Models to Predict Stock Market Crashes
Center for Financial Studies Research Seminar Series
William T. Ziemba, Professor Emeritus of Financial Modeling and Stochastic Optimization
Sauder School of Business, University of British Columbia
In this paper, we discuss the theoretical setting of the bond-stock earnings yield differential model (BSEYD), test empirically its ability to predict equity market corrections and compare it with the Campbell-Shiller price-earnings ratio model. Our results extend the existing literature in three important ways. First, we derive a relation between the Campbell and Shiller price-earnings ratio model, an empirical model, and the Gordon growth model, a theoretical equity valuation model. Second, we show that developing a statistical test of the predictive power of a measure is easier when it comes to equity market corrections and crashes than for market downturns. Third, we find that the P/E ratio contains statistically significant evidence about future equity market corrections, although its significance is not as strong as that of either the BSEYD measure or its parent the log BSEYD measure. We conclude that all three models can be used as predictors of equity market corrections.
Professor William T. Ziemba taught at University of British Columbia from 1968-2006 and has been a professor at the ICMA Centre, University of Reading. He has been a visiting professor at Cambridge, Oxford, London School of Economics, and Warwick in the UK, at Stanford, UCLA, Berkeley, MIT, University of Washington and Chicago in the U.S., as well as at other universities across the globe. He has been a consultant to a number of leading financial institutions, such as Frank Russell Company, Morgan Stanley, Buchanan Partners, RAB Hedge Funds, and in the gambling area, to the BC Lotto Corporation, SCA Insurance, Singapore Pools, Canadian Sports Pool, Keeneland Racetrack. His research is in asset-liability management, portfolio theory and practice, security market imperfections, Japanese and Asian financial markets, hedge fund strategies, risk management, sports and lottery investments and applied stochastic programming. He has published widely in journals such as Operations Research, Management Science, Mathematics of OR, Mathematical Programming, American Economic Review, Journal of Economic Perspectives, Journal of Finance, Journal of Economic Dynamics and Control, and JFQA and in many books and special journal issues. His Ph.D. is from the University of California, Berkeley.