This study explores the “Sell-in-May” effect in environmental, social and governance (ESG) indices and compares the seasonal effects in ESG equity indices with conventional equity indices.
The authors use ordinary least squares (OLS) models and M-estimation as a robustness check, as OLS estimates may be sensitive to outliers. The authors also employ bootstrap simulations to use the data efficiently and to test whether seasonal trading strategies can produce abnormal returns.
The regression results reveal that seasonal effects in USA ESG equity indices are similar to those in conventional equity indices. Higher returns are noticeable from November through April, mainly in ESG indices including small and medium capitalization stocks. When the authors extend the Sell-in-May strategy from October through April, the authors find that the seasonal effect is significant for multiple ESG indices, even after accounting for the January effect. Bootstrap simulations show that the Sell-in-May and Extended Sell-in-May strategies appear to beat a buy-and-hold strategy on a risk-adjusted basis and that this result is stronger in medium and small capitalization ESG indices.
Although previous research has considered the effectiveness of seasonal equity trading strategies and the general performance of ESG stocks, this is the first study to specifically examine the “Sell in May” effect in ESG indices. The authors also consider an “Extended” Sell-in-May strategy where stocks are purchased one month earlier and show that the strategy produces higher returns.
