PORTFOLIO PERFORMANCE AND THE INTERACTION BETWEEN SYSTEMATIC RISK, FIRM SIZE AND PRICE-EARNINGS RATIO: THE CANADIAN EVIDENCE Said Elfakhani* While some American studies relate portfolio performance to PIE ratios, others reject such a hypothesis orfind evidence ofa confounded PIE-size effect. This prevents conclusive inferences. Canadian markets are structurally differentfrom American ones; thus American evidence may not apply to Canadian stocks. This study examines how interaction between PIE ratio, beta andfirm size affects the portfolio performance of Canadian stocks. The results show a relative support for the firm size effect, even after proper adjustment for risk and alternate change in control variables. This evidence is not uniform across different quarters of the year but not restricted to year-end effect. The findings also demonstrate a positive correlation among the three variables. However, one cannot generalize conclusions since the analysis may not capture all other pertinentfactors. I. Introduction The performance superiority of stockportfolios characterized by low price-earnings (PI E) ratio continues to be a debatable issue. On the one hand, risk-adjusted excess returns are a violation of the semi-strong form of the efficient market hypothesis. One argument advocates that lack of proper tests that correctly adjust for risk may have led to misleading results. Similarly, the PIE ratio may be acting as a proxy for other pertinent variables. On the other hand, many portfolio managers favorably view stocks with predictableperformance and are, therefore, willing to pay generously for them. Thus, they often hold portfolios of high PIE ratios stocks expecting to achieve targeted returns at acceptable levels of risk. Such a controversy leaves room for further theoretical and empirical investigation. Basu (1975, 1977) finds that returns on a portfolio of low PIE ratio stocks tend, on average, to earn higher absolute and risk-adjusted rates of return than higher PIE portfolios. This holds after controlling for infrequent trading bias (Goodman and Peavy, 1983 and 1985), but only if transaction costs are minimal (Levy and Lerman, 1985). * Associate Professor of Finance, Universityof Saskatchewan,Saskatoon, Saskatchewan, Canada, S7N OWO. Telephone (306) 966-8427, fax (306) 966-8709.1. The author would like to thank the Social Sciences and Humanities Research Council of Canada ResearchFundandthe Department of Financeand Management Science,University of Saskatchewan, for financialsupport. Helpfulinsightsof anonymousrefereesof thisjournal and valuablecomments by the seminar participants at the Financial Management Association meeting, 1991, are also acknowledged. The author, however, is responsible for any remaining errors.