A comprehensive analysis of calendar anomalies in mature and emerging markets

Abstract

The research begins with a comprehensive study on 4 major calendar anomalies at industry level in the US market for 90 years which include a break point period in 1952. Daily returns are examined for the weekday, turn-of-the-month, January and Halloween effects and the results confirm that the effects of these anomalies exist and persist uniformly across almost all industries in both periods before and after the break point. Hence, calendar effects are driven by economic events affecting all industries rather than by industry-specific factors. The thesis starts with the US market given the maturity of the market and the extended data available which will provide an overall understanding of the topic of calendar anomalies. To delve deeper into the topic of calendar anomalies the research attempt to investigate the famous weekday effect in an emerging market, Saudi Arabia. This will give a deeper understanding of the topic since investors attributed to this market have many behavioural aspects that could affect the anomaly like cultural and religious beliefs. The research studies the existence of the anomaly in 15 industries in the Saudi stock market by applying a break point in June 2013 where there was a change in the weekend days. The findings confirm the existence of the anomaly pre-June 2013 only, providing evidence that the break point event of changing the weekend days directly affected the anomaly. From this standpoint, the research continue to explore the pre-holiday effect in Saudi Arabia to further investigate the effect of culture and religion on calendar anomalies. The study examines the anomaly in both market and industry level to confirm whether the effect is limited to certain sectors or is a wide-market phenomenon affecting all industries similarly. The data covers daily returns for both the general market and 15 industries over a period of almost 11 years, from 2009 to 2020. The findings confirm the existence of the pre-holiday effect at the general market and industry level for religious holidays, however, there is no evidence found on the existence of the anomaly on non-religious holidays.