THE EFFECT OF THIN TRADING ON MARKET PARAMETERS IN THE SINGAPORE MARKET AND A TEST OF SECURITY PRICING

Abstract

Financial theories assume that the stock market is perfect and continuously observable. However, in practice, most stock markets suffer from varying degree of infrequent trading, or thin trading. Thin trading introduces biases in the measurement of market parameters, one of which is the popular systematic risk or beta factor. The problem of thin trading in the Singapore Stock Market has not been given much attention in most of the financial research done on the local market. The objective of this ASP is to study the various published theories for thinness correction by applying them to the correction of beta measurement in the Singapore Stock Market for the period 1975 through 1988. This ASP examined the thinness correction models by Scholes and Williams, Dimson, and Fowler and Rorke, with one, two and three periods lead/lag . Where necessary, the theories were re-examined to develop some of the unpublished formulae. The study was carried out over the whole period (1975 - 1988), and three subperiods (I: 1975-79, II: 1979-83, III: 1983 -88). The results revealed that all the three models provide more or less similar correction to thin trading in the Singapore Market. For the whole period, a lead/lag of I three periods works best for the Scholes-Williams model, while for the Dimson and Fowler-Rorke model, a lead/lag of two periods is better. For subperiod III, a lead/lag of three periods works best for all the three models. Variations of beta by deciles of trading frequency and industry sectors were also studied. In the test of Capital Asset Pricing Model, it was observed that the corrected data supports the Capital market theory, and fits the Security Market Line better than the ordinary least square data. It is concluded that correction of thin trading is necessary for the Singapore Stock Market. For economic calculation and use of computer resources the Fowler-Rorke model is recommended.