Seasonality, January-effect, and Momentum in Stock Returns
Abstract
One of the most studied phenomena in empirical finance is the January-effect, which implies that stocks experience significantly larger returns in January compared to the rest of the year. In parallel, a number of investment strategies tend to accumulate most of their annual returns in this month (Haug and Hirschey, 2006).
Although the persistence of the January-effect is often observed, to our knowledge, a sufficient theoretical model of the anomaly has not yet been devised. In this study, we present a number of approaches and highlight their shortcomings. We address the tax-loss selling hypothesis (Reinganum, 1983), the window-dressing hypothesis (Lakonishok et al., 1991) and behavioral approaches (Ciccone, 2011).Furthermore, several empirical results confirm that investment strategies based on past returns are related to seasonal patterns (Haug and Hirschey, 2006, Heston and Sadka, 2008, Yao, 2012). Hence, it follows naturally to address the persistence of the January-effect regarding these strategies. The stock return momentum (Jegadeesh and Titman, 1993; Carhart, 1997) is a commonly examined strategy that tends to perform well in January (Haug and Hirschey, 2006). The explanation of the profitability of return momentum has been linked to many theories such as information uncertainty (Zhang, 2006) or investor overreaction (Daniel et al., 1998). By observing how the January-effect and the momentum-strategies interact, one is able to construct hypotheses that explain the January anomaly based on the momentum-related stylized facts.
Motivated by such considerations, in this study, we explore the linkages between the January-effect, seasonality in general and various momentum-strategies on the Budapest Stock Exchange using monthly stock return data spanning February 2000 to October 2018. We address the main properties of the January-effect: whether there is a robust positive excess return of small-cap stocks in January (Keim, 1983) and whether portfolios based on the stocks' past return perform better in this month (Haug and Hirschey, 2006). We utilize the standard methods of cross-sectional stock return analyses. We apply portfolio-level analyses and firm-level cross-sectional regressions (Fama and MacBeth, 1973). Several empirical findings stand out, indicating mixed results on the persistence of the January-effect in the Budapest Stock Exchange. Small-cap stocks do not appear to achieve any sort of excess return in January. This result is robust to a wide variety of settings and to controlling for other risk factors. On the other hand, some strategies based on short-term past returns (i.e. the formation period is less than a year) display strong January seasonality, which is not observed in other months. Strategies based on longer term past returns (formation period of 2-5 years) appear not to be driven by seasonality patterns, either. In fact, equally weighted long-short strategy based on portfolios with winner and loser stocks based on long term past returns do not provide a return different from zero with statistical significance. Hence, it cannot be concluded that there is a persistent January-effect observable on the Budapest Stock Exchange. However, the lack of the January-effect cannot be concluded, either, as some strategies based on past returns clearly display strong January seasonality.
Our results are in line with previous empirical findings in the Hungarian Stock Exchange. Andor et al. (1999) document the absence of the January-effect in the Hungarian financial markets. Lakatos (2016) finds that short-term momentum strategies provide high returns, however, the strength of the effect diminishes over longer horizons. Mérő et al. (2019) and Csillag and Neszveda (2020) report strong presence of stock return momentum on the Budapest Stock Exchange as well.
We view our contributions to the empirical finance literature as follows. We explore the seasonal patterns of short-term momentum strategies and reveal their strong January seasonality, although we observe no such patterns in the case of long-term momentum strategies. On the other hand, we argue that small-cap stocks on the Budapest Stock Exchange do not experience any sort of January-effect. Our results therefore contribute to the understanding of the January anomaly and how it drives certain strategies. A future strand of research could assess the connections between the drivers of the momentum strategies and the January-effect, such as investor sentiment or information uncertainty.
