Decreasing Returns or Reversion to the Mean? The Case of Private Equity Fund Growth
This article analyzes the interaction between the growth and the performance of the two most prominent groups of private equity (PE) funds: buyout funds and venture capital (VC) funds.
These funds are usually structured as closed-end, limited-life vehicles. Therefore, in order to keep earning fees, PE firms periodically attempt to raise a new fund, typically every 3 to 6 years. The success of a PE firm’s fundraising efforts critically depends on the performance of the previous funds it managed. In particular, the best-performing PE firms are able to raise follow-on funds that are often 2 or 3 times larger than their preceding funds. These dynamics motivate this study’s main question: what is the impact of fund growth on performance?
The answer to this question is far from obvious. On the one hand, there are reasons to expect that PE funds might face decreasing returns to scale. When PE firms raise a larger follow-on fund, they have to make more and/or larger investments than they did in their previous funds. Yet, general partners’ time and energy are limited, and there are only so many potential portfolio firms for which returns are high enough to offset the funds’ fees and illiquidity.
On the other hand, PE firms might be able to keep decreasing returns at bay. First, larger funds earn substantially higher fees, which they can use to expand their operations, hire new associates to help source deals, and attract additional experienced general partners. Second, anecdotal evidence and academic studies suggests that a PE firm’s reputation (which is arguably correlated with fund growth and fund size) might help to generate higher returns. Specifically, studies have shown that more reputable buyout funds have better and cheaper access to credit. Moreover, young firms and startups might prefer to be financed by larger and more experienced VC firms because they are able to offer more resources, better networks, and higher certification value. Indeed, empirical evidence shows that more reputable VCs are more likely to be able to access hot deals and obtain favorable contractual terms and pricing.
Another important issue is whether the aggregate size of the PE fund industry has an effect on average fund returns – for example through competition for deals. In this research article, I focus exclusively on fund-level effects of fund growth and abstract away from aggregate effects. In practice, this is done by comparing returns across funds that operated in the same market environment; for instance, I compare returns across US-focuses VC funds that started investing in 2005.
This article uses a commercially-available dataset of 2,500 buyout funds and VC funds to study the effect of fund growth on performance. A first glance at the data seems to confirm previous findings: there is a strong negative correlation between changes in fund size and changes in returns within the funds managed by a given PE firm. In other words, when a PE firm raises a larger follow-on fund, that fund tends to underperform the preceding ones. This kind of negative correlation is usually interpreted as evidence of decreasing returns to scale by both academics and practitioners. For example, institutional investors often report being disappointed by the returns of the larger follow-on funds raised by PE firms that were very successful in the past.
This study argues that this interpretation of the data is unwarranted because it fails to properly account for reversion to the mean in fund performance. Reversion to the mean is present is any setting where observed performance depends both on a fixed component (which is sometimes referred to as skill or persistence) and on a transitory component (which is sometimes referred to as idiosyncratic shocks, noise, or luck). This phenomenon has been observed, for instance, in the performance of corporate managers, fighter pilots, athletes and mutual fund managers.
How does reversion to the mean help us to understand the relation between fund growth and performance in private equity? We know that, on average, (i) funds with high returns tend to be followed by larger funds, and (ii) funds with high returns tend to be followed by funds with relatively lower returns. Therefore, we should expect that larger follow-on funds underperform their preceding funds, on average, even if there were no decreasing returns to scale.
Based on the argument presented above, I analyze the data to understand what really drives the negative relation between fund growth and performance. The main finding is that, once I account for reversion to the mean, the direct effect of fund growth on performance becomes 80% smaller and statistically insignificant. Therefore, the main reason why larger follow-on funds tend to have lower returns than their preceding funds is not because fund growth has hindered performance, but rather because the preceding funds’ returns where unusually high. In other words, PE firms that have been able to raise larger follow-on funds were in part lucky in the past.
There is, however, a silver lining: the negative correlation between fund growth and returns is not caused by PE firms’ decision to raise larger follow-on funds. This means that, historically, PE fund managers have been able to scale up quite effectively and have avoided growing to the point where decreasing returns are strong enough to harm performance significantly. Consistent with this interpretation, we know that successful buyout and especially VC firms sometimes restrict the size of the funds they raise, which often go oversubscribed.
So how are PE firms able to scale up so well? I studied an additional dataset containing tens of thousands of fund-level investments to find out exactly how fund portfolios change as a function of fund size (1). For buyout firms, larger follow-on funds make only marginally more investments that their smaller preceding funds. Rather, they simply take over larger firms. VC firms behave somewhat differently: increases in the dollar amounts and in the number of portfolio investments account for approximately 60% and 40% of a given increase in fund size, respectively. Therefore, buyout funds appear to be easier to scale than VC funds, yet both seem to be able to do so without hurting returns significantly.
1: This set of results will be available in the next draft of the paper.Download the complete