Venture Capital: A Business Model Built on Failure
4 steps · one page
In short
Venture capital funds companies too young, too small, and too uncertain to borrow or to list.
The distinctive thing about it is not the technology or the enthusiasm — it is the return distribution, which is unlike anything else in this portal and which explains every feature of how the industry behaves. Most investments lose everything, and the model is designed around that rather than in spite of it.
The power law, computed
In most asset classes, outcomes cluster around an average. In venture capital they do not. A small number of investments produce almost all the return, and the majority produce nothing — which is called a power-law distribution and is the single most important fact about the asset class. The table below runs a representative twenty-company portfolio at $1 million each. The outcome shares are illustrative rather than measured, and they are this pillar's canonical venture case.
| Outcome | Share of portfolio | Companies | Return multiple | Proceeds |
|---|---|---|---|---|
| Total loss | 50% | 10.0 | 0× | $0m |
| Capital returned | 20% | 4.0 | 1× | $4m |
| Modest success | 20% | 4.0 | 3× | $12m |
| Strong success | 7% | 1.4 | 10× | $14m |
| Outlier | 3% | 0.6 | 50× | $30m |
| Total | 20 | 3.00× | $60m |
Three things fall out of that table, and they define the asset class. Half the portfolio returned nothing and the fund still tripled. A 3.00× gross multiple over ten years is about 11.6% annually — and after the standard fee stack it is 2.43×, or about 9.3% net, per the pillar's arithmetic. The concentration is extreme. The outlier tier — well under one company in this portfolio — produced 50% of all proceeds, and the top two tiers together produced 73%. Which means the entire result depended on roughly two investments out of twenty. And that explains the industry's behaviour. A venture investor is not trying to avoid losses, because losses are the expected case; they are trying to be in the small number of companies that return the fund, which is why they seek outcomes that could be very large rather than outcomes that are likely to be positive. A strategy that reliably returned 2× on every investment would be a bad venture strategy, which sounds paradoxical and follows directly from the arithmetic. Three mechanics that follow from the distribution. Dilution. Companies raise repeatedly, and each round issues new shares — so an early investor's percentage falls unless they keep investing, and a headline valuation increase does not automatically mean a proportional gain for an early holder. Preferences and structure. Venture shares usually carry liquidation preferences and other terms giving them priority over ordinary equity, which means a company sold for a modest amount can return capital to preferred holders and nothing to founders and employees — the priority logic the distressed article established, appearing at the other end of the corporate life cycle. And the exit dependency. Returns require a sale or listing, so the asset class depends on exit markets being open, and a fund holding good companies in a closed exit market has unrealised marks rather than money.
Why access is the whole question
The dispersion problem the private-equity article described is more extreme here, for a structural reason: the best investments are competitive and the best firms win them. A company with several willing investors chooses, and it frequently chooses on reputation and prior track record — which means the outcome distribution is not drawn randomly by every fund. The top firms see and win a different set of opportunities. Three consequences. Access to the strongest funds is closed in practice. They are oversubscribed and allocate to existing relationships, so the arithmetic above describes a fund a reader probably cannot invest in. Persistence is stronger here than elsewhere in alternatives, which sounds like good news and is not: if returns persist because the best firms get the best deals, that is a barrier rather than a signal a new investor can act on. And the accessible versions are structurally different — later-stage vehicles, funds-of-funds with an extra fee layer, or platforms offering single-company exposure, which is the worst version of a power-law strategy because a single venture position is overwhelmingly likely to be one of the losses. What the model genuinely produces, stated properly. Venture capital funds companies that cannot be funded any other way — no bank lends against an idea, and no public market will list a pre-revenue company. It is the mechanism by which a substantial portion of new technology gets built, and the failures are the cost of that mechanism rather than a defect in it. The governance and expertise that good investors bring to young companies is also real. What a reader should carry: the arithmetic works only across a portfolio large enough to contain an outlier, the outliers are concentrated in firms a reader cannot access, and any exposure that is not diversified across many companies is not running the strategy — it is holding one position in a distribution where most positions go to zero.
Worked example
Worked example (fictional; figures computed). Kesterline Ventures raises $20 million and invests $1 million in each of twenty companies. The outcome. Ten fail completely. Four return the capital. Four return 3×. One and a half, on the distribution above, return 10×. And a fraction of one company returns 50×, producing $30 million — half of everything the fund made. Total proceeds $60 million, a 3.00× gross multiple, 11.6% annually, and 9.3% net after fees. Now the point of the article. Remove the single outlier and the fund returns $30 million on $20 million — a 1.5× multiple, roughly 4.1% gross annually, and after fees about 2%. The difference between a good venture fund and a poor one was one investment out of twenty. And the diversification arithmetic, which is the part that matters for anyone considering a smaller version. If roughly one company in ten produces an outsized return, then across twenty positions the chance of holding at least one is about 88%. Across five positions it falls to 41% — so an investor who can fund five companies is more likely than not (59%) to hold none of the winners, and their expected experience is the losses without the outlier that pays for them. That is not bad luck. It is the distribution working exactly as described on a sample too small to contain its own tail. (All names fictional; outcome shares illustrative and computed from this pillar's canonical venture case; probabilities assume independent draws at a one-in-ten rate.)
Frequently asked
9 questions
What makes venture capital different?
The return distribution. A small number of investments produce almost all the return and the majority produce nothing — a power law. Every feature of how the industry behaves follows from that.
How concentrated are the returns really?
On the illustration here, the outlier tier produced 50% of all proceeds and the top two tiers produced 73% — so the entire result depended on roughly two investments out of twenty.
Why don't VCs try harder to avoid losses?
Because losses are the expected case. They're trying to be in the few companies that return the whole fund, which means seeking outcomes that could be very large rather than outcomes likely to be positive. A strategy that reliably returned 2× on every investment would be a bad venture strategy — which sounds paradoxical and follows from the arithmetic.
What is dilution and why does it matter?
Companies raise repeatedly and each round issues new shares, so an early investor's percentage falls unless they keep investing. A headline valuation increase doesn't automatically mean a proportional gain for an early holder.
What are liquidation preferences?
Terms giving venture shares priority over ordinary equity. They mean a company sold for a modest amount can return capital to preferred holders and nothing to founders and employees — the same priority logic that governs distressed situations, appearing at the other end of the corporate life cycle.
Why do exit markets matter?
Because returns require a sale or listing. A fund holding good companies in a closed exit market has unrealised marks rather than money.
Isn't it good news that returns persist in venture?
Not for a new investor. If returns persist because the best firms see and win the best deals, that's a barrier rather than a signal you can act on — the outcome distribution isn't drawn randomly by every fund.
Can I do this with a handful of investments?
The arithmetic says no. If roughly one company in ten produces an outsized return, twenty positions give you about an 88% chance of holding at least one. Five positions give 41% — so you're more likely than not (59%) to hold none of the winners, and your expected experience is the losses without the outlier that pays for them. That isn't bad luck; it's the distribution working on a sample too small to contain its own tail.
What does venture capital actually accomplish?
It funds companies that cannot be funded any other way — no bank lends against an idea and no public market lists a pre-revenue company. It's the mechanism by which a substantial portion of new technology gets built, and the failures are the cost of that mechanism rather than a defect in it.
References
- SEC Investor.gov — Accredited Investors: Updated Investor Bulletin (venture capital funds as exempt private offerings; reduced prescribed disclosure) —
- SEC Investor.gov — Private Equity Funds (private-fund structure, illiquidity, long horizon) —
- FINRA — Alternative and Emerging Products (private placements and non-traditional structures) —
Educational and informational only — not investment advice, a recommendation, or an offer to buy or sell any security. Investing involves risk, including the possible loss of principal. Worked examples use fictional companies and figures.