Market Timing Abilities of Fund Managers: Parametric and Non-Parametric Tests
Abstract
The ability to measure the investment performance of fund managers is a perennial issue for potential and committed investors. The performance of fund managers influences the manner in which investors place their investible wealth and their choice of fund managers. Obviously, it also affects the compensation of the fund managers (see Kershot, 1978; and Smith, 1978). Past studies on the subject of investment performance have attempted to draw the distinction between the ability to 'time' the market and the ability to 'select' the right security to obtain high portfolio returns. These have been termed 'timing' and 'selectivity' abilities respectively. This distinction has been viewed as a useful device for the attribution of a manager's performance. An ability to distinguish between these two sources of superior performance allows a better assessment of the services provided by fund managers. This study focuses on the assessment of market timing abilities of fund managers in Singapore. It addresses the approach in measuring fund managers' abilities to forecast price movements of the general stock market as a whole and in the identification of under- or overvalued equities relative to fixed-income securities. It builds on the theoretical model on market timing developed by Merton (1981) and empirically extended by Merton and Henriksson (1981). The latter paper derived the statistical framework for non-parametric and parametric tests of timing ability used in this study. Using the same data set, this study used both parametric and non-parametric tests to examine the performance of Singapore fund managers. The results obtained are consistent with past studies. In particular, it confirms the theoretical studies by Admati and Ross (1985) and Jagannathan and Korajczyk (1986) who argued that the parametric test of market timing can give erroneous results. [ABSTRACT FROM AUTHOR]