Power law
A power law is a relationship where one quantity changes as a fixed power of another. In finance, the term often describes a distribution in which a few very large outcomes account for a large share of the total.
Key takeaways
- A power law is a specific mathematical pattern, not just another name for unequal results.
- Venture investors use the idea to explain why one exceptional exit can outweigh many losses.
- Fund size and ownership at exit determine how large an outcome needs to be to matter to an investor.
- A concentrated portfolio return does not, by itself, prove that the underlying distribution follows a power law.
How it works
The general power law formula is:
y = Cx^k
Here, C is a constant and k is the exponent. Multiplying x by two multiplies y by 2^k. For example, if k = 2, doubling x quadruples y.
In a power-law distribution, the probability of exceeding a given outcome size falls as a negative power of that size. Extremely large outcomes are rare, but more likely than under an exponential-tail model. Other distributions, including lognormal distributions, can also produce big outliers, so researchers must test the fit rather than infer it from a few winners.[1]
In Zero to One, Peter Thiel and Blake Masters apply the idea to venture capital: investors should seek companies with the potential to return an entire fund.[2] That is an investment strategy, not a promise that every portfolio will produce such a winner.
One exit returns all $50 million invested
A hypothetical fund invests $50 million across 20 startups, with $2.5 million in each. Ignore fees, carried interest, taxes, timing and follow-on investments.
| Outcome | Companies | Total invested | Total proceeds |
|---|---|---|---|
| Complete loss | 15 | $37.5M | $0 |
| Each returns twice its investment | 4 | $10M | $20M |
| One returns 20 times its investment | 1 | $2.5M | $50M |
| Total | 20 | $50M | $70M |
The fund's gross investment multiple is:
$70M ÷ $50M = 1.4x
The winner supplies $50M ÷ $70M = 71.4% of proceeds. It returns the fund's entire original investment, yet the portfolio as a whole earns only 1.4x before fees and carried interest.
If the fund owns 10% of the winner at exit, that company needs a $500 million equity exit to deliver $50 million. This calculation treats proceeds as proportional to ownership and ignores preferences or other payout adjustments.
For a founder, this explains why an investor might reject a business with a credible, profitable exit: the likely proceeds may be too small for that fund. A smaller fund or an angel could judge the same opportunity differently.
What the investment document actually says
"Power law" is an investment thesis, not a payout clause.
YC's post-money SAFE with a valuation cap and no discount sets conversion mechanics for an equity financing and the treatment of a liquidity event or dissolution.[3] The investor's contractual rights come from those provisions, not from an expectation of outsized returns.
Question a return forecast that assumes a huge winner without showing the required exit price and the investor's ownership after dilution. Neither an optimistic portfolio model nor a large addressable market changes the signed payout terms.
Power law versus the 80/20 rule
| Term | What it means |
|---|---|
| Power-law distribution | A mathematical distribution whose tail follows a power relationship. It has no universal 80/20 split. |
| 80/20 rule, or Pareto principle | A rule of thumb that a small share of inputs produces a large share of results. It need not be exactly 80% and 20%, or follow a power law. |
Use "highly concentrated returns" when that is all the evidence supports. It is more precise than claiming a statistical distribution you have not tested.
Questions founders ask
What is power law in simple terms?
In venture-capital shorthand, it means a few exceptional investments can do most of the work. Strictly, a power law describes how quantities relate mathematically; it does not mean merely that some outcomes are much bigger than others.
What is an example of a power law?
The fund above illustrates the concentrated returns investors have in mind. It does not establish a power-law distribution. To make that claim, you would need data and a statistical comparison against alternative distributions.
Does the power law mean founders should take more risk?
No. A fund spreads its bets across companies; a founder's time and wealth are concentrated in one. An investor's need for an enormous exit is not proof that the riskiest strategy is your best choice.
