The following is a nice chart (courtesy of Focus Ventures) showing the large gaps in performance across quartiles in venture capital. Anyone surprised that most smart investors want to invest in the first quartile of venture funds, or none at all? After all, risk-adjusted performance of the VC second-quartile is less than that of the public markets.

## Latest Stories

# Skewness in Venture Capital Returns

By July 22, 2005 · ·

·
Skewness in ReturnsPaul Kedrosky has an interesting graphic on weblog post Skewness in Venture Capital Returns. It reminds me of when I used to look at junior oil and gas companies in Calgary. Arc Financial did and probably still does release a…

Hi Paul,

A few comments on the whole top quartile issue intended to stir the pot a bit:

1. It should not be surprising that the top quartile of VC funds produces superior returns. In fact, you don’t even need any actual data on VC returns to determine this; just simple statistics and the basic constructs of Venture Capital. From a mathematics standpoint, what you have is a one-tailed distribution with a rather long tail representing the extreme variance that can occur in a business where you can never lose more than 100% on a single deal but you can sometimes make a 10,000%+ return. When you divide such a tail into quartiles, the top quartile is guaranteed to “perform” better than the other three in any kind of reasonable distribution. Thus, there’s no way that the 2nd and 3rd quartiles will ever catch up unless there is a change in the independence of the individual outcomes or a substantial change in the shape of the distribution, both of which are possible but not likely.

2. As economists have determined other infamous quartile studies, perhaps the most prominent of which being studies of income distribution in the US, the question is not really “Can the other quartiles catch the top quartile?”, because that’s a logical impossibility, but “Is there sufficient turnover in the top quartile to provide for the reasonable prospect of upward mobility?”, i.e. how consistent is the cohort within a given quartile over time. In the VC space, you hear a lot about certain firms that are “consistent” top quartile performers, but I have never actually seen a study that reliably quantifies this consistency. My guess is that while there are a number of consistent performers there are probably quite a few firms that wander between quartiles from fund-to-fund. If this is the case, then for LPs life is a bit trickier than simply picking a top quartile fund because there’s always a decent chance that the firm that just turned in a second or even a third quartile fund, will turn in a 1st quartile performance next time around and that their top quartile firm will turn in a real clunker next time.

Perhaps the most interesting studies to see then would be how consistent the cohorts are within each quartile are over a certain time period and what, if any, factors are most predictive of top quartile performance over time. Such studies may prove to be better investment tools for LPs than their rearview mirrors.

Monsieur Burnham made a great point that there may be churn within each of the quartiles.

I believe that “A Random Walk Down Wall Street” refers to various studies that show that past performance is not an indicator of future performance for mutual funds.

Significant turnover in the quartiles could be the reason LPs are giving money to today’s poor performers. The zeros of today could be the heros of tommorrow. But are the LPs venture allocations beating the S&P 500?

You could move this study of quartiles up a level and segment LP performance (just their venture allocation) by quartile. I wonder what percentile of LPs are beating the S&P 500?

And then, what is the turnover in the quartiles for LPs?

I am not a VC but I would like to know something.

Whilst your meetings with LP’s are sure to include a healthy discussion of venture returns and distribution there of…

Isn’t the primary motivation for LPs plain old diversification? In other words, to better practice modern porfolio theory, don’t the LPs *have to* sign up for your funds as a measure to diversify risk rather than get all hyped up about your returns?

Isn’t it just the matter of having exposure to *all* assets which is what drives the investment decision, rather than any kind of “let’s pick the winners” kind of funny business?