Abstract
I find that nominal equity returns respond to unexpected inflation more negatively during contractions than expansions. In particular, returns on firms with lower book-to-market ratio, or of medium size, demonstrate strong asymmetric correlations with unexpected inflation across the business cycle. The cross-sectional correlations of returns on book-to-market and size portfolios with unexpected inflation mostly reflect the heterogeneous factor loadings of these portfolios on one of the Fama–French factors, namely, the excess market return. By examining the cyclical responses to unexpected inflation of the three primitive forces which determine stock prices: the discount rate, the expected growth rate of real activity and the equity risk premium, I find that changes in expected real activity and the equity premium, signalled by unexpected inflation, are important in explaining the asymmetric responses of the stock market to unexpected inflation across the business cycle.
Acknowledgements
I thank Tara Sinclair and Tong Yao for helpful comments and suggestions.
Notes
1 In Wei (Citation2005), I have also used Gibbs sampling method to extract unexpected inflation from the nominal interest rate and inflation series. Alternative measures of inflation, such as the Gross Domestic Product (GDP) deflator, have been used to examine the robustness of the findings. The results of this article are robust to alternative measures of unexpected inflation. The results are available from the author upon request.
2 I have run regressions with additional independent variables, including the 30-day treasury bill return. I only keep those regressors which have statistically significant coefficients.
3 I use a forecasting model to construct a proxy for the unobserved unexpected inflation series. This gives rise to the well-known errors-in-variables problem, meaning that the estimated slope coefficients will be biased towards zero. The direction of bias suggests that the business cycle asymmetry of inflation betas can be even more salient than reported in the regression results.
4 I thank John Y. Campbell for providing me with data on the cross-sectional risk premium.