Journal zyxwvutsrq of Buriness Financc zyxwvuts &Accounting, zyxwvu 18(5), September 1991, 0306 686 X $2.50 zy MACROECONOMIC FACTORS AND THE UK STOCK MARKET s. POON AND zyxwv S.J. TAYLOR* INTRODUCTION Chen, Roll and Ross (CRR) (1986) claimed recently to have found exogenous economic factors in a multivariate Arbitrage Pricing Model. Using US data for the period 1958 to 1984, CRR asserted that sources of risk from industrial production, changes in a risk premium for bonds, and the term structure are significantly and systematically priced in the stock market. Measures of unanticipated inflation and changes in expected inflation have some influence as well, but only when these variables were highly volatile. More surprisingly, they reported that the value-weighted New York Stock Exchange Index, although explaining a significant portion of the time series variability of stock returns, has an insignificant influence on pricing when macroeconomic factors are also considered. CRR is the first, in a series of recent studies,' to employ specific macroeconomic factors as proxies for the otherwise theoretically undefined state variables in the Arbitrage Pricing Model. The earlier literature on tests of the APM is based on using factor analysis to extract systematic factors influencing stock returns.' CRR compared time series of the values of five common factors, obtained from factor analysis, with macroeconomic variables based on economic theory. These macroeconomic variables are assumed to have influenced either future cash flows or the risk-adjusted discount rate, two key variables when stocks are priced by the expectation of the present value of future cash flows. They report in their footnote 7 that the null hypothesis that each of the macroeconomic factors is not related to any one of the five common stock factors is rejected in every case, except for the case of inflation. If the claims in CRR are true then an implication of their findings is that the Arbitrage Pricing Model (APM) is superior to the Capital Asset Pricing Model (CAPM). Given that there is so much interest, emphasis and effort from both practitioners and academics in computing and assessing security CAPM betas this implication is important. The two objectives of this study are to reconsider the results in CRR and to see if they are applicable to UK stocks. The CRR methodology and the The authors are respectively, Senior Lecturer in the School of Accounting and Financial Services, Lancashire Polytechnic; and Reader in the Department of Accounting and Finance, Lancaster University. They are very grateful to Professor K.V. Peasnell, Professor C.W.R. Ward and the anonymous referee for several helpful suggestions on an earlier draft of this paper. (Paper received November 1988, revised November 1989) 619