On Mutual Fund Family Diversification, Performance, Persistence and Flows

The first essay introduces a portfolio theory motivated approach to measuring
mutual fund family-level diversification and hedging strategies. Diversification of
idiosyncratic risk (systematic risk) is measured by the average cross-fund correlation in
idiosyncratic returns (predicted returns from the multifactor model). Using new
methodology, I find evidence of cross-sectional variation in family-level diversification
and examine several fund families’ characteristics as the determinants of this crosssectional
variation. On average, fund families that offer more objectives are more
diversified in terms of both idiosyncratic and systematic risks; however, in the subsample
of larger fund families, greater number of objectives is associated with increase
(decrease) in idiosyncratic (systematic) risk diversification. Families that concentrate in
the retail sector are more diversified. I also find that less diversification of idiosyncratic
risk on the family level is associated with better risk-adjusted performance, while greater diversification of systematic risk is associated with greater performance during an
economic downturn.
The second essay examines whether new measures of diversification are
additional determinants of fund family flows and flow volatility. I find that fund family
capital flows increase in systematic risk focus, as more of the fund family’s assets are
held by institutional investors. Family flow volatility decreases in diversification of
systematic risk during market downturn, increase in market uncertainty and during
recession. I further find that families with greater concentration in the retail sector
(institutional sector) exhibit less family capital flow volatility as the diversification of
systematic risk (idiosyncratic risk) increases. Fund-level volatility of focused and
concentrated funds within diversified families is greater than in less diversified families,
signaling that diversification on the family level may decrease participation costs for the
investors. Moreover, in support of participation cost hypothesis, I find that the
performance of worst performing funds within fund families increases in the family-level
diversification; thus, family-level diversification affects the convexity in the fund flowperformance
relation documented in the previous studies. On the family-level,
diversification is associated with convexity in flow-performance relation, while family
focus with more direct flow-performance relation.