주식형 펀드의 성과와 현금흐름에 관한 새로운 시각

Net flows of equity funds are a measure of advisory firm’s profitability as well as investors’ decision making. Previous studies have attempted to understand investors’ investment decision making by studying fund net flows. Among the determinants of net flows, the most important is the past return of an equity fund. Chevalier and Ellison (1997), Sirri and Tufano (1998), Lynch and Musto (2003), and Huang et al. (2007), in both empirical and theoretical senses, show that better-performing funds attract more cash flows disproportionately than worse-performing funds. This is called “convex” flow-performance relationship of equity funds. The convex relationship can be interpreted in two ways. First, the worst-performing funds do not lose cash flows as much as expected by the linear flow-performance relationship. Second, best-performing funds attract extraordinarily high cash flows than the other funds. For this reason, fund managers and advisory firms are trying to enhance fund returns in a fair or an unfair way to induce more cash flows. More cash flows are directly related to better profitability of advisory firms and more payoff to fund managers. Consequently, the convex flow-performance relationship may cause an agency problem between fund investors and managers (or advisory firms). This study investigates the equity fund flow-performance relationship from a new point of view. To investigate this convex relationship, Spiegel and Zhang (2013) employ the concept of market-share changes while Chevalier and Ellison (1997) and Sirri and Tufano (1998) use net flows of equity funds. Net flows are a conventional measure of cash flows that is used by most studies on this issue, however, in a practical sense, market-share changes are a more directly related measure to advisory firms’ profitability. This study use both measures to study the flow-performance relationship. On the other hand, previous studies assume the homogeneous volatility of cash flows within the same performance domain. Unfortunately, this is not true. Heterogenous volatility of cash flows may deter us from estimating the correct flow-performance relationship. To resolve this issue, we employ quantile regression developed by Koenker and Bassett (1978). Quantile regression allows us to consider heterogenous responses of fund cash flows to past performance and to identify the characteristics of each fund quantile. Last but not least, previous studies have not been interested in the characteristics of the lowest and highest quantile fund groups. If an agency problem is related to this flow-performance relationship, the consequences should be observed in the characteristic analysis. Empiricl findings are summarized as follows. First, regardless of the measure of fund cash flows, domestic equity funds show an asymmetric convex relationship between past perfomance and fund flows. Second, quantile regressions reveal that the response of change in market share is different among quantile fund groups. In terms of market-share changes, the convex relationship is strongly affected by high quantiles in the high-performance domain, and by low quantiles in the low-performance domain. Third, in the case of changes in market share, the fund size of the highest quantile is much larger (305 billion won) than that of the lowest quantile, which means that, in the high-performance domain, large funds are easier to advertise and promote than small funds. Such a phenomenon can be interpreted as an agency problem between fund managers (or advisory firms) and investors concerning artificial enhancement of fund performance. This study contributes to the extant literature in that it attempts to take the heterogenous fund cash flows into account to understand the response of fund investors to past performance.