When Mean Square Error Becomes Variance: A Comment on “Business Risk and Return: A Test of Simultaneous Relationships”

In a recent article, Oviatt and Bauerschmidt (1991) investigated risk-return relationship by employing the square root of the mean square error of returns as a measure of risk and found no significant relationship existed in those terms. Ruefli (1991) has suggested that under the assumption of stable distributions there is the possibility of spurious correlation in estimating the risk-return relationship in mean-variance terms. This comment identifies the commonalities between mean square error and variance measures, shows that the former measure is subject to many of the problems of the latter, and presents further evidence regarding the likelihood of spurious correlation in industry studies of risk and return. The results suggest an alternative and more parsimonious explanation for Oviatt and Bauerschmidt's findings as well as for the findings reported in the wider strategic management research literature.

• 저널Management Science · 40(6) · 750–759
• 토픽Market Dynamics and Volatility · Economics and Econometrics
• DOI10.1287/mnsc.40.6.750
• 저자Timothy W. Ruefli, Robert R. Wiggins