VC’s are bizarrely private about the returns of their funds, eschewing the calls of other asset managers for transparency–so hoi polloi (you and me) need to rely on whatever sources we can have.
Fortunately, our friends at CalPERS do believe in disclosure and twice a year publish the performance of funds in their alternative asset class. This includes the California Emerging Ventures Program. This is one of the few public sources of the performance of a wide basket of VC funds and as of August 6th, it’s data was updated to reflect performance to 31/12/2008.
I think it’s valuable for entrepreneurs and angels to get a flavour for how VC performs as an asset class over time; and the variety of performance tracks a given fund can take. These will reflect on the GPs you are dealing with and might explain their behaviour during a negotiation or board activity.
What does the CEV programme tell us?
Before we get into that, let’s be clear that the CEV programme invests in a lot of funds, in tranches of $5m and above. So while this isn’t an accurate, representative survey of the industry, it certainly gives wide, deep coverage of VC and PE. And CalPERS is a large pension fund which can probably bully its way into any PE fund it wants.
VC funds (like private equity funds) face a J-curve, that is the first few years after a fund starts investing, it’s returns tend to be negative. For this reason, VC funds of a 2005 vintage or later may plausibly argue they are in the dip of the J-curve. Firms of a 2004 vintage or earlier, should be clambering out of the J-curve, enjoying the bountiful and rapid M&A market of 2006/7.
Timing also matters. Funds with a 1999 vintage faced horribly expensive seed rounds. Those of a 2001 – 3 vintage were able to enjoy shell-shocked, post-crash entrepreneurs settling for lower valuations. Valuations steadily increased until late 2007 (so once again late vintage funds may have disproportionately lower returns as reported by CEV.)
So here are some interesting extracts, with an unapologetic European-VC bias:
Atlas VI, a 2001 vintage, has an IRR of -5.7%; and has paid out $1.1m of $6.2m committed. This means it is reporting on about $4m book value from companies in that fund. The fund may significantly outperform if the remaining companies in that fund exit well. (This is the only Atlas fund CEV seems to have invested in, although I believe they raised one more since).
Battery Ventures (firm behind BazaarVoice and BlueKai) is a favourite of the CEV. Battery VI (2000) has an IRR of 2.7%, Battery VII (2005) IRR of 5.7%; Battery VIII (20008) an IRR of of -11.1% (deep in the J).
Draper Fisher VII (2000 vintage) received $25m from CEV, no doubt on the strength of the franchise. This fund has an IRR of -2.8%
CEV also like Highland Capital Partners, having invested four funds. The 2000 vintage has been a mediocre, IRR of -2.9%; the 2001 vintage (in which CEV ponied up $35m) has a commendable IRR of 8% (during that period the S&P dropped 20-30%); the 2006 fund (-16.1%) and 2007 fund (-33.6%) look deep in the J.
Two Index funds have received investment from CEV. Index II (a 2001 vintage) is a stand out success, with an IRR of 30.7%, and has returned hard-working Californian pensioners twice their money. Index III, a 2005 vintage, is still being invested but my hunch would be that with a non-meaningful IRR after 3 yrs, this fund won’t be a knock-out like its older sister.
NEA 9 and 10 (1999 and 2000) were both flunks, the 99 vintage losing 60 cents on the dollar; and the 2000 vintage returning close to what was put in.
The full data is available for the four different tranches of CEV: 1999, 2000, 2001 & 2006.
It is also helpful to look at the overall range of returns that Calper’s has seen through CEV. The fund has placed around $2bn into various tech, biotech and clean VC and mid-market PE over several years. They have appointed a highly-regarded advisory firm to run the program. And CEV itself has been the subject of an HBS case study.
Despite this, the overall returns are not great. (Will someone pass this stuff through a spreadsheet and share it back out?) You can understand why CalPERS treasurer has put the asset class and fee structure under review.(Remember, any GP whose fund returned less than 7.5% or so over eight years, would have been paid more for management fees than from performance.)
At a time when venture capital has an opportunity to ask how best to configure itself (see the likes of Fred Wilson|Destin) more transparency would help. So it is great that CalPERS provides this, because smart people from across the ecosystem can take a look at it and come up some radical, practical, wonderful or daft suggestions on how to improve it.
As Linus’ Law states: “With many eyes, all bugs are shallow” 
P.S. If you are an LP or GP in any of these funds, or an investee company, then feel free to send me an anonymous mail via hushmail about your investment terms and final exit so that we can piece together a clearer picture.
Popularity: 78% [?]