This thesis contributes to an understanding of the influence that recent performance has on investors' choice of outside portfolio manager. The recent growth and importance of professionally managed investment funds is unquestionable, yet very little work has been devoted to understanding the nature of the relationship between investor and outside fund manager. The work conducted here provides important insight into the relationship between past performance and investors' retention/replacement decisions, extending the literature on professional portfolio management in several ways.
Professionally managed funds have been studied extensively. Research has focussed primarily on performance per se: developing methods for the evaluation of performance, and empirical studies of the performance of fund managers through the application of proposed measures. Other aspects of delegated funds management remain
relatively unexplored territory. In studying the investor/investment manager relationship, this thesis provides important insight into problems and resolution mechanisms in delegated investment management.
The work conducted here is aimed at understanding the decisions taken by investors about which fund manager to engage. The study begins with an exploration into the problems faced by investors in selecting and motivating outside investment managers, which leads to the identification of recent performance as an important choice criterion. While the importance of recent performance to fund manager choice has been accepted widely by those interested in delegated investment from either a practical or academic perspective, very few studies analysing and measuring and the relationship have been conducted. Another contribution of this thesis is its investigation into theoretical reasons for the existence of a relationship between past performance and fund manager
choice, and the empirical evidence regarding the relationship.
This study also contributes to knowledge of a little-researched subject through its focus on the institutional investor. A majority of the research into the performance of managed funds has been conducted using data of United States mutual funds, serving individual investors. The tests performed in this study are conducted using data on size and performance of funds serving large investors in Australia, primarily institutions such as superfunds. This thesis contributes to knowledge about the behaviour of institutional investors that can be compared to that of individuals, as well as a generalisation of results based on US data.
The results of the tests conducted in this study provide an unequivocal confirmation of the influence of recent performance on the movement of assets-undermanagement reflective of investors' retention/replacement decisions. A number of different
performance measures in addition to those used in prior work are used here; the robustness of the results across measures allows greater validity to be placed on results of studies using one or two performance measures. In addition to hypothesising and testing a basic relationship between recent performance and the retention/replacement of outside fund manager, several factors are hypothesised to affect investors' decisions, such as fund age and size, as well as transactions costs. The results indicate that investors are more responsive to the performance of small funds than to that of large funds; this also appears true of young versus old funds. This study's data overlap (ie., many of the small funds are also young) predicates the finding of both effects. Results of distinguishing tests indicate that a size effect, rather than age, appears to be present. In addition to size effects, investors appear more responsive to superior performance than to inferior performance. This can be
interpreted as the effect of higher transactions costs associated with replacing poor performers and/or opportunistic behaviour of investors chasing recent stellar performance.
The results reported here generally corroborate those of prior studies in finding a statistically and economically significant relationship between past performance and the movement of assets-under-management. There are two notable differences in the results of this study. First, the evidence indicating support for a size effect as opposed to an age effect, contradicts that of Chevalier and Ellison (1995) who report the opposite. Second, the pronounced asymmetry in the response to superior versus inferior performance reported by Sirri and Tufano (1993) suggests that mutual fund investors appear to flock to recent top performing but do not flee from recent bottom performers. Evidence of this behaviour is not as pronounced here, with many tests indicating that Australian institutional
investors reward recent winners as well as discipline recent losers as predicted by the theoretical models.