The Persistent Effect of Early Success: Evidence from Venture Capital

Unlike other assets classes, private equity -- especially venture capital -- has been found to exhibit performance persistence. Money managers that perform well in one period have a higher likelihood of being above-average in their performance in the future. To understand better what might account for this persistence, we examined how the performance of VC firms’ investments –- in terms of having successful exits, either through IPO's or trade sales –- depended on their past success. Whereas prior studies have analyzed this persistence at the level of the fund, our analysis shifted the unit of analysis to the individual investment.

Consistent with prior studies, we found that prior performance predicted future performance. Each additional IPO among a VC firm’s first ten investments, for example, predicted as much as an 8.5% higher IPO rate on its subsequent investments. Although the performance of VC firms converged with increasing numbers of investments, we found that initial success predicted future success for as many as 60 subsequent investments. Initial success, moreover, had far more predictive power than more proximate success, success in the ten prior investments.

Both initial and future success depended in large part on being in the right places at the right times, investing in ``hot'' industries and regions. But VC firms did not appear to differ in their abilities to select these attractive segments. Those that invested in a hot segment in one period had no greater likelihood than chance that they would invest in the next hot area.

VC firms also did not appear to differ in their abilities to select or nurture specific companies. At the firm level, the companies in which successful VC firms invested had no better odds of a successful exit than those in the same industries and regions in which those VC firms had not invested. VC firms enjoying early success did, however, shift their investments to later stages and to syndicated investments. Initial success also allowed these firms to move into more central positions in the co-investment network.

The picture that emerges then is one where initial success gives the firms enjoying it preferential access to deal flow. Both entrepreneurs and other VC firms want to partner with them. VC firms therefore get to see more deals, particularly in more mature startups, when it becomes easier to predict which companies might have successful outcomes. Entrepreneurs and fellow investors will often even accept less favorable deal terms to partner with these successful venture capitalists. Even if venture capitalists do not differ in their abilities to identify more promising ventures, this access channel can still perpetuate differences in initial success over an extended period of time.

Even though these differences do not emerge from heterogeneity in the abilities of venture capitalists, investors in venture capital, limited partners, can potentially still invest in them to earn excess returns. Whether they can do so, however, depends in large part on whether investors have enough information about the performance of previous funds at the time that they must decide whether to invest in future ones. Phalippou (2010), for example, notes that a large share of the correlation in returns across funds stems from investments made within only a few years of one another, when the outcomes of the earlier ones would not necessarily have yet been realized. Our results, nevertheless, suggest that at least a small portion of the performance persistence associated with early success lasts long enough for investors to react to it.