Asset pricing model parameters and infrequent trading in an emerging market

Abstract

Direct empirical evidence on the relationship between the asset pricing model parameters and infrequent trading in emerging markets seems sparse at best. This thesis provides empirical evidence on the potential impact of infrequent trading on the estimated asset pricing model parameters in Thailand. As a comparison sample, Singaporean data are also employed. Three main empirical results are found in this thesis. First, there is a pattern indicating that the alphas without adjusting for infrequent trading (unadjusted alphas) may be upwardly biased and the betas without adjusting for infrequent trading (unadjusted betas) may be downwardly biased in both the Thai and Singaporean markets. Second, although the differences between the unadjusted alphas and betas and the adjusted alphas and betas are statistically insignificant, the differences between the unadjusted alphas and the adjusted alphas are economically significant. Third, the serial correlation problem in portfolio returns is alleviated after adjusting for infrequent trading. Hence, it seems worthwhile to adjust for the impact of infrequent trading on the asset pricing model parameters in both Thailand and Singapore.