Stock Return Predictability: Is it There?

By Andrew Ang and Geert Bekaert

Previous research has documented the predictability of aggregate stock excess returns, cash flows and interest rates by the dividend yield (see e.g. Campbell 1991 or Cochrane 1992). High dividend yield forecast high stock excess returns.

This paper by Andrew Ang and Geert Bekaert revisits this common wisdom:

- They find that the predictability is in fact quite poor at long horizon when they correct the standard errors for heteroskedasticity.

- They find predictability only on the short run using the dividend yield in conjunction with the short-term interest rate.

- They show that the dividend yield does not robustly predict future dividend growth but only high future interest rates.

They use a new non-linear present value model for stocks consistent with the predictability evidence to explain the empirical findings. They run predictability regressions and test their model not only on US data but also on international data.

The US stock data is the S&P Composite Index and comes from the S&P’s Security Price Index Record from June 1935 to December 2001. The UK and German stock data are the Financial Times Actuaries Index and the Deutsche Borse CDAX Index from June 1953 to December 2001 and come from Global Financial Data. They also use a short sample data from MSCI and manage to cover the US, UK, Germany and France.

They first investigate the return predictability in the US and run the regression of annualised stock market excess returns at annual or quarterly horizon on a set of independent variables. To compute the standard errors they follow Hodrick (1992) that they compare to the methods of Newey-West (1987) and Hansen and Hodrick (1980) (Appendix A of the paper). Over the long US sample data, the Hodrick t-statistic of the dividend yield is significant only for 2 to 4 quarters. At one quarter, it needs the addition of the short-term interest rate. The short-term rate is a stronger predictor with a negative coefficient (higher rate means lower future returns).

In their model, the risk free rate and the log dividend growth follow a VAR. The discount rate depends on its lagged value and the contemporaneous value of the risk free rate and the log dividend growth. The authors calculate the formula for the price-dividend ratio and show that it is a non-linear function of interest rates, excess returns and cash flows. They can estimate the model using Simulated Method of Moments on US data (Appendix E).

They consider five different specifications for the discount rate process: two nulls and three alternatives. The Third alternative is the most general specification without any restrictions on the parameters.

After estimation of the model, the authors investigate the predictive regressions that have been run in the literature. They compare the regression coefficients (expected excess return regressions, dividend growth regressions and risk-free rate regressions) implied by their model to the values in the data. The most general model (Alternative 3) is the most consistent with the data.

They revisit the earnings yield and excess return predictability but find that the predictability is not robust however they find that dividend and earning yields predict dividend growth.

The paper is available at http://www2.gsb.columbia.edu/faculty/aang/papers/pred.pdf and http://rfs.oxfordjournals.org/cgi/content/abstract/hhl021v1

References:

Campbell, J. Y., 1991, “A Variance Decomposition for Stock Returns,” Economic Journal, 101, 157-179.

Cochrane, J. H., 2001, Asset Pricing, Princeton University Press.

Fama, E., and F. French, 1988, “Dividend Yields and Expected Stock Returns,” Journal of Financial Economics,22, 3-26.

Fama, E., and G. W. Schwert, 1977, “Asset Returns and Inflation,” Journal of Financial Economics, 5, 115-146.

Hansen, L., and R. Hodrick, 1980, “Forward Exchange Rates as Optimal Predictors of Future Spot Rates: An Econometric Analysis,” Journal of Political Economy, 88, 829-853.

Hodrick, R. J., 1992, “Dividend Yields and Expected Stock Returns: Alternative Procedures for Inference and Measurement,” Review of Financial Studies, 5, 3, 357-386.

Lettau, M., and S. Ludvigson, 2001, “Consumption, AggregateWealth and Expected Stock Returns,” Journal of Finance, 56, 3, 815-849.

Lettau, M., and S. Ludvigson, 2005, “Expected Returns and Expected Dividend Growth,” Journal of Financial Economics, 76, 583-626.

Lewellen, J., 2004, “Predicting Returns with Financial Ratios,” Journal of Financial Economics, 74, 209-235.

Menzly, L., J. Santos, and P. Veronesi, 2004, “Understanding Predictability,” Journal of Political Economy,112, 1, 1-47.

Newey, W., and K. West, 1987, “A Simple, Positive Semi-Definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix,” Econometrica, 55, 703-708.

Polk, C., S. Thompson, and T. Vuolteenaho, 2005, “New Forecasts of the Equity Premium,” forthcoming Journal of Financial Economics.

See also {ln:How Well Do Financial and Macroeconomic Variables Predict Stock Returns}.