VC returns trump all
Truths, Myths and the Roadmap Part Two
If I heard it once in Calgary last week (see prior post “The juices are flowing at annual CVCA conference” May 28-09), I heard it 1,000 times: “venture capital returns have been poor”. Therefore, we need to cut our exposure to zero; and for those who might continue to fund the asset class, there is a desire to shrink the Canadian venture industry to just 8 or 10 firms nationally.
I’ve just revieved the most recent U.S. return numbers, and it is true. VC returns have been poor this decade. Except, not surprisingly, on a relative basis. On that metric, VC returns beat both Buyout and the public markets on a 1, 3, 10 and 20 year basis. Nevertheless, the vast majority of assets invested by pension fund managers go into buyout and public equities. Curious.
Here are the most recent stats (1, 3, 5, 10, 20 year horizon as at 12/08):
NASDAQ: -38.1%, -10.3%, -4.6%, -3.2%, +7.3%
S&P 500: -36.1%, -10.0%, -4%, -3%, +6.1%
All Buyout: -23.9%, +2.2%, +8.4%, +5.9%, +9.8%
All Venture: -20.9%, +4.2%, +6.4%, +15.5%, +17%
Who knew? Except for the leverage-loan blip in the 5 year data, VC returns beat buyout in every category. These figures are a bit skewed downward by the Mega Buyout class, but in the outer years it matters not a whit which pool you were fishing in.
One reader (Mark Z.) made the case yesterday that only 25% of VC funds make decent money, and the balance should be shuttered. According to a 2008 study by Harvard Professor Josh Lerner and the U.S. National Bureau of Economic Research, that is just another one of those myths that need to be put to rest. At least 54% of U.S. VC funds had a positive return for their investors (from inception, as at 3/08) according to his work.
It was certainly the case that the top 3% of VC funds saw returns in excess of 100%, but the 2008 presentation by former BDC staffer Brian Elder that Mark Z. relied upon doesn’t do the topic justice given the new return data that exists.
Reader Andy wisely pointed out yesterday that many pension fund types may actually seek liquidity over returns, which is why they put capital into the public markets over private strategies. Something for those of us on pension boards to reflect upon.
The other old saw is that “the risk didn’t warrant the returns”. It is easy to use that line from a lecturn, but what does it actually mean? That the risk of losing money in one investment exceeds the risk in another? Better update your financial risk algorithm now that AIG, GM and Citigroup have gone in the tank. The risk/return calculation of an individual manager can be easily determined using methods such as the Efficient Frontier, for example, and should be embraced.
The suggestion to back out the tech “bubble years” is fair, as long as one also finds a way to intelligently back out the impact of recent leverage bubble that private equity utilized to enhance their returns in the 2005-2007 period. (According to a presentation by Colony Capital’s CEO at last week’s CVCA AGM, up to one third of a private equity return is generated by the use of leverage.)
If a fund manager has weak, unjustifiable returns, don’t hire them. Don’t re-up. Put them out of business. But don’t kid yourself: VC returns have been great when compared to every other asset class.
Just ’cause the cool kids are hanging around a particular asset class, doesn’t mean it is healthy for your pension fund annuitants. Remember the advice that your Mom gave you about “the kids at the smoking door”?
MRM
Mark – Part 1 of your post talks about Canadian venture whereas Part 2 quotes U.S. data. The Canadian equivalent data, compared with the U.S. data that you use is at:
http://www.cvca.ca/files/Downloads/Session_1_Gilles_Duruflé.Calgary.Final.2.pdf
and shows that ‘all’ U.S. VCs (the data you quote above) perform about as well as top-quartile Canadian VCs. So an LP can go to the U.S., randomly pick some VCs and do about as well as trying to figure out which Canadian VCs are going to be top-quartile in Canada. I know where I’d put my money…
Discounting the tech bubble years does indeed make as much sense as backing out income trust private equity exits or leverage normalizing large buyouts, such adjustments which are all routine parts of a track record assessment. Negative outcomes are also potentially adjusted for where circumstances might warrant. But the whole point of track record analysis is to gauge likely future behaviour and relevant skill-set positioning for forward looking opportunities. That’s why a weak track record in itself does not disqualify one from consideration, and a strong one may be viewed is transient. Its all about what’s to come. The subtlety of private equity investing is what makes this whole asset class challenging, but intellectually interesting for those who appreciate how hard it is to balance all of the factors.
The venture industry has not made the case the skills, attitudes and business models largely developed during the bubble years, and unchanged to date, will work going forward. Don’t tell investors why they are wrong in how they make their assessment, just tell them what you want to do, and why. The returns history lesson suggests you are pining for a return to the glory days of favoured captial allocation to venture. The next wave of venture investing will look more like regular business investing, all investing actvities are in a way converging.
Mark,
There is no question that on average US venture capital has outperformed other asset classes, including the S&P equity index. Frankly, if it didn’t there wouldn’t be a VC industry today. So, I am in total agreement with you that US venture capital is a superior asset class to equities and buyout.
My previous comments, however, were focused on examining US VC returns by quartile rather than focusing on median returns. More specifically, the data indicates that 3/4 of US VC funds (ie., the bottom three quartiles) have underperformed the S&P over a 20 year period. This does not suggest that these funds didn’t generate positive returns (as the Lerner study indicates) but rather that they did not generate the minimum risk-adjusted returns to support their existence. Put simply, the LPs of these bottom three quartile funds would have been better off putting their money in an S&P index fund.
All that being said, let’s focus on Canadian VC returns for a moment. The data that Roger Chabra of Growthworks sent around yesterday (http://twitter.com/rogerchabra/status/2008183804) indicates that Canadian VC returns have generated negative returns over a 1, 3, 5 and 10 year horizon. Below is a comparison of Canadian and US VC median returns by time horizon.
US Venture: -20.9%, 4.2%, 6.4%, 15.5%
Cdn Venture: -7.4%, -1.8%, -1.8%, -2.8%
Perhaps more damning than these negative returns, is that the top quartile of Canadian VC funds have generated only a 3.9% return since inception. Investors would have been better off putting their money in GICs than in Canadian venture capital (even the top quartile of Canadian venture capital!) over these time horizons. What does this say about the viability of this asset class in Canada?
As you know, my beef with the VC industry, both in the US but particularly here in Canada, is that it lacks transparency and doesn’t provide its customers (ie., start-ups) with sufficient information for us to make informed buying decisions. That’s right, VCs need to start acting more like sellers and change the way they do business if they have any hope of attracting buyers (and investors) in this new environment.