Modeling Mutual Fund Flows and Fund Risk from a Spatial Perspective
29th EBES Conference - Lisbon
Mutual fund market has reached a significant size in the last 10 years worldwide. Besides being a way to channel resources among economic agents, their increased role in the financial markets has drawn attention of the academics. These studies mainly show that the relation between mutual fund flow and past performance is not linear but convex in shape. This study aims to reinvestigate this long-established convex fund flow-past performance relation from a spatial modeling perspective. More particularly, we aim to include the effects of rival/neighboring funds’ position on a risk-return analytical surface when evaluating the above-mentioned relationship for Portuguese equity mutual funds. The studies examining fund flow-past fund performance indicate that fund managers must compete with each other to be among winners and to attract more investors to the fund. In other words, their compensation is directly linked to the fund’s “relative position in the market”, which causes a constant ranking inside the industry (i.e. Chevalier and Ellison, 1997; Brown et al. 1996). This payment scheme creates incentives for fund managers to take excessive risk in order to beat their “rivals”. A recent study also notes that even there are no financial reasons, fund ranking is still important for fund managers for a positive self-image or public status (Kirschler et al., 2018). This study refers to these non-financial incentives to outperform the “rivals” as ranking incentives. Although “the ranking” is determined according to the funds’ position, to the best of our knowledge, no study has considered its effects. Likewise, Akerlof (1997) criticizes influence of others while making decisions. The same criticism is valid for the studies examining mutual fund investors’ behavior. Therefore, we try to explain how the existence of neighboring funds affect the decisions of fund investors and managers in an industry where the funds are evaluated according to their positions. This study begins with regressions which serve as a benchmark. It benefits from 3 different types of spatial regressions in modeling both fund managers’ and investors’ behaviors. The distance among funds and the neighbors will be determined by data envelopment analysis (DEA). Computing distances in such a way also contributes the literature by showing an alternative usage of DEA. The sample includes all the domestic equity mutual funds (liquidated or active) in Portugal between the years 2010 and 2015. Further studies will extend this study to a worldwide sample and compare the results for international fund markets.
mutual fund cash flows,mutual fund performance,neighboring funds,spatial econometrics,DEA