The Components of Property Fund Performance

Executive Summary. The evaluation of investment fund performance has been one of the main developments of modern portfolio theory. Most studies employ the technique developed by Jensen (1968) that compares a particular fund's returns to a benchmark portfolio of equal risk. However, such studies implicitly assume that the risk level of the portfolio is stationary through the evaluation period, something that would seem unlikely in an actively managed portfolio. Fama (1972) and Treynor and Black (1973) have suggested that such active management can be separated into the distinct components (1) selectivity (the ability to select undervalued assets), and (2) timing (the ability to adjust security holdings in anticipation of general market movements). The application of simple regression techniques to calculate the Jensen index of performance will, therefore, be biased and any tests of significance distorted (Fama, 1972; Jensen, 1972). Recognising this, a number of researchers have developed models of performance that permit identification and separation of timing and selectivity skills. The simplest and most widely used is the approach developed by Henriksson and Merton (1981). This study analyses the performance of thirty-seven property funds over the period 1987 to 1995 in order to test for the presence of timing and/or selection skills on the part of the fund managers. The results indicate that very few funds appear to engage in timing activities and those that do were generally unsuccessful, while the results for selection skill are much more encouraging.