The Effect of Intervaling on Estimating Parameters of the Capital Asset Pricing Model

Empirical research has played an important role in recent theoretical developments in the theory of finance, particularly in the formulation and testing of various theories of capital asset pricing. A common procedure in much of that empirical research is to use historical price and dividend data to estimate the parameters of a characteristic line which relates the return on an asset or portfolio to the return on the market. While several possible limitations of such procedures have been explored, one recurring question is the appropriate length of each interval used in the estimation. The purpose of this study is to investigate intervaling in greater detail so as to better understand its impact on the results of empirical research and hence of further developments in the field of finance. This is accomplished by examining the effect of different intervals on the return distributions and estimated characteristic lines of 200 common stocks over the two decades 1950–1969. Section II reviews the relevant literature and attempts to place the intervaling effect in perspective. Research design for the investigation is described in Section III, and findings are presented in Section IV. A brief conclusion appears as Section V.