A re-examination of the behavior of common stocks on ex-dividend days.

Abstract

Over the past thirty-six years most of the research on the ex-dividend day price behavior of common stocks has centered on either the tax-induced clientele hypothesis or the short-term trader hypothesis. However, neither theory has fully explained why, on average, the price drop-off on the ex-dividend day is less than the declared dividend. This paper investigates both of these hypotheses and introduces a new explanation. The tax-induced clientele hypothesis is considered by replicating the Elton and Gruber 1970 study and by examining the effects of the 1986 Tax Reform Act on ex-dividend day price movements during the years 1984 through 1988. The replication of the Elton and Gruber study in Chapter 2 will provide results contrary to the tax-induced clientele hypothesis. Also, the tax-induced clientele hypothesis cannot fully explain the price behavior of common stocks on ex-dividend days during the 1984 through 1988 period, as shown in Chapter 3. The short-term trader hypothesis is examined in Chapter 4 by determining if the actions of the short-term traders are truly riskless as proposed by Kalay (1982). If so, the price drop-off will be examined to see if these short-term traders are able to "arbitrage" the excess profits up to their marginal transaction costs. As with the tax-induced clientele hypothesis, the results will show that the short-term trader hypothesis cannot explain the ex-dividend day price movements of common stocks. A new explanation, a risk shift hypothesis, is considered in Chapter 5. This hypothesis proposes that the excess profits earned on the ex-dividend day are merely premiums for taking on more risk by holding securities during the ex-dividend day period. Unfortunately, the results do not support this theory.