Team Stability and Performance: Evidence from Private Equity

In the paper “Team Stability and Performance: Evidence from Private Equity”, we empirically study the effect of team stability on performance in a team production environment. The effect of team stability on organizations is ambiguous. Stability (or lower turnover) can induce individuals to invest in relationship specific (team specific) capital and allows members of the team to get a better knowledge of the team members’ abilities, which leads to an improved allocation of tasks within a team. However, some turnover might be optimal as firms need to get rid of the underperformers and teams need to adapt in a changing environment.

We focus on the PE industry which provides an ideal setting to study this question. The PE industry is highly human capital intensive. Furthermore, it is an industry where both private equity firms and their investors put a lot of emphasis on team stability: when investors select in which fund to invest their money, stability of the team is one of their most important criteria. In addition, Private Equity organizational structure has some distinct advantages (compared to other industries) that allow for meaningful comparisons: it is a simpler organization compared to one where individuals have multiple relationships with direct hierarchical structures and PE teams are doing a relatively similar task, which makes comparisons easier.

We hand-collect data on 138 fund managers (or PE companies) by reading through the due diligence of an investor in PE funds. We have information on the due diligence of the funds in which the investor ultimately decided not to invest, alleviating selection concerns. Since this specific investor was concerned about team stability, there is ample information about the individuals in the team, with an emphasis on departures from the team as well as additions. Moreover, we can observe which individuals were in charge of a specific deal, thus allowing us to measure individual performance.

We define turnover as the average number of people leaving the PE firm normalized by the size of the team. First, we find a positive and significant relation between turnover computed over a five-year period from the start of a given fund and IRR of the current fund, or average IRR of the current and the subsequent fund within the same fund manager. Our intuition is that these results capture an immediate effect of replacing bad performers in the current fund, which thereby allows “fixing” performance of the current fund. Since the investments have already been made in the first 5 years, improvements can mainly be achieved through successful restructuring and exit of the invested deals.

We next examine a more long-term effect of turnover on performance at the time of raising the subsequent fund. We define turnover the average number of people joining and leaving the PE firm normalized by the size of the team over a five-year period before the start of a given fund. These five years correspond to the time period when the fund manager starts planning the new fund, wherein managers talk to investors about their new investment proposals and prepare their teams to reflect the changing needs and skills required to better respond to shifting external conditions. We find a positive and significant effect of joiners’ and leavers’ turnover over five years prior to the start of the fund on performance of the next fund. We are unable to replicate the same findings when we focus instead only on leavers’ turnover, which gives statistically insignificant results. These results suggest that turnover before the start of the fund is not about firing bad performers, but rather captures teams adapting to future conditions by bringing in different skills and fresh ideas.

We provide further evidence in support of the two channels. As evidence that turnover allows PE companies to replace underperforming individuals, we show that deals attributed to individuals who subsequently left the PE firm are associated with lower performance as compared to other deals within the same manager, invested in the same region and year, or exited in the same year, while those attributed to people who recently joined are slightly overperforming. To test the team adaptability channel, we examine whether there is a change in skill composition of better performing PE firms following recessions. Here, we implicitly assume that PE firms need to successfully restructure their invested firms during recessions and therefore their demand for operational skills increases (moreover, as leverage is less available during recessions, new investments will have to be more focused on operational restructuring). We find that, one year following recessions, better performing PE firms, as proxied by their past funds’ performance, have a higher share of individuals with operational skills in their teams and experience higher turnover of team members with operational skills.

This evidence suggests that turnover is associated with an increase in future performance. However, the optimal level of turnover in equilibrium implies no relationship with performance. Thus, an implication of our results is that there may be frictions that result in lower team turnover than it may be optimal. We discuss two of these frictions. First, the presence of asymmetric information between the firm management team and the investors may make the firm less willing to change the team because of a signal extraction problem. If the external investors cannot decipher whether the change in team composition reflects on the team ability or the changing external conditions, this would make the management more reluctant to alter the team. This problem will be more severe when there are large information asymmetries between investors and PE firms, leading to sub-optimally stable teams. Second, scarcity of resources (i.e. high skill individuals) may make the firm less willing to fire underperformers if it has low chances to hire high-skill individuals. We provide evidence in support of both those frictions.