Published at: 2026-05-09T21:07:37+05:30
Thesis
Venture capital is not a game of picking many “pretty good” companies. It is a game of not missing the few companies that compound into category-defining outcomes. That single fact reshapes everything: portfolio construction, ownership targets, reserves, deal terms, and even what kind of help you offer founders.
My claim is simple: if you want venture-style returns, the only rational strategy is to buy more of your winners.b In practice, that means designing your fund, your time, and your decision-making so that you can identify the handful of breakouts early enough, and then keep (or increase) ownership as conviction rises.
This sounds obvious until you look at how most portfolios are actually managed. Many investors spread initial checks widely, underwrite follow-ons lazily, and treat pro rata like a courtesy rather than a core asset. The result is a portfolio that correctly “finds” a winner but fails to own enough of it to matter.
The power law is not a motivational poster. It is arithmetic.
Context
The modern venture market has two defining tensions.
First, the distribution of outcomes is brutally skewedb. A small fraction of companies drive most of the value. Cambridge Associates has described the dynamic plainly: nearly 90% of venture’s value has been driven by the top 10% of companies.[1]ahttps://www.cambridgeassociates.com/insight/2026-outlook-private-equity-venture-capital-views/
Second, the industry is operating in an era where liquidity has been inconsistentb and fundraising has become more selective. The PitchBook-NVCA Venture Monitor has emphasized the continued pressure created by lack of distributions and slower exit markets.[2]ahttps://nvca.org/pitchbook-nvca-venture-monitor/
When exits are slower, markups are noisier, and the “average” company is harder to sell, the power law becomes even more tyrannical. In a tough regime, mediocre outcomes do not quietly become fine outcomes. They become dead weight.
So the question is not “How do I build a portfolio of good startups?” It is:
- How do I find the companies with a realistic path to outlier scale?
- How do I avoid being diluted out of the handful that actually get there?
- How do I reserve capital and attention so I can double down when reality reveals itself?
Key ideas
1. The power law is not a belief. It is a portfolio constraint.
In public markets, diversification smooths outcomes because returns are less skewed and liquidity lets you rebalance. In venture, the distribution is lopsided and liquidity is episodic. One company can return the entire fund, and a large number of companies can return near-zero.
This is why pro rata, follow-ons, and ownership are not “optional optimizations”. They are the mechanism by which you turn a correct early signal into actual fund-level returns.
There is a practical corollary: your initial check is often not your main betb. It is an option on future ownership.
You pay for the right to learn.
Later, once you have real evidence of product pull, retention, founder execution, and market expansion, you decide whether to exercise that option.
That is the venture version of rationality: invest small to buy information, then invest large when uncertainty collapses.
2. Follow-on strategy is where funds quietly win or lose
Many new investors imagine venture as a single moment: the seed check, the story, the memo. But most of the economic outcome is determined by what happens after.
Follow-on investing is not just “investing again.” It is a disciplined approach to:
- Reserving capitalb so you can actually participate.
- Measuring tractionb in a way that predicts compounding.
- Concentratingb exposure into the few companies that demonstrate the right kind of momentum.
Carta’s overview of follow-on investing captures the basic idea: follow-ons are a key way for GPs to “double down” on the most promising companies as the market concentrates capital into later stages.[3]ahttps://carta.com/learn/private-funds/management/portfolio-management/follow-on-investment/
This matters because the biggest value creation is usually later than seed. If you are diluted down to an irrelevant percentage before the company becomes obviously great, you have, functionally, donated your early pattern recognition to later-stage capital.
3. Pro rata is not about fairness. It is about being allowed to compound.
Pro rata rights exist because dilution is the default outcome of successful companies. The best companies raise multiple rounds. Each round reduces early ownership unless you keep buying.
The investor’s job is to treat pro rata like an asset.
Fred Wilson has written that USV values pro rata and exercises it frequently, often making multiple investments in the same company over time.[4]ahttps://www.goingvc.com/post/why-investors-fight-for-pro-rata-rights-and-what-founders-should-know
That quote is not a flex. It describes how ownership is actually built.
If one company can return your fund, then not being diluted out of that one company is the most important thing you do.
4. “Buying more of your winners” requires reserves and decision rules
The strategy sounds clean. The execution is messy. The mess comes from two constraints.
Constraint A: capital.b If you do not reserve meaningful capital, you cannot follow on. You might have a contractual right to buy, but no budget to exercise it.
Constraint B: attention.b Follow-on decisions require updated beliefs. Updated beliefs require signal. Signal requires time with founders, customers, metrics, and the market.
So the follow-on strategy must be designed upfront:
- Reserve policy (for example, a large portion of the fund held back for later rounds).
- A simple set of “promotion” criteria: what evidence upgrades a company from “option” to “core position.”
- A kill list: what evidence downgrades a company quickly so you do not waste reserves.
Without rules, the fund will follow on based on feelings, relationships, or sunk costs.
Rules do not make you correct. They make you consistent.
5. The skill is not predicting unicorns. The skill is recognizing compounding early.
Most investors over-index on narrative. Great narratives are cheap.
Compounding is not narrative. Compounding shows up as:
- Customer behavior that improves over time.
- Distribution that gets cheaper as the product spreads.
- Expansion dynamics: new use cases, new buyer types, new geographies.
- A team that becomes more effective as complexity increases.
The art is to identify the causal engine.
Many startups can grow for a year by spending money.
Far fewer can grow for a decade by building something that earns momentum.
If you want to buy more of your winners, you need to know what “winner” means beyond a mark-up.
6. Capital formation data hints at why ownership matters
The U.S. private markets run heavily through exempt offerings. The SEC’s Regulation D statistics show the sheer scale and persistence of private fundraising.[5]ahttps://www.sec.gov/data-research/statistics-data-visualizations/regulation-d-offerings
You do not need to love this fact to understand its implication: companies can keep raising private rounds for a long time. That long runway can be good for company building. It can also mean that the highest value inflection points occur years after your initial check.
If value inflects later, then early ownership without follow-on capacity is a mirage.
7. A portfolio is not a list of companies. It is a curve.
Most investors talk about portfolios as if they are collections.
But in venture, a portfolio is a curve: a distribution of outcomes. Your job is to shape the curve so that:
- You have enough shots to encounter an outlier.
- You have enough ownership in the outlier when it happens.
The key mistake is to treat the two as independent.
A large number of small checks increases the chance you “touch” a winner.
But if you do not reserve and concentrate, you will not own the winner.
You will have bragging rights and disappointing DPI.
8. Why this strategy feels emotionally hard
“Buy more of your winners” is rational. It is also psychologically uncomfortable.
It forces you to:
- Admit that most of your picks will not matter.
- Say no to follow-ons in companies you like.
- Put large amounts of money into a small number of outcomes.
Humans prefer symmetry and fairness. We want every company to get a chance.
Capital does not care about fairness.
A venture fund is a machine for concentrating exposure into reality.
Reality reveals itself unevenly.
Counterarguments
Counterargument 1: Concentration increases risk. Diversification protects the fund.
This is true in many domains.
But venture is a domain where returns are already concentrated. Diversification can reduce idiosyncratic risk, but it cannot erase the shape of the distribution. If the asset class is power-law driven, you still need exposure to the top tail.
Over-diversification creates a different risk: the risk of being right but not being paid.b
If you own 0.2% of your breakout instead of 2%, you can be correct and still lose.
Counterargument 2: Follow-ons are overpriced. The best returns come from early entry.
Early entry can be amazing when it works.
But the biggest value creation often happens as a company moves from “credible” to “inevitable”. If the company’s advantage is real, later rounds can still be excellent investments, especially if they let you preserve ownership through the steepest portion of the value curve.
Also, early entry without pro rata is not really early entry. It is a lottery ticket with forced dilution.
The right comparison is not “seed vs. later.” It is “seed + follow-on” vs. “seed only.”
Counterargument 3: Pro rata is founder-unfriendly and signals distrust
Some founders resent aggressive pro rata. They prefer a cap table optimized for new brand-name investors.
There is a legitimate point here. A company is not obligated to serve early investors’ ownership goals.
But a good founder also understands incentives. If an investor is truly helpful, aligned, and willing to keep supporting the company, it is not irrational for that investor to want the option to maintain their stake.
The best version of pro rata is quiet. It is not used as leverage. It is used as a tool of alignment.
Counterargument 4: This strategy favors big funds and incumbents
It often does.
Big funds can reserve more, follow on more, and sometimes get better access. That is part of the structural inequality of the asset class.
But smaller funds can still apply the principle.