Volatility of holiday effects in Thai stock market

Abstract

This paper aims to examine the volatility of holiday effects on Thai stock market. The holiday effect is phenomenon in which high returns could be found around the holiday. The event provided new insight about the weak-form efficiency in the market. The SET index were collected five day daily data from the 1st January 1992 until 31st December 2016. The size of the data was 6,523 days. The holiday data were collected from 1992 to 2016. The holidays, announced by the Bank of Thailand, were 317 days along 25 years. This paper tested the holiday effect by using GARCH(1,1) model and EGARCH(1,1) model both of which are considered as appropriate for time series data. The result showed that there are statistically significant positive higher returns rate than normal days in both pre-holiday and post–holiday. Indeed, we found the abnormal positive returns in pre-holiday is higher than post–holiday at the significant level of 1%. Finally, we tested the model fit to data of two methods by using AIC and SIC. From the result it appeared that EGARCH(1,1) is more appropriate than GARCH(1,1).