چگونه استراتژی های خانواده بر عملکرد صندوق تاثیر می گذارد؟ هنگامی که حداکثر عملکرد تنها بازی ممکن در شهر نیست

This is a first attempt to study how the structure of the industry affects mutual fund behavior. I show that industry structure matters; the mutual fund families employ strategies that rely on the heterogeneity of the investors in terms of investment horizon by offering the possibility to switch across different funds belonging to the same family at no cost. I argue that this option acts as an externality for all the funds belonging to the same family, affecting the target level of performance the family wants to reach and the number of funds it wants to set up. By using the universe of the U.S. mutual fund industry, I empirically confirm this intuition. I find evidence of family driven heterogeneity among funds and show that families actively exploit it. I argue that the more families are able to differentiate themselves in terms of non-performance-related characteristics, the less they need to compete in terms of performance. Product differentiation—i.e., the dispersion in the “services” (fees, performance) that the competing funds offer—affects performance and fund proliferation. In particular, I show that the degree of product differentiation negatively affects performance and positively affects fund proliferation.

The most glaring stylized fact about the mutual fund industry is the existence of a very high number of funds, differentiated into market “categories” and belonging to relatively few families. The number of mutual funds in the U.S. has reached 8,171, more than the total number of stocks traded on NYSE and AMEX added together. Over the period 1990–2000 the number of mutual funds grew from 3,081 to 8,171 while the number of families only slightly increased, from 361 to 431. During the same time, the degree of segmentation of the industry also grew, reaching around 33 different categories. This fact can hardly be explained in terms of the standard finance literature, not only because there already exists a number of securities presumably sufficient to pursue optimal investment strategies, but also because market segmentation makes it harder to improve absolute performance. Indeed, segmentation reduces the scope and range of activity of the manager and forces him to invest only in the assets specific to the fund's category, potentially hampering his market timing skills. Few attempts to address this issue and to model the mutual fund industry have been made. Dermine et al. (1991) lay out a model in which mutual funds locate themselves on the portfolio frontier, together with other primitive assets. Massa (1998) argues that market segmentation and fund proliferation can be seen as marketing strategies used by the families to exploit investors’ heterogeneity. Category proliferation is justified in terms of the positive “spillover” that having a “star” fund provides to all the funds belonging to the same family. Nanda et al. (2000) develop a model of the mutual fund industry in which management fees and load fees are determined endogenously in a competitive setting. Investors’ clienteles and heterogeneity in managerial skills generate different fee structures. Mamaysky and Spiegel (2001) derive from first principles the first equilibrium model of the mutual fund industry in which funds are shown to exist to overcome investors’ hedging needs. They analyze mutual funds as trading devices, set up by investors who cannot remain in the market to trade at all times. Khorana and Servaes (1999) and Khorana and Servaes (2000) empirically analyze the determinants of mutual fund starts, identifying a series of factors that induce the family to set up new funds, such as economies of scale and scope, the family's prior performance, and the overall level of funds invested. Nevertheless, the link between market structure and family strategies has never been directly investigated, neither from a theoretical perspective nor from an empirical one. Nor has anyone studied the relationship between fund performance and market structure. I address this issue by studying how mutual funds behave in a framework in which the interaction between fund family and industry structure matters. I show that fund proliferation is an alternative, non-performance-related strategy devised to make inter-fund comparison harder and to increase market dispersion. In particular, I consider funds as heterogeneous products differentiated in terms not only of fund-specific characteristics but also of family specific ones. Family specific characteristics pertain to the way investors evaluate funds that belong to different families but are otherwise identical in terms of performance and fees. Perhaps the most relevant of these characteristics is generated by the possibility of moving money in and out of funds belonging to the same family at very low cost. This can be seen as an option that the family provides to its investors that reduces the effective fees they pay. The higher the number of funds in the family, the greater the value of this option, because the effective fees decrease as a function of the number of funds. Moreover, the value of this option is investor-specific, as it is a function of investors’ investment horizons: investors who plan to turn over their portfolio more frequently will value it more. By testing this intuition on the U.S. mutual fund industry, I show that investors with a shorter or more volatile investment horizon tend to go for the funds with lower load fees that are parts of big families. I also find evidence of family driven heterogeneity in mutual fund investors’ demand. That is, investors perceive funds as differentiated products, and family affiliation plays a major role in segmenting the market. The existence of this free-switching option affects the degree of competition between funds, as investors evaluate them in terms of the returns provided and the value of the free-switching option: the greater the value of the option, the lower the degree of competition between funds and the greater the segmentation of the industry. Indeed, this option effectively segments the mutual fund industry in terms of family affiliation. Two funds that provide the same return and have the same investment profile may cater to different investors. Their appeal depends on investor-specific (i.e., their investment horizon) as well as family specific (i.e., the number of other funds offered by the same family) characteristics. In this context, I argue that fund proliferation becomes an additional tool that can be used to limit competition and increase market coverage. If the family realizes that the level of performance it offers is much lower than that of its competitors and that it would be very costly to compete on the performance dimension alone, it will focus on other ways of attracting investors, such as reducing fees or increasing the number of funds within the family. It follows that performance-maximization is not necessarily the optimal strategy. In fact, the profit-maximizing mix of fees, performance, and number of funds could even involve a level of performance that would otherwise be defined as “inferior” in a standard performance evaluation analysis. Each fund positions itself on the basis of the fund-specific performance/fee/family combination it offers and targets the investors whose preferences are closer in the product space to the offered combination. The choice among alternative strategies depends on both the structure of the market and its degree of competitiveness. In particular, I argue that the level of performance of a fund will be negatively related to the degree of product differentiation in the category the fund is in, measured as the dispersion in the “services” (fees, performance) that the competing funds offer. If families are able to differentiate themselves in terms of non-performance-related characteristics (e.g., a higher degree of fee differentiation), they have less need to compete in terms of performance. The empirical evidence supports this intuition, suggesting a strong and statistically significant negative relationship between performance and the degree of product differentiation. Moreover, not only is the performance of a family in a category negatively related to the degree of differentiation within that category, but also this negative correlation is present at the category level. That is, categories characterized by higher degree of product differentiation—i.e., a lower degree of competition—systematically provide lower performance. Analogously, I argue that fund proliferation and category proliferation—i.e., the number of funds and the number of categories where the family has funds—should be directly related to the degree of differentiation. In particular, fund proliferation should be positively related to the degree of product differentiation in the category. Indeed, if fund proliferation is an alternative way of attracting investors, fund proliferation should be higher in the very cases in which performance is lower. I find empirical evidence that suggests that product differentiation positively affects the decision of the family to set up a new fund or to enter into a new category. Families that pursue a strategy of category proliferation tend to operate in categories characterized by a higher degree of product differentiation. This effect is present at the category level. That is, fund creation and fund turnover are greater in the very categories in which the degree of differentiation is stronger. The paper is structured as follows. In Section 2, I lay out the framework and derive the main testable restrictions between performance, fund proliferation, and industry characteristics. In Section 3, I describe the data, and in Section 4 I outline the econometric approach. In Section 5, I provide evidence of the degree of competition of the mutual fund industry, by studying whether investment horizon affects investor choice and whether there is evidence of family driven segmentation. In 6, 7 and 8 I test the main working hypotheses in terms of performance and fund proliferation. A brief conclusion follows.

I show that market structure affects mutual fund performance. I argue that performance is only one of the dimensions along which mutual fund families compete. In particular, I show that performance enhancement and category proliferation are alternative ways of attracting investors. The choice between them is directly related to the degree of product dispersion in the market, measured as the dispersion in the “services” (fees, performance) that the competing funds offer. By testing the theory on the U.S. mutual fund industry, I show that the performance of the mutual funds is negatively related to the degree of product differentiation of the category in which they are active, while the degree of category proliferation is positively related to it.