Searching for seasonal patterns in exchange traded funds' trading characteristics

Abstract

This paper investigates the seasonal characteristics of exchange traded funds (ETFs) return, risk, tracking error and volume and reveals the existence of a strong November effect in performance. On the other hand, this study finds that the well-known January effect does not affect the performance of ETFs. Moreover, this paper demonstrates that a semi-strong seasonality effect on ETFs' risk exists in November and that ETFs achieve their best index replication in this month. The combination of substantial average performance and low average risk and tracking error signals an opportunity for investors to gain sufficient returns by exposing themselves in modest or low volatility and tracking failure. A straightforward relationship between risk and tracking error is also revealed. Finally, the study indicates that the trading activity of ETFs is lacking in any seasonal pattern but there is some evidence on the direct conjuncture between risk and volume. This connection implies that ETF investors sell their shares when their investments in ETFs are over risky.