Venture Capital Returns: Power Law and the J-Curve

Venture capital is the most misunderstood asset class in institutional investing. The headlines celebrate the outliers — the fund that turned a seed check into a billion-dollar position — while obscuring the base rate: the median venture capital fund barely matches public equity returns after fees. Understanding why requires grappling with two defining features of VC: the power law and the J-curve.

The Power Law

Venture capital returns follow a power law distribution, not a normal distribution. This means a tiny number of investments generate the vast majority of returns. Data from Horsley Bridge Partners — one of the few institutional LPs to publish disaggregated data — shows that the top 5% of investments by a given fund typically account for 60-80% of total fund returns.

This has profound implications. In public equities, diversification works because individual stock returns are roughly normally distributed. In VC, diversification is necessary but insufficient — you must be in the right deals, not just enough deals. A fund that makes 30 investments but misses the power law winners will underperform despite apparent diversification.

The distribution is stark: roughly 30% of VC funds lose money for their investors (return less than 1.0x net of fees). Another 22% barely return capital. Only the top quartile — funds returning above 2x — meaningfully outperform public markets after accounting for illiquidity and fees. The top decile, returning 3x or more, is where the asset class reputation is made.

The J-Curve

The J-curve describes the pattern of returns over the life of a VC fund. In the early years (years 1-4), the fund draws capital from LPs, pays management fees, and marks down its losers — producing a negative net asset value. It is only in years 5-10, as the winners mature and are realized through IPOs or acquisitions, that the fund's net value rises above the invested capital.

The J-curve creates real challenges for institutional investors. A pension fund that allocates to VC must tolerate years of negative reported returns before any gains materialize. This is why VC allocation requires long time horizons and committed capital — you cannot invest in venture capital with money you might need in the next five years.

DPI vs TVPI vs IRR

VC performance is reported through three metrics, each telling a different story:

DPI (Distributions to Paid-In) measures actual cash returned to investors relative to capital contributed. A DPI of 1.5x means the fund has distributed $1.50 for every $1.00 invested. DPI is the only metric that measures realized returns — it cannot be gamed by optimistic markups.

TVPI (Total Value to Paid-In) adds the unrealized value of remaining portfolio companies to the distributions. TVPI is always higher than DPI because it includes the estimated value of companies that have not yet been sold. This metric is useful for tracking a fund's trajectory but can be misleading — unrealized value is an estimate, not cash.

IRR (Internal Rate of Return) measures the time-weighted annualized return. IRR accounts for the timing of capital calls and distributions. A fund can show a high IRR by returning capital quickly on a small number of investments, even if the total multiple is modest.

The experienced LP focuses on DPI. Young funds will always show attractive TVPIs because their unrealized portfolios have not yet been tested by reality. The question is not what the portfolio is marked at, but how much cash has come back.

Vintage Year Matters

The year in which a VC fund begins investing — its vintage year — has a larger impact on returns than manager selection. Funds that deploy capital during periods of low valuations (2001-2003, 2009-2010) have structurally higher returns because they buy ownership stakes at lower prices. Funds that deploy during valuation peaks (1999-2000, 2021) face the opposite headwind.

This creates a paradox: the best vintage years are the ones that feel worst at the time of commitment. Allocating to a VC fund in a recession requires conviction that the long-term opportunity outweighs the short-term uncertainty — which is precisely the bet that produces the best returns.

Sizing the Allocation

For institutional investors, the consensus VC allocation ranges from 5% to 15% of total assets. University endowments — the most sophisticated VC allocators — tend to sit at the higher end: Yale's endowment has maintained a 20%+ allocation to venture capital for decades. Pension funds, with shorter time horizons and greater liquidity needs, typically allocate 5-10%.

For individual investors, direct VC access is limited, but the principles apply to anyone considering illiquid alternatives. The question is always the same: can you commit capital for 10+ years, tolerate the J-curve, accept that the median outcome is mediocre, and size the allocation so that a total loss does not impair your overall portfolio?

The honest answer for most investors is that venture capital is better admired than owned. The asset class generates extraordinary returns for the best managers, but the best managers are the ones you cannot access. The rest of the industry provides illiquidity, fees, and median returns that roughly match — and sometimes lag — a simple index fund.