I recently tweeted two charts from a Preqin report on Venture Capital. Since that tweet got a lot of interest (relative to the amount of interest my tweets generally get) ranging from celebratory to cynic, I thought I would dissect it further.
The top chart (Fig. 1) shows the performance of first-time funds and non-first time funds compared to the venture industry as a whole, by vintage year (2006 through to 2013), measured by the median net IRR as of September 2016. The bottom chart (Fig. 2) plots the performance (again measured by net IRR) of first-time funds vs non-first time funds from all vintages from 2003 to 2013 relative to the standard deviation of that performance (standard deviation in statistics is a measure of the dispersion of a data set around its mean, and applied to the financial world is a common proxy for risk). I can infer the size of the bubble represents the size of the population for each bucket.
Since the measure of performance is net IRR, this is all to be read with a LP hat on.
The point I was trying to convey in the tweet is that while first-time funds do outperform non-first time funds on an absolute basis (over that period), the outperformance is (in part or entirely?) a function of first-time funds being riskier than non-first-time funds, as measured by the standard deviation of their returns. In crude terms and as a way of example, a first-time fund is more likely to return <1x than a non-first time fund, but equally more likely to return >5x, while the returns from a non-first time funds are less disperse and more concentrated around the mean of the distribution. That is, the distribution of returns from first-time funds are more “fat-tailed“.
The relationship between risk and expected return should ring bells to anyone who has taken portfolio management class at school and recalls what the efficient frontier is: investors only accept higher risk if it is compensated by higher returns, or in other words there is no ‘free lunch’ in an efficient capital market i.e. the only way to earn an extra unit of return is by taking an extra unit of risk, and therefore assets are priced accordingly by the market.
So why are first-time funds riskier?
- Size. To start with, first-time funds are likely to be smaller in size, just like a Seed round is smaller than a Series A round. Preqin data shows how first-time fund size targets have ranged between $100-170m for the past 10 years, compared to a range of $200-370m for non-first-time funds. Smaller funds tend to play more earlier stage where the attrition rate is higher, as is the dilution risk from not having enough funds to follow on in the winners, but also where the potential return from hitting the outliers is much higher (this is also referred to as convexity of returns, in that the downside is capped at 0x but the upside is, theoretically, limitless);
- Focus. As a first corollary to the smaller size, first-time funds tend to focus on a single region or on a single sector, thus not benefiting from a more diversified portfolio that would reduce the volatility of returns;
- Incentives. A second corollary to fund size is GP incentives: raising larger subsequent funds allows GPs to earn larger, and cumulative, fixed cash compensation regardless of fund performance. Management fees on first-time funds tend not to be substantial in absolute terms, particularly when netted off of GP commitments. So one could argue the attitude towards risk taking is higher in managers of first-time funds;
- Impatience. As sub-corollary to incentives, first-time funds main goal is to demonstrate enough proof points to allow them to raise a second fund; in doing so, they may be more tempted to accept quick early exits as validation, which might generate higher IRR (though lower multiple and less cash creation) compared to more patient non-first time fund managers;
- First-time funds can be managed by first-time managers who don’t benefit from previous investment experience or come with a completely different background from existing funds. That can either make them more susceptible to making mistakes, but equally more receptive to ground breaking, yet dismissed, ideas in the absence of preconceived notions or powered by unique perspectives;
- First-time funds can also be managed by experienced GPs who may have spun-off from more established funds; they may come with proprietary experience and networks, taking them away from non-first time funds, which coupled with a clean sheet may yield very different returns;
- Innovation. First-time funds are more likely to be exploring innovative strategies and funds models, which are yet unproved and could yield very different (read more dispersed) outcomes from those on non-first-time funds; such strategies can also be developed in response to markets, conditions and opportunities that non-first-time funds cannot adapt to (quickly enough), thus missing out on them;
- Brand. Lastly, a first-time fund, by its very nature, has no institutional heritage nor brand to leverage (though as mentioned above “spinoff GPs” in first-time funds do), which could potentially lead to a different quality of dealflow from non-first-time funds.
It would indeed be very interesting to have the same top chart, but showing returns adjusted for risk, to see if first-time funds do indeed generate true alpha (i.e. that portion of excess return that isn’t explained by extra risk) compared to the industry as a whole. It would equally be very interesting to split performance for first-time funds into first-time GPs vs first-time vehicles managed by spinoff GPs, and also first time fund performance by fund size or strategy.
While I don’t have that data at hand, and the preconceived notion in the industry is that success breeds even more success, I like to believe that first time funds have an increasingly good chance at capturing alpha in the industry. A recent Cambridge Associate report demonstrates how venture capital itself is undergoing disruption: returns are getting more democratised and are no longer exclusive property of the top 10 firms on Sand Hill road.
As innovation and company creation happen literally everywhere, the next unicorn can be hiding where non-first-time funds aren’t closing looking. There is also a softer argument made by an investor who commented to my tweet, that first time fund managers work harder to to prove themselves. There may be some truth in there too!
While entirely anecdotal, the delta between performance of first-time and non-first-time funds in Fig. 1 of the Preqin report appears to be increasing over the period under consideration: is that because first-time funds are getting riskier or could there be some alpha in there?
The good news is that LPs are taking notes. According to Preqin, half of the c. 4,200 investors tracked by the platform and confirmed as actively investing in venture capital have either committed to first-time funds or are open to the idea. Also, according to Shai Goldman‘s spreadsheet that tracks <$200m venture funds closed since 2011, first-time funds continue to attract strong LP interest, with average fund size having increased from $54m in ’13 to $81m so far in ’16, showing more conviction. And notably 2016 is already higher than 2015 in both number and amount raised for first-time funds (though likely to end up down from the 2014 peak by 30-40% in numbers and 20-25% in dollars).