New Evidence on the Behavior of Canadian Stock Prices in the Days Surrounding the Ex-Dividend Day

Athanassakos, George, Fowler, David, Quarterly Journal of Business and Economics

INTRODUCTION

Ever since Elton and Gruber (1970) used the ratio of the ex-dividend day price change to the dividend per share to infer the typical investor's marginal tax rate and advanced the dividend (tax) clientele hypothesis there has been considerable debate about the interpretation of their results.(1) Miller and Scholes (1982) and Kalay (1982) have challenged Elton and Gruber's dividend clientele hypothesis. They argue that the Elton and Gruber analysis ignores the short-term trades by members of the exchange and by tax-exempt investors whose activities tend to eliminate abnormal returns around the ex-dividend day. The findings of Lakonishok and Vermaelen (1986) support Kalay's short-term trading hypothesis in that trading volume increases around the ex-dividend day, particularly in the period following the introduction of negotiable commissions in 1975. Karpoff and Walkling (1988) provide further support for the short-term trading hypothesis. Their study of stock returns on the ex-dividend day shows that short-term traders are the marginal investors in high yield stocks on the NYSE, particularly since the advent of negotiated commissions. Short-term trading, however, is not evident in low yield stocks at any time.

Foster and Oldfield (1986), on the other hand, find that transactions costs are large enough to limit the ability of short-term traders to effect particular price adjustments. Heath and Jarrow (1988) demonstrate that the ex-dividend day stock price may differ arbitrarily from the dividend for each individual stock; therefore, short-term traders cannot generate riskless arbitrage profits. As a result, ex-dividend stock returns must include a risk premium. Their argument does not rely on transactions costs, but on risk considerations.

Lakonishok and Vermaelen (1983) and Booth and Johnston (1984) have performed similar studies using Canadian data. Lakonishok and Vermaelen examine the ex-dividend day behavior of Canadian stock prices around the introduction of a tax reform in 1972 and conclude that the Elton and Gruber tax clientele

hypothesis is invalid. Booth and Johnston, on the other hand, investigate the ex-dividend day behavior using the Elton and Gruber methodology over four distinct tax regimes between 1970 and 1980. They find that the marginal tax rate implied by the ratio of the ex-dividend day price change to the dividend is greater than the maximum marginal tax rate payable. The only firm conclusion they are able to reach is that the ex-dividend day price changes for stocks interlisted on Canadian and U.S. exchanges are determined by U.S. rather than by Canadian investors.

Green (1980) extends the Elton and Gruber analysis by introducing costs of delaying or accelerating a transaction. His work sets the theoretical foundation for the development of the model of delay and acceleration of trade in response to different tax rates. Grundy (1985) extends and tests Green's model by examining the extent to which investors accelerate or delay trades in response to the payment of the dividend. He argues that a tax-induced incentive to either delay or accelerate trades would result in specific patterns of abnormal returns around the ex-dividend day. He shows that when capital gains are taxed on realization rather than accrual, the stock price behavior is affected not only on the ex-dividend day, but also during a few days around it. His empirical results lead him to conclude that the evidence is consistent with the predictions of the model of delay and acceleration of trade and the existence of dividend clienteles.

This study tests the clienteles and short-term traders hypotheses employing a modified version of the model of delay and acceleration of trade to Canadian data over different tax and transactions cost regimes from 1970 to 1984. There were significant differences between the structure of tax rates and transactions costs in Canada and in the U.S. over the period studied. In addition, there were several different tax regimes in effect in Canada between 1970 and 1984, each of which had different implications for the taxation of dividends and capital gains. Between 1970 and 1984 there were six major changes in the taxation of dividends and capital gains. Over the same period there were also regulatory changes in transactions costs that started in 1976 and led to initially lower transactions costs for large trades and then to fully negotiable commissions in 1983. If taxed investors are the marginal traders, as in the Elton and Gruber model, then stock prices around the ex-dividend days should be affected only by tax changes, not by the reduction in transactions costs. If transactions costs changes impact stock prices around ex-dividend days, however, then taxed investors will not be the marginal traders; hence, the Elton and Gruber tax clienteles argument is invalid. Canada, therefore, provides an excellent testing ground for the clienteles and short-term traders hypotheses that otherwise may be difficult to detect.

This paper, based on the argument that untaxed/short-term investors(2) are the dominant and marginal traders, develops and tests hypotheses on how abnormal returns should behave under both the clienteles and short-term traders hypotheses over the different tax and transactions cost regimes in Canada between 1970 and 1984. The evidence supports the modified model of delay and acceleration of trade and is consistent with the short-term traders hypothesis. Short-term traders/untaxed investors are shown to transact around the ex-dividend days with the intention of capturing or avoiding dividends subject to the prevailing tax and transactions cost regime in order to maximize their after tax returns. This gives rise to abnormal returns on dividend paying stocks in the days surrounding the ex-dividend days. Moreover, the higher the stock's dividend yield, the greater is the abnormal return during these days.

THE CANADIAN TAX AND TRANSACTIONS COST REGIMES

CANADIAN TAX RATES AND TAX REGIMES(3)

Canada's tax system is unusual in at least two respects. Dividends have been given special treatment in order to minimize the effects of double taxation since 1949, and capital gains taxes were introduced only at the beginning of 1972.(4) As a result, the difference between the tax on dividends and that on capital gains was larger in Canada than in the U.S. before 1972. The situation was reversed, however, subsequently.

Dividends in Canada have been taxed according to the following formula:(5)

(1) Pre-1972: T = |t.sub.fp~ - |C.sub.fp~

Post-1971: T = (1 + G) (|t.sub.f~ - C) (1 + |t.sub.p~)

where:

T = Tax payable per dollar of dividend;

G = Gross-up factor;(6)

|t.sub.f~ = Marginal federal tax rate;

C = Dividend tax credit;

|t.sub.p~ = Marginal provincial tax as a (constant) percentage of the federal tax payable;(7)

|t.sub.fp~ = Combined federal/provincial marginal tax rate;

|C.sub.fp~ = Combined federal/provincial dividend tax credit.

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: New Evidence on the Behavior of Canadian Stock Prices in the Days Surrounding the Ex-Dividend Day. Contributors: Athanassakos, George - Author, Fowler, David - Author. Journal title: Quarterly Journal of Business and Economics. Volume: 32. Issue: 4 Publication date: Autumn 1993. Page number: 26+. © 1999 University of Nebraska-Lincoln. COPYRIGHT 1993 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.