Venture Power-Law Returns and Portfolio Construction
Venture capital returns don't follow a bell curve where most investments cluster near an average outcome — they follow a power law, where a small handful of investments return the entire fund and most investments return close to nothing, which changes almost everything about how a sensible venture portfolio is built.
Prerequisites: Venture Rounds, Preferred Stock and Liquidation Preferences
In most investing contexts, a portfolio of many holdings produces a return distribution that clusters somewhere near the average — a handful of standout winners, a handful of losers, most positions somewhere in between. Venture capital returns don't behave that way at all. Most startups a venture fund backs fail outright and return close to zero, a smaller number return a modest multiple of the money invested, and the fund's entire performance is typically driven by one or two investments that return 50, 100, or even 1,000 times the capital put into them. This shape — a small number of extreme outcomes dominating the total, rather than results clustering near an average — is called a power-law distribution, and it is fundamentally different from the roughly bell-curve-shaped outcomes typical of, say, a diversified public equity portfolio.
Why this changes how a venture portfolio gets built
If returns clustered near an average, a venture investor could reasonably diversify broadly and expect a fairly predictable blended return. Under a genuine power law, broad diversification alone doesn't help nearly as much — what matters far more is not missing the handful of enormous winners, because a fund that passes on the one company destined to become a generational success cannot make up for it with a larger number of merely-good investments. This is why venture investors put enormous weight on reserving follow-on capital: once a portfolio company shows early signs of being a breakout winner, doubling down in later rounds to maintain or increase ownership captures far more value than spreading that same capital across new, unproven companies instead.
What this means in practice
A venture fund manager's job looks less like building a smooth, diversified portfolio and more like running enough independent bets to have a realistic shot at holding one of the rare extreme winners, while reserving enough capital to concentrate further into that winner once it starts to reveal itself — a very different discipline from traditional portfolio-risk management, which typically works to reduce the influence of any single position's outcome rather than deliberately seeking it out.
Because venture returns are dominated by a tiny fraction of extreme winners rather than clustering around an average, a venture portfolio's success depends far more on capturing and doubling down on rare outliers than on broad diversification or avoiding individual losers — most losses are expected and largely irrelevant to the fund's overall return.
Don't judge an individual venture investment, or even a whole vintage year, by its early loss rate. A portfolio with a majority of failed investments can still be an excellent fund if it holds the one or two outsized winners that power-law returns predict will carry the entire result — the failure rate alone tells you almost nothing about fund performance.
Related concepts
Practice in interviews
Further reading
- Peter Thiel, Zero to One (ch. 9, Follow the Money)