Most venture investments lose money. The fund is saved by one or two outliers, so every new deal is judged on whether it could be that outlier.
Venture returns are not spread evenly. In a typical early-stage portfolio, around half the companies return less than the money put in, a handful return a few times their money, and one or two return more than all the others combined. Statisticians call this a power law. VCs live by it.
Picture a fund that backs 30 tech startups with equal cheques (an illustrative mix, not data):
That one outlier often produces more than everything else in the portfolio put together. Remove it and a good fund becomes a poor one. Run it yourself in the VC portfolio simulator.
If half the portfolio returns little or nothing, the winners must make up for them and still deliver a strong overall return. So at the moment of investing, a VC needs to believe each company could return at least ten times their money. Most will not. That is understood. But a company that could only ever return two or three times cannot fix the maths, however safe it looks. This is the 10x return model we wrote about in The 10x return model of VC funding.
Many partners go further and ask whether a single investment could return the entire fund. The sum is simple:
Exit value needed = fund size ÷ the stake the VC owns at exit.
A £50m fund that owns 10% of your company when it is sold needs a £500m exit to get £50m back. Owning 7.5% after later rounds dilute it, the figure rises to about £670m. Try your own numbers in Can you return the fund?
General information, not financial, legal or tax advice. Figures marked as rules of thumb vary by fund, sector and market conditions.
The exit your company would need for one investment to return a whole VC fund, and which fund sizes suit your likely exit.
Open tool →Investor viewNewPlay a fund of 30 tech startups and see how one or two outliers carry the rest.
Open tool →Investor viewNewWork back from a future exit to the most a VC can pay today, and test your asking price.
Open tool →Work through it in Execution Sprints with tech founders who have raised, bootstrapped and exited.