Risk loving VCs

A reader (whose reply email address sadly did not work) wrote in regarding a recent post about VCs. It's a good letter, so I'll reprint most of it

After reading your article "Fixing Venture Capital," I was surprised that you (and the various commentators, some of whom are VCs!) failed to mention the more fundamental reason why VCs prefer riskier investments. It's not about diversification of portfolios, although you are correct that a diverse portfolio incentivizes the VC to make risky investments. The more fundamental reason is that the VC has a direct financial interest to maximize risk, even setting aside portfolio diversity.

Overwhelmingly, and simplifying only slightly, VCs are compensated in two ways. VC firms receive a "Management Fee," a flat annual amount that is supposed to cover their salaries, rent, secretaries, etc. The Management Fee is usally 2% of committed capital (ie capital they can drawdown from investors). The Management Fee is nice, but its not where the action is. VCs also receive a "Carried Interest" in the profits of the fund, a minimum of 20% (top-tier funds can get up to 35%, even in this bad environment). This means that to the extent that the fund is net profitable, the VC will receive 35% of the profit. In other words, because the VC takes a share in the profits, but is not investing his own money, the VC shares in the upside potential but has a limited downside potential; this is a classic case of assymetrical risk-taking encourages the VC to act riskier than he would if he had a straight interest in returns, rather than profits.

Just to make this overwhelmingly clear, lets suppose the VC had an average size fund of $100 million with a 20% carry that had two potential investment targets (each will take the entire $100 million capital). Target 1 has a 50% change of being worth $90M, and 50% chance of being worth $120M, for an expected net value of $105M. Target 2 has a 90% chance of being worth $0M, and a 10% chance of being worth $1,000M, for an expected net value of $100M. Obviously, Target 1 has a larger overall expected net value than Target 2, but now look at the VC's incentives.

With Target 1, the VC can expect his Carried Interest to be worth $4M [50%*20%*($90M-$100M, i.e. 0)+50%*20%*($120M-$100M)]. With Target 2, the VC's Carried Interest is worth $18M [90%*20%*($0M-$100M, i.e. 0)+ 10%*20%*($1,000M-$100M)]. Thus, *even where the net expected value of the riskier investment is lower, the VC is incentivized to make the riskier investment.*

Why is this so? Well, the investors who invest in VC funds are looking for extremely risky asset classes. They don't want VCs to turn their money around and invest it in companies with less risk (after all they can always buy S&P 500s for themselves). What are the implications of all this? I'm in agreement with Naval, VC money is not for every type of business. And thats not necessarily a bad thing, after all entrepreneurs with slower, less riskier business plans can get secured bank financing (which of course has the opposite incentives of overvaluing stability ...).

Essentially, the argument is that the carry ("carried interest") rewards VCs on the upside but does not penalize them on the downside, and so motivates them to be risk loving, not risk neutral. He demonstrates this by showing why a VC would choose an investment with higher variance but lower expected return over one with lower variance but higher expected return. Note that a risk neutral investor would pick highest expected return and ignore variance.

Also note that this structure is very similar to a call option, and we know that volatility (risk) makes options more valuable. An options trader, then, likes risk for its own sake. But a VC investment, even with carry, is not the same as a call option. Clawback provisions mean that the VC has to return the fund investors' money first before they can keep any for themselves. Also, if you want to raise a fund in the future, you better do a good job with the one you have now. These factors help VCs go from risk loving to risk neutral, but I don't know by how much. I don't think that institutional investors want risk just for the hell of it, otherwise they would allocate some of their pension money to Vegas.

You can actually model VC funds by looking at their payoff structure and putting together a portfolio of options that match it, both for the institutional investors and the VCs (so gross returns and net returns). It turns out that the generous carries VCs give themselves makes little difference to what the institutional investors get back, but it also seems that the VCs who actually outperform keep all their outperformance in compensation.