How well are venture capital firms really performing?
You’d think they are doing fine, glancing at the official statistics released by the VC lobby group each quarter.
Venture firms have seen a five-year annual return of 2.7 percent, according to data compiled through March 31, 2007 (see table below). That’s not great, but the VC business is a long-term one. Firms investing for ten years are seeing 21 percent annual returns, on average, according to data published by the NVCA and Thomson Financial. Long term, VC performs better than popular market indexes like the Nasdaq or S&P. Those benchmarks show returns of only 7.1 and 6.5 percent, respectively, over the same ten years.
The VC data deceives somewhat, however, because it captures “average” returns. The better performing venture firms distort the data, pushing up the average. In reality, the vast majority of venture firms aren’t doing that well, something Dan Primack documents using data from funds raised post-bubble. The median firm is losing money. There’s also the continued problem of self-selection. A venture capital firm on the rocks is bound to stop providing data in this voluntary system, and so won’t show up in the data.
Also, look at the divergent fortunes of “seed stage VCs” and “early stage VCs.” Seed stage VC investors made a meager 0.8 percent annually over ten years. That compares to 40 percent for early stage VCs. However, when we asked representatives at both the NVCA and Thomson about this strange discrepancy, they dismissed the 0.8 number, saying it was based on a small sample size and shouldn’t be trusted. They don’t usually publish the “seed stage VC” number, but “it found its way to the table this quarter.”
Speaking of which, we’re tracking two more data related stories next week. One will look at the Thomson/NVCA data about how much money VCs are investing in companies, to be released on Tuesday. This comes after they failed last week to meet their deadline on releasing the quarterly data. Thomson and NVCA blame a server failure. This second data story is about Glam, the controversial women’s network that says its the fastest growing on the Web. See audio file below for preview of these stories.
3 Comments
-
Seth said:
Matt, this is interesting but it seems like there’s another deception. I may be mistaken, but the numbers don’t seem to include a factor to account for the differences in liquidity between a public index and a venture fund. (And those differences can skew the math.)
Specifically, if you look at public company indexes and measure Return on Investment, it’s simple because there is a clear valuation for every company. There is an explicit market-determined value on different dates. If you invest $1000 today, next week, you can pick a day and find out exactly how much your investment has changed. Nasdaq is up or down, simple, liquid.
In a VC portfolio until there is a “liquidity event” it’s obviously different. That $1000 could be tied up for 5 years and then be worth $100k or be worthless.
So the question I have is, how do these studies value and account for money during the five years it’s tied up? Do they account for illiquid investments.
If a venture fund, for example, has only seen returns on 20 percent of its investments - and a study only uses that 20 percent for measurement of return - 80 percent of capital under management didn’t get factored in. That doesn’t make for a very accurate measure of performance.
If they’ve accounted for all the capital in the fund, different story (but it doesn’t look like they’ve done that).
-
Peter Cranstone said:
Matt,
You might want to take a look at this linkL http://hosting.mansellgroup.net/enablemail/ThomsonNewLetter/HostedWires/NewsLetters/aug3-07.htm
for more information on VC returns. It paints a slightly different (but maybe more realistic) picture than your post above.Peter
-
Joe said:
Seth - all VC and PE funds are required to value their portfolios at FMV, so the issue you raise is actually addressed to some large degree. The return data factors in these protfolio FMV’s.
