Stock returns and inflation: A long-horizon perspective

Two main empirical facts regarding the statistical relation between stock returns and inflation emerge from the current literature in finance. The first is that ex post nominal stock returns and inflation are negatively correlated. Financial economists consider this result surprising since stocks, as claims against real assets, should compensate for movements in inflation. The second, and related, empirical result documents a negative relation between ex ante nominal stock returns and ex ante inflation. Since the Fisher model implies that expected nominal rates should move one-forone with expected inflation, this negative correlation strikes at the heart of one of the oldest and most respected financial models (see e.g., John Lintner, 1975; Zvie Bodie, 1976; Charles Nelson, 1976; Eugene Fama and G. William Schwert, 1977; Jeffrey Jaffe and Gershon Mandelker, 1977; N. Bulent Gultekin, 1983; Gautam Kaul, 1987). Existing studies, however, have focused almost exclusively upon short-term asset returns with time horizons of one year and less.1 Since the Fisher model (and its corresponding intuition) might be expected to hold at all horizon lengths, this void in the empirical literature is unfortunate for several reasons. First, from a practical perspective, many investors hold stocks over long holding periods. Therefore, it is important to know the manner in which stock returns move with inflation over longer horizons. Second, the relation between stock returns and inflation at long horizons is of particular interest given that the results at short horizons (both ex ante and ex post) appear to be anomalous. That is, evidence at these longer horizons may provide additional information regarding explanations for the negative correlation between nominal stock returns and both ex ante and ex post inflation. In this paper, the relation between stock returns and inflation at long horizons is examined. In approaching this issue, several problems arise which must be addressed. The first difficulty is the necessity for a long data sample in order to capture long-term movements in the time series of returns. We have been able to accumulate two centuries of data on stocks, short-term and long-term bonds, and inflation in both the United States and the United Kingdom in order to fulfill this requirement. The second difficulty results from the inability to model ex ante long-term inflation accurately. We circumvent the absence of any long-horizon inflation model by using an instrumental* Finance Department, Stern School of Business, New York University, New York, NY 10012, and Finance Department, Wharton School of Finance and Commerce, University of Pennsylvania, Philadelphia, PA 19104-6367, respectively. An earlier version of the paper was circulated under the title, Stocks Are a Good Hedge for Inflation (in the Long Run). We thank Pierluigi Balduzzi, Bob Cumby, Silverio Foresi, Yasushi Hamao, Julie Richardson, Jeremy Siegel, Tom Smith, participants at the Western Finance Association meetings (San Francisco, 1992), and two anonymous referees for their helpful comments and suggestions. Special thanks to Tim Opler and Jeremy Siegel for use of the data. The authors gratefully acknowledge research support from the New York University Salomon Center (Boudoukh), and the Geewax-Terker Research Foundation (Richardson). IAs a representative sample of this literature, in their study of a variety of assets, Fama and Schwert (1977) document a negative relation between ex ante stock returns and expected inflation using monthly, quarterly, and semiannual data.