Macroeconomic variables and the E/P ratio: Is inflation really positively associated with the E/P ratio?

We examine the economywide E/P ratio both over a long time period (1952–2003) and by dividing the entire period into subperiods. We have two main objectives. First, we reexamine the puzzling result documented by Modigliani and Cohn (1979), who find that, contrary to theory, the economy wide E/P ratio and inflation are positively correlated. Our longer period of analysis allows us to divide the entire period into subperiods with differing macroeconomic environments. We find that the Modigliani and Cohn anomaly is period specific. The association between the E/P ratio and inflation reverses from one period to another. Hence, the relation between inflation and the economywide E/P ratio is not stable over a long time period. Second, we analyze the associations between the economywide E/P ratio and its two main drivers as given by the Gordon (1962) model. We find that the economywide E/P ratio (a) is not associated with the real interest rate and (b) is weakly negatively associated with the expected growth rate. Findings for inflation do not change when we include or exclude other E/P drivers in regression specifications.