Executive Summary:
Predictive Power in Stock Market Returns: Glamour Stocks
A model developed by researchers Venkat Eleswarapu and Marc Reinganum of SMU Cox analyzes how glamour stocks have a unique predictive power in aggregate stocks market returns that other classes of stocks lack. Glamour stocks is a term used to denote high growth stocks and represent the “trendy” stocks for which investors are interested. In the 80s and 90s, technology stocks were king. In the 90s, telecommunications stocks were hot; Internet stocks were glamorous for a period as well in the late 90s. This research documents the predictability of stock market returns that may be caused by "behavioral" factors like excessive optimism that leads to mis-pricings. The study “The Predictability of Aggregate Stock Market Returns: Evidence Based on Glamour Stocks” in April’s Journal of Business adds one more piece of evidence for the behavioralist point of view in the world of financial economics.
Key Findings
The study documents that annual returns on the stock market can be predicted using the prior performance of a portfolio of glamour stocks. Specifically, it was found that when glamour stocks have done very well in the prior 36-months, this trend is followed by very low or negative returns in the stock markets in the subsequent year. In such periods, the chances of big negative returns (i.e. market crashes) also dramatically increase. Since traditional rational asset-pricing models do not predict such relations, the evidence appears to suggest some "behavioral" phenomenon.
In contrast, returns from value stocks and from the aggregate market, purged of glamour stocks, have no predictive power of future returns. The probability of large stock market declines or “corrections” are high for the total stock market and in particular glamour stocks, following the 36-month run up of glamour stocks. In other words, a “boom” period for glamour stocks implies future weakness in the overall market. In the study, stock market returns were analyzed from the period 1951 to 1997. “The big run-up in Nasdaq stocks and subsequent stock market declines in 2001-2002 would have been predicted by our empirical model. Since our study only included data up to December 1997, the actual market behavior in 1998-2002 of speculative bubble followed by a significant downturn would serve as an out-of-sample validation of our results,” noted Dr. Eleswarapu.
Predictions and Conclusions
In the past year glamour stocks have performed well, but not over the last three years. “Based on this model, there is no imminent correction in the near term, in spite of the fact that the stock market has risen considerably,” Dr. Eleswarapu explained. “When the overall stock market declines, glamour stocks tend to decline even further as they are often where the trouble lies.”
This model is consistent with the behavioralist point of view which has been gaining ground and recognition over the last few years. When Daniel Kahneman of Princeton won the Nobel Prize in 2002 for his work in this field, it further legitimized its growing relevance and power of explanation—integrating insights from psychological research into economic science, particularly concerning human judgment and decision-making. From an academic viewpoint, Eleswarapu believes the burden of reconciling this “behavioral” phenomenon from the rationalists’ camp is now in their court. “Stock market crashes are difficult to predict based on the rational view in financial economics, and so we further the argument on the behavioralist side,” explained Dr. Eleswarapu.
March 31, 2004 |
