The Princeton PhD Program Fails Hacking Venture Capital


Venture capitalists get plenty of good press but it’s hard to figure out how much money they’re making for their investors. That may be because you’d be better off putting your money in an S&P 500 index fund than giving VCs 2% of your investment plus 20% of the profits.

To be fair, the averages tend to gloss over the outliers who outperform the industry. This month I spoke with a Princeton Ph.D. program dropout who says he’s figured out an effective algorithm for beating the odds. Read on to learn how he’s doing it.

Before getting into that, let’s look at the performance of VC funds. What I’ve learned is that VC fund performance used to appear regularly on the website of the National Venture Capital Association (NVCA). It usually appeared there in the form of 10-year internal rate of return calculations for funds — based on the year they started investing. For example, in the early 2000s, I would visit the NVCA site and find that the average IRR of a fund started in 1989 was a whopping 86%.

But in the last few years, those performance figures have been difficult to find.

Fortunately, the firm that calculates these IRRs is still around and I recently found a site posting the returns for VC funds as of September 2016. In a nutshell, their performance after fees has been mediocre. For the 10 years ending September 2016, the pooled return for multi-stage funds was 9.43%. If those limited partners had instead invested in an S&P 500 index fund, they would have paid a fee of 0.05% and their money would have appreciated at a 5% annual rate — not taking into account dividends. It is hard to compare the two because of the much bigger fees that VCs charge. But limited partners have really suffered the most if you look at the performance of VC funds during the five years ending September 2016. For that time period, the VC rate of return was 13.7% after fees — compared to a 13.8% compound average growth in the S&P 500 for the five years from September 2011.

In short, since the dot-com boom, venture capital has not generated enough return compared to a stock market index to warrant the much higher fees and risk.

But Paul Martino, founder of Bullpen Capital — which provides post-seed stage financing, after an institutional seed fund but before a series A — thinks he can do better. He sounds to me like a math and computer science prodigy which inevitably gets me interested. As he said in a November 7 interview, “I grew up in Lansdale, Penn. and got my undergraduate degree in computer science and mathematics from Lehigh in three years, earned an MS in computer science at Princeton and dropped out of its Ph.D. program in high-performance computing. I hated being in school and entered graduate school at 19. It was the default option. I wish I had been born 20 years later. I wanted to be an entrepreneur since I was five years old.”