CAUSALITY RELATIONS AMONG STOCK RETURNS, INTEREST RATES, REAL ACTIVITY, AND INFLATION: SOME INTERNATIONAL EVIDENCE

Abstract

Variou theoretical framework have been developed to explain the negative correlations observed between inflation and real a et return in the po t war period. In light of the Fisher hypothe i , and of the commonly held views that tocks and bonds should be a hedge against inflation, these relation are indeed puzzling. Fama (1981), Ge ke and Roll (1983), Ram and Spencer(1983), Stulz (1986), and Kaul (1987) all attempt to explain the negative a sociation between tock return and infla- tion; and Fama and Gibbons (1982) attempt to explain the negative assoc~tions between inflation and real in- tere t rate . Fama hypothesizes that the negative correlation between stock returns and inflation is not a causal rela- tion but that it is proxying for a positive relation be- tween stock returns and real activity and i induced by a negative relation between real activity and inflation. Ge ke and Roll argue that stock returns cau e (or sig- nal) changes in inflationary expectation becau e of a chain of macroeconomic event . When tock price de- cline in response to anticipated change in economic condition , the government, given largely fixed expen- diture , will tend to run a deficit. To the extent that the deficit is monetized, expected inflation will ri e. Ram and Spencer, however, find evidence of unidirectional causality from inflation to tock returns. All of these tudies recognize the importance of cau al relation and di cu s them in one way or another. However, two potential drawbacks remain . Fir t, a Mehra (1978) and Sim (1980) point out, the causal re- lations ob erved ba ed on a bivariate causal te t may not be robu t when other relevant variables are intro- duced into the vector autoregressive (V AR) ystem . Thu , Geske and Rolls (1983) and Ram and Spencer (1983) findings, both of which are ba ed on a bivariate cau al test, may not be robust in a larger ystem of vari- able. Second, while empirical tudie on the U.S . capital market abound, few tudie have been performed on other countrie 1• The di ruption of the international monetary y tern in 1971, and the re ulting introduc- tion of the flexible exchange rate ha permitted inde- pendence of national monetary policies and divergent national inflation rates; this means that different infla- tionary policie may be observed in different countrie wherea this was theoretically not possible under the fixed exchange rates. Indeed, the ame theoretical model need not apply to every country if it is a model uch a Geske and Roll, which relies on the behavior of the do- me tic fi cal and monetary authorities. The purpose of this paper, therefore, i to investi- gate the causal relations among tock returns, interest rates, real activity and inflation u ing the FPE/Multi- variate Grange-Cau ality Modeling Technique for five major industrial countrie over the 1976-1995 period. Our empirical result are con i tent with the reverse causality model proposed by Ge ke and Roll. In par- ticular, we find a strong link between stock returns, a proxy for expected real activity, and the growth rate in the money upply. Moreover, we find that expected change in real activity and money supply growth are important predictor of changes in expected inflation. The evidence, therefore, strongly supports the hypoth- esis that stock return ignals changes in expected in- flation and nominal interest rates. The remainder of the paper is organized as follow : In Section I, we discuss briefly the causal relation pro- posed by Fama and Geske and Roll. In Section II, we de cribe our data. The methodology is di cu ed and result are reported on Section III . Concluding remark are pre ented in Section IV.