Research Methodology, Multicollinearity

Edgar L. Feige, Douglas K. Pearce, The Casual Causal Relationship Between Money and Income: Some Caveats for Time Series Analysis, The Review of Economics and Statistics, Vol. 61, No. 4 (Nov., 1979), pp. 521-533

This paper re‐examines the effects of nominal contracts on the relationship between unanticipated inflation and an individual stock's rate of return. This study differs in three main ways from previous research. First, announced inflation data are used to examine the effects of unanticipated inflation. Second, a different specification is used to obtain more efficient estimates. Third, additional nominal contracts are considered. The empirical results indicate that time‐varying firm characteristics related to inflation predominately determine the effect of unanticipated inflation on a stock's rate of return. A firm's debt‐equity ratio appears to be particularly important in determining the response.

This paper re-examines the effects of nominal contracts on the relationship between unanticipated inflation and individual stock's rate of return. This study differs in three main ways from previous research. First, announced inflation data are used to examine the effects of unanticipated inflation. Second, a different specification is used to obtain more efficient estimates. Third, additional nominal contracts are considered. The empirical results indicate that time-varying firm characteristics related to inflation predominately determine the effect of unanticipated inflation on a stock's rate of return. A firm's debt-equity ratio appears to be particularly important in determining the response.

The concept of "economically rational" expectation formation is developed for a regime in which the acquisition and use of some information sets are nonnegligible. The concept provides a middle ground between "autoregressive" expectation formation and "rational" expectation formation. A necessary condition for the use of nonnegligible cost information sets is that such sets serve as leading indicators or more formally satisfy the causality conditions of Granger (1969). Using a time series modeling identification methodology developed by Box and Jenkins (1970) and Haugh (1972), we test the causal relationship between the rate of inflation and various monetary and fiscal aggregates. Surprisingly, we cannot reject the hypothesis that the rate of inflation is independent of the monetary and fiscal aggregates considered. Since the proposed leading indicator series contain no incremental predictive power once the information contained in the past history of inflation is efficiently utilized, we conclude that autoregressive expectation models may indeed represent economically rational price expectation formation.

The concept of "economically rational" expectation formation is developed for a regime in which the acquisition and use of some information sets are nonnegligible. The concept provides a middle ground between "autoregressive" expectation formation and "rational" expectation formation. A necessary condition for the use of nonnegligible cost information sets is that such sets serve as leading indicators or more formally satisfy the causality conditions of Granger (1969). Using a time series modeling identification methodology developed by Box and Jenkins (1970) and Haugh (1972), we test the causal relationship between the rate of inflation and various monetary and fiscal aggregates. Surprisingly, we cannot reject the hypothesis that the rate of inflation is independent of the monetary and fiscal aggregates considered. Since the proposed leading indicator series contain no incremental predictive power once the information contained in the past history of inflation is efficiently utilized, we conclude that autoregressive expectation models may indeed represent economically rational price expectation formation.