 1, 1993
Previous Article:Kuznets's inverted-U hypothesis: reply.
Next Article:The Rise of the Japanese Corporate System: The Inside View of a MITI Official.

Related Articles
Economic Complexity: Chaos, Sunspots, Bubbles, and Nonlinearity.
From Catastrophe to Chaos: A General Theory of Economic Discontinuities.
The Science Book. (Books: evolving science).
Political Numeracy: Mathematical Perspectives on Our Chaotic Constitution. (Books: a selection of new and notable books of scientific interest).
Dot.Con: The Greatest Story Ever Sold. (Political booknotes: net loss).
How the Universe Got its Spots: Diary of a Finite Time in a Finite Space.
Why Stock Markets Crash: Critical Events in Complex Financial Systems.
Political Numeracy: Mathematical Perspectives on Our Chaotic Constitution.
Critical Mass: How One Thing Leads to Another.
Coincidences, Chaos, and All That Math Jazz.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters