Financial markets and anomalies – the weekend effect

This paper’s main objective is for one to test whether the weekend effect in the context of financial markets, i.e. higher stock returns on Fridays and lower stock returns on Mondays, is present among nine selected European stock indices under study and for another to examine whether investor sentiment is able to explain this market anomaly. The presented statistical analyses reveal that the traditional interpretation of the weekend effect, which implies the direct comparison between Monday and Friday returns, is of existence in the data under study as I find that Monday returns are significantly lower compared to Friday returns, ceteris paribus. Even though this finding is statistically significant at the 5% level, the magnitude of the effect is rather low as the fixed effects regression outputs show that the estimated difference in returns between Monday and Friday lies below 0.1%. Furthermore, I provide statistical evidence that investor sentiment as indirectly measured by average daily returns of German government bonds with a duration of 10 years to a limited extent is capable of explaining the weekend effect from a behavioural point of view. Based on a popular three-step approach for mediation testing, I find that average daily returns of German government bonds with a duration of 10 years serve as a valid mediator variable for the weekend effect. Additional effect path analyses based on structural equation modelling (SEM) reveal that the degree to which this indirect proxy of investor sentiment drives the weekend effect lies below 5% and thus can be considered as rather insignificant from an economic perspective. This paper provides interesting insights into weekly patterns in stock returns among indices that have experienced little to no interest by researchers in the context of examining the weekend effect to date. Similar to the well-document weekend effect among US stocks, the detected traditional weekend effect among selected European stock indices is small in size. Therefore, trading strategies that aim to exploit this market anomaly are likely to turn out unprofitable since trading costs and noise trader risk would offset a possible, small gain. Intraday analyses of stock returns, as well as the application of direct measures of investor sentiment within appropriate statistical models, constitute promising areas for future research in regards to the weekend effect that would provide further meaningful insights into this puzzling anomaly.