The Ex-Dividend Pricing of Real Estate Investment Trusts
Hardin, William G.,, III, Liano, Kartono, Huang, Gow-cheng, Real Estate Economics
Past studies have shown that ex-dividend stock prices are not fully reflective of dividend payments. A tax-induced clientele effect and micromarket limitations in stock pricing have been used to explain this pricing anomaly. This study focuses on the ex-dividend behavior of real estate investment trusts (REITs). Due to a low correlation between dividend size and dividend yield, REITs permit a cleaner examination of a tax-induced clientele effect. The results indicate that tick constraints in pricing ex-dividend stocks create the appearance of a tax-induced clientele effect in REITs when none should exist.
Financial market theory postulates that ex-dividend stocks should have a concurrent reduction in value reflecting the payment of dividends. However, documenting this reduction has been problematic. The inability to empirically demonstrate ex-dividend pricing efficiency has produced additional theories to explain abnormal returns and repricing anomalies surrounding ex-dividend dates. Perhaps the most cited work on the topic is Elton and Gruber (1970). Expanding studies by Campbell and Beranek (1955) and Durand and May (1960), Elton and Gruber show that the change in a stock's ex-dividend price is less than the dividend amount. They interpret this result as evidence of a marginal tax-rate-induced clientele effect, as postulated by Miller and Modigliani (1961).
In response to Elton and Gruber, subsequent studies have investigated the effects of changes in tax policy (Grammatikos 1989 and Michaely 1991) and transaction costs (Eades, Hess and Kim 1984, Boyd and Jagannathan 1994 and Frank and Jagannathan 1998) on the ex-dividend pricing of stocks. Miller and Scholes (1982) provide the framework for much of this research by highlighting the incongruence between a tax-induced clientele effect as modeled and defined by Elton and Gruber and the ability of traders to arbitrage away any possible tax effect. The goal of this subsequent research has been to control for the extraneous variance in the data to allow for other explanations of ex-dividend stock pricing anomalies.
The present study differs from prior work because of the use of a unique subset of stocks: real estate investment trusts (REITs). REITs are characterized by high-dividend yields and little correlation between dividend size and dividend yield (0.05). (1) Because REITs are required to pay out a significant percentage of earnings in the form of dividends to retain their favored tax status, they have limited ability to generate substantial retained earnings for reinvestment. Thus, dividend yield is an important determinant of REIT value. Also, unlike other high-yield stocks such as banks and utilities, where regulations foster retaining and reinvesting capital to reduce financial risk, REITs are required to reduce their internally generated capital regardless of actual company-level reinvestment opportunities. This creates a situation in which dividend payments constitute a substantial portion of total investment return. (2)
The use of REIT stocks has the additional benefit of making typical market participants subject to similar tax consequences. As Lee and Kau (1987) point out, dividends provide the majority of return for REIT investors. Long-term investors purchase REIT stocks knowing that they will receive a high dividend yield during their ownership term with some potential for capital appreciation. (3) This implies that the majority of their returns will be taxed as short-term income. Market makers and dividend payment arbitragers are also subject to similar short-term or current income tax consequences. (4) Consequently, investors, market makers and arbitragers generally face similar tax consequences for investing in REIT stocks.
This study incorporates and expands the work of Bali and Hite (1998) on the relative importance of tick constraints in the pricing of ex-dividend stocks. …