How Portfolio Size Affects Early-Stage Venture Returns

With the caveat that the bulk of returns are unrealized and may fall in value in the future, the aggregate returns from early-stage venture look appealing. Cambridge Associates data suggests that venture capital has outperformed the public markets over the past decade. Additionally, our own data agrees with a provocative conclusion from Invesco: venture capital and public market performance are generally uncorrelated. Competitive and uncorrelated performance suggests that early-stage venture is a materially different asset class which investors could benefit from.

But at an individual level we commonly hear a different story from small-scale seed investors. There are plenty of posts around the Internet about folks giving up on or disparaging angel investing based on their own bad experiences, and these posts align with many anecdotes I have heard from angel investors who have abandoned investing in startups. While there could be issues for these investors around access or selection, I do not think that these stories represent especially unlucky or unskilled cases. In fact, our data suggests that the median investor exposed to three or fewer startups on AngelList has a negative portfolio value.

So, how can early-stage venture look like an attractive asset class but be repellent to individual investors at the same time? My colleague Nigel Koh and I looked at the performance of more than 10,000 AngelList investor portfolios to find the answer: Investors who invest in more deals do better, both in average and typical return.

As an example, the below figure displays the median IRR for investors on AngelList grouped by the number of investments they are exposed to — so if, for instance, an investor had participated in two syndicate deals and a microfund that has made 11 investments they would be counted as having made 11+2=13 investments. The dots are the grouped medians, and the black line is a linear regression fit:

Performance generally increases in the number of investments made; the coefficient of the linear regression line here is 9 basis points per investment (one basis point is one one-hundredth of a percent), suggesting that the typical investor with a 100-investment portfolio outperforms the typical investor with a single investment by almost 9% a year.

One of the easiest ways that an investor could take a broad and systematic approach to early-stage venture capital is through investing in vehicles that broadly index into early-stage venture capital, such as the AngelList-managed Access Fund. In our paper we also look at the performance of investors who participated in the Access Fund against those who did not. We found that, across many years of investment activity, Access Fund investors typically and in expectation outperform those investors who did not participate in an Access Fund.

Our whitepaper “How Portfolio Size Affects Early-Stage Venture Returns” has all of our results and methodology.

‍