We explore the properties of equity mutual funds that experience a loss of assets after poor performance. We document that both inflows and outflows are less sensitive to performance, because performance-sensitive investors leave or decide not to invest after bad performance. Consistent with the idea that attrition measures the sorting of performance-sensitive investors, we find that attrition has less of an impact on the fund’s flow–performance sensitivity for institutional funds where there is less dispersion in investor performance sensitivity. Also, attrition has no effect on the flow–performance sensitivity when attrition arises after good performance or investors invest for nonperformance reasons.

Firms whose human capital is concentrated in a few irreplaceable employees lack diversification in their human capital stock, exposing them to key human capital risk. Using disclosures of “key man life insurance” to measure this risk, we show that exposed firms are riskier. These younger, smaller, growth firms have abnormally high volatility, and following announcement of key employee departures, the most exposed firms lose 8% of their value. Key employees tend to be highly educated. They are four times more likely to hold PhD degrees than top managers, and firms with key human capital are more innovative.

We study data from an organization in which fund managers privately share and discuss detailed investment recommendations. Buy recommendations generate positive abnormal returns, and sell recommendations result in negative abnormal returns. In the context of these results, we explore an important economic question: Why do skilled investors share profitable ideas with others? Evidence suggests that the managers in our sample share to receive feedback on their ideas and to attract additional arbitrageur capital to the securities they recommend in order to correct mispricings.