New evidence on the impact of size and taxation on the seasonality of UK equity returns.

1. Introduction and objectives of the study

The empirical phenomenon of high monthly equity returns in January relative to other months of the year was first documented by Wachtel (1942) for the New York Stock Exchange (NYSE) over the period of 1927-1942. Rozeff and Kinney (1976) confirmed Wachtel's finding of a January seasonal from 1904 to 1974. Keim (1983) concluded that about half of the January seasonal was associated with small firms during the period of 1962-1979. Reinganum (1983) investigated the year-end tax-loss selling hypothesis that had been advanced by Wachtel (1942) to explain the unusual January returns. He noted that tax-loss selling of small capitalization stocks explains a significant part of the high January returns. The interaction between firm size and tax-loss selling in January was refined by Ritter (1988). He argued that individual investors who primarily invest in small capitalization stocks tend to sell them in December to realize capital losses and buy into such stocks in January. Eakins and Sewell (1993) and Ligon (1997) found evidence consistent with this hypothesis.

Haugen and Lakonishok (1988), on the other hand, argue that January seasonality could be induced by fund managers rebalancing their portfolios to either hedge or window dress their performance. In performance hedging, fund managers lock in their gains, especially from small stocks, and move into stocks that will closely replicate the features of their performance benchmark during the course of the calender year. They re-enter the market after the last reporting day to acquire undervalued small stocks that consequently appreciate in price at the start of the calender year. In window dressing, fund managers include stocks that have done well in the recent past and eliminate under-performing small stocks for year-end reporting purposes. At the start of the new reporting period, managers choose stocks, especially those of small firms that are likely to provide higher returns during the year. The prices of small stocks therefore rebound at the start of the next reporting period. Evidence consistent with this hypothesis has been reported by Ritter and Chopra (1989) and Porter et al. (1996) for the US and by Athanassakos (1992) and Athanassakos and Schnabel (1994) for Canada.

In light of the above, it is clear that tax-loss selling and portfolio rebalancing could interact to generate January seasonality in markets with a December tax year-end. This makes it difficult to test the two hypotheses independently on the same data, as is usually done in US studies. This paper contributes to the literature by using data from the UK, which has a tax year (in April)(1) that is different from the calender year, to separately test the tax-loss selling and portfolio rebalancing hypotheses. Specifically, the portfolio rebalancing hypothesis is tested as an explanation of January seasonality, while tax-loss selling and window dressing explanations are investigated for April seasonality. This contrasts with previous UK papers that either focused on tax effects (Reinganum & Shapiro, 1987) or size effects (Levis, 1985; Corhay et al., 1988; Clare et al., 1995) to the neglect of other explanations of January and April seasonalities.

The next section describes the evolution of UK capital gains taxation and assesses its impact on tax related explanation of seasonality. The data and sample selection methods are described in section 3. Using a GARCH-M model, section 4 identifies the months and portfolios that show evidence of seasonality during the sample period. Section 5 examines the extent to which January and April seasonalities could be explained by either the portfolio rebalancing hypothesis or tax-loss selling. A summary of the study and its findings are presented in the final section.

2. UK capital gains taxation

Compared to the US, which introduced capital gains taxation in 1913, it was not until half a century later that a short-term speculative gains tax became law in the UK.(2) A comprehensive capital gains tax was introduced in 1965 after years of public debate.(3) The major capital gains tax law changes covered by this study and their possible effects on …