The Power Law & Portfolio Math

What the power law is

A "power law" is a statistical distribution in which a few outcomes are enormously larger than all the others. Most things in life follow a different pattern — heights of people, IQ scores, daily commute times — what statisticians call a normal distribution, the bell curve. A normal distribution has a sensible average, and most observations cluster close to it.

A power law is the opposite. The average is misleading. The median is far below the average. A small number of observations dominate everything. Wealth follows a power law (a few people have most of it). City populations follow a power law (a few cities are huge, most are small). Earthquake magnitudes follow a power law.

So do venture capital returns.

The shape of VC returns

In a typical 25-investment VC fund, outcomes look roughly like this:

The numbers vary by fund and stage, but the rough pattern is consistent across credible studies. Roughly half of all VC-backed companies return less than the original investment. The fund's actual return comes from the right tail — the few outliers that compounded extraordinarily.

Peter Thiel articulated the resulting principle most cleanly in Zero to One: the best investment in a successful fund equals or outperforms the entire rest of the fund. One company carries everything else.

What a "fund returner" is

A "fund returner" is a single investment that, on its own, returns 100% of the fund's committed capital. On a $100M fund, a fund returner is one ~$1M investment that turned into $100M+ of proceeds — roughly a 100x outcome on that single check.

Top funds typically have 1-3 fund returners per fund. A few legendary outcomes:

• Sequoia's investment in WhatsApp — reportedly $60M in, $3B+ out at the Facebook acquisition.
• Benchmark's investment in eBay — $5M in, several billion out.
• Lowercase Capital's investment in Uber — small early checks, hundreds of millions out.
• Andreessen Horowitz's investment in Coinbase — a fund returner several times over.

These aren't the GPs being lucky once. The whole fund's return profile depends on hits like these. A fund without a fund returner usually returns 1-2x at best, which is below what LPs need to justify the illiquidity and risk.

Why this drives investor behavior

If you internalize the power law, a lot of VC behavior starts to make sense.

VCs swing for outliers, not consistent base hits

In a normal-distribution business, a steady 2-3x return on most investments would be a great result. In VC, that produces a mediocre fund. The math forces investors to look for companies that could be 50-100x outcomes, not companies that are likely to be 3-5x.

"Okay" companies are worse than passes

This is counterintuitive. If you spend two years and meaningful diligence resources on a company that ends up returning 2x, you've made money. But you've also spent the time and capital that could have gone into hunting for the actual outliers. In a fund where one bet drives most of the return, okay companies are an opportunity cost masquerading as a result.

This is why experienced investors say things like "I'd rather have a clear 0 or a clear 100x than a maybe-3x." It's not posturing. It's the math.

"Failure" is priced in

A fund where 50%+ of investments are losses isn't broken. That's the structure working as designed. The job isn't to avoid losses, it's to make sure you don't miss the outliers.

"Concentration is your friend" is mostly wrong

It's tempting to think: if outliers drive returns, write fewer, bigger checks. The catch is that you can't tell ex ante which company is the outlier. The outlier looks like every other early-stage company in year 1 or 2. You need many shots on goal to give the outlier a chance to surface. Top funds typically write 25-40 checks per fund for this reason — not because they're indecisive, but because the math demands enough swings.

(There are exceptions. Concentrated funds exist and sometimes outperform — Benchmark for example — but they require very high pick rates and a different operating model.)

Common misunderstandings

▸Average vs median — why headline numbers lie

A fund where one investment returns 100x and twenty-four return 0x has an "average" return of 4x per investment. But the median is 0. Headline averages are almost always misleading in venture; ask for the distribution.

▸Confusing power law with "high failure rate"

The power law isn't just "many bets fail." It's that the right tail is enormous. A business with 90% failures and 10% modest wins is not a power-law business; it's a bad-bet business. VC works because the 1-2 winners go to the right-tail extreme.

▸Treating it as a probability you can game

It's tempting to think: "The power law says one in 25 will be huge — so if I invest in 25 companies, one will hit." That's not how it works. The 1-in-25 is empirical, not guaranteed. A specific portfolio of 25 mediocre companies will return zero outliers. The skill is in picking the 25.

▸"I'll just back the obvious winners"

By the time a winner is obvious, you can't get into the round, and the price has moved past the venture math. Outliers are obvious in retrospect. They aren't obvious at Series A.

What it means for an analyst

A few practical implications for the seat you're in or about to be in.

1. Read every deal as if it might be the outlier. Most won't be. But the analyst's job is to keep the fund's pattern recognition fresh, and you do that by treating every pitch seriously enough to notice when something different walks in.

2. Don't be the one who passed on the outlier. Every famous VC has a list of misses. Pattern: the things they passed on weren't crazy at the time. They looked merely "interesting." The discipline is to ask, every time, "what would have to be true for this to be a 100x?" — and to take the answer seriously instead of dismissing it.

3. Don't argue against losses; argue against missed winners. If a partner is looking at a deal and the worst case is "we lose $1M and look stupid," that's fine. The fund can absorb it. The much worse case is "we don't write the check and it returns the fund for someone else."

4. Calibrate your IC writeups to the power law. A memo that says "this could realistically return $30M on a $1M check" is more useful than a memo that says "we're 80% confident in a 3-5x outcome." The IC is making a portfolio-level decision; show them the upside scenario, not just the central case.

The power law shapes everything: portfolio construction, deal selection, founder targeting, even how you interview for a VC job. Internalize it, and a lot of "VC weirdness" becomes ordinary economics.