The Power Law in Venture Capital, Explained with Real Numbers

The power law in venture capital means returns are not spread evenly across a portfolio: a handful of investments produce most of the gains, and most investments lose money. Correlation Ventures' study of about 21,000 US financings from 2004 to 2013 found that roughly 65% returned less than the money invested, while about 4% returned ten times or more. A venture fund's result therefore depends on whether it owned a meaningful stake in one or two outliers.

The clearest illustration is a single check. In 2005 Accel invested a reported $12.7 million in Facebook's Series A. At the 2012 IPO, that stake was reportedly worth more than $9 billion. No amount of careful work on the rest of the portfolio could have mattered as much as that one decision. That is the power law in practice, and it shapes nearly everything venture investors do.

What is the power law in venture capital?

A power law is a distribution in which a few very large values dominate the total. In a normal distribution, like human height, the average is close to the typical case. In a power law, like city sizes or book sales, the average is pulled far above the typical case by a few giants.

Venture returns follow that second shape. The book's research companion fitted a distribution to the Correlation Ventures figures with a Pareto tail. It has a mean deal multiple of about 2.8x and a median of about 0.5x. In other words, the typical deal loses half its money, yet the average deal nearly triples it. Both statements are true at once because of the tail.

The evidence comes from several datasets and decades:

These are reported estimates from different samples, but the pattern recurs. For the full data set, see the performance and power law study.

Why do a few deals generate most venture returns?

The cause is the asymmetry of the payoff. The most an investor can lose on a deal is 1x the check. The most it can gain has no ceiling. A company that grows from nothing into a global platform can return a thousand times the seed check. When losses are capped and gains are not, the right tail decides the outcome.

History is full of single positions that made entire funds:

Peter Thiel, who later co-founded Founders Fund, popularized the "power law" framing in his 2014 book Zero to One. The data behind it were there long before.

What does the power law mean for a venture fund?

The book's worked example makes the arithmetic concrete. A $200 million fund pays 2% a year in fees for ten years, leaving $160 million to invest in 30 companies at about $5.33 million each. Outcomes follow a power-law shape:

Gross proceeds are $538.7 million, and after 20% carry LPs receive a net multiple of about 2.35x. Half the companies went to zero, and the fund is still good. Now remove the single 50x company. Net TVPI falls to about 1.29x, well below the roughly 2.6x a decade of public equities compounding at about 10% a year would have returned. One company is the difference between a strong fund and a poor one.

That leads to the most useful formula in venture:

Exit value needed to return the fund = Fund size ÷ Ownership at exit

At 8% ownership, the example fund's best company must sell for $2.5 billion to return the fund by itself. A $1 billion fund owning 10% needs a $10 billion exit. Fund size is strategy, because it sets how large the outliers must be. Chapter 4 of The Outlier Equation builds portfolio construction from this equation.

Does diversification protect a venture portfolio?

Partly. The research companion simulated 20,000 funds at each portfolio size, using outcomes calibrated to the Correlation data:

The expected multiple is about the same, roughly 2.7x, at every size. More companies raise the median and almost remove the risk of losing money, but they barely raise the chance of a great fund. A great fund needs a large position in an outlier, not just exposure to one. That is the real trade-off: breadth protects against failure; ownership decides whether success matters.

Ownership also shrinks. Ownership after several rounds equals starting ownership times the product of one minus each round's dilution. A Series A investor's 20% can fall to about 13% after two more rounds, and a seed fund's 10% often becomes 4–6% by exit. Funds hold reserves to defend their stakes in winners, and follow-ons into the top few companies often turn a good fund into an excellent one. Sequoia's WhatsApp position is the textbook case of concentrated follow-on into a winner.

Why hit rate is the wrong way to judge a VC

If returns are a power law, the share of deals that lose money tells you little about skill. The book puts it bluntly: a 64% loss rate with one 20x outlier beats a 30% loss rate without one. The research companion's worked fund had 16 of 25 companies return less than 1x, with 49% of invested capital in money-losing deals, and one company supplying 55% of gross value. That is what a good venture fund looks like.

Buyout investors commonly see capital loss ratios of about 10–20%. An LP who applies buyout-style loss screens to venture managers will systematically select against the managers who take enough risk to catch outliers. Experienced LPs ask a different question: show me your 10x outcomes, what you owned at exit and how you got into the deal.

The power law also changes which error matters. The cost of investing in a company that fails is the check. The cost of passing on the company that becomes Google is many times the fund. Sequoia's famous misses are instructive: it never invested in Facebook, and it reportedly sold its Apple stake in 1979, before the 1980 IPO. Errors of omission are invisible on a fund's books but often larger than every write-off combined.

How the power law shapes pricing

Investors price deals so that the success case returns the fund. In the book's example, a Series A investor expects an $800 million success case, keeps 60% of its stake through later rounds and wants 10x. That caps the post-money at $48 million. Weighted for realistic odds, with a 50% chance of failure and a 5% chance of a $2 billion outlier, the expected multiple is only about 2.2x, and the 5% outlier scenario contributes about 56% of the expected value.

The implication is uncomfortable for founders of solid but limited businesses. A company that can plausibly sell for $200 million, however safe, has little value to a fund that needs $2 billion outcomes. That is not a judgment on the company. It is the math of the fund. The venture math and cap tables study works through these calculations.

Does the power law hold at every stage?

It is steepest at the earliest stages and flattens as companies mature. At seed, information is scarce, checks are small and most companies fail. At growth stage, most companies do not fail outright; typical upside on good outcomes is 2–5x rather than 10–100x, and losses come mostly from overpaying. That is why growth investors are judged more on price discipline than on picking.

The seed stage also shows how the power law interacts with the cycle. Carta's data show about 12% of the Q1 2022 seed cohort reaching Series A within two years, against 30% or more for 2018–2020 cohorts. When graduation halves, so does the number of companies that can become outliers. The seed and angel study tracks these rates.

The power law and the four factors

The Outlier Equation frames a venture return as four factors that multiply: whether a company can become an outlier, the price and ownership at which the investor holds it, the governance that protects value in the middle, and the realization that turns it into cash. The power law is mostly about the first two. A fund must find companies capable of extreme outcomes, and it must own enough of them, at a reasonable price, for those outcomes to matter. A fund that owns 0.5% of a decacorn may have backed the right company and still earned almost nothing on it. The factors multiply: picking the outlier is necessary, but owning too little of it at exit makes the result small anyway.

For how this connects to fees, carry and the business of venture as a whole, see our guide to how venture capital works and our explainer on how VC funds make money.

The takeaway

The power law means that a venture fund's result is set by its best one or two investments, and that most deals losing money is normal rather than a sign of failure. Diversification protects against disaster but does not create great funds; meaningful ownership in outliers does. Fund size, entry price, reserves and the willingness to back non-consensus ideas all follow from that fact. Chapter 1 of The Outlier Equation sets out the logic, and Chapter 23 shows how to read a track record built on it.