Wealth taxes, billion-dollar ambitions: what two founders wrote this month

Two essays from Valley founders landed in May 2026, each addressing a question early-stage builders quietly wrestle with but rarely discuss out loud: how should I think about my own equity? And is aiming for a billion-dollar outcome a sign of clarity or delusion?

Paul Graham's latest piece is a policy essay on its surface. Garry Tan's is a morale boost. Read together, they form something more useful — a framework for how founders should calibrate ambition and understand what the equity they're working for actually represents.

"How to Convert Between Wealth and Income Tax" — paulgraham.com, published 2026-05-22 1

Graham wrote this as a contribution to the wealth-tax debate, but the underlying math is directly useful to any founder carrying meaningful equity. His core argument: a 1% annual wealth tax is not a small thing — it is equivalent to a roughly 20% income tax on the returns that wealth would otherwise generate.

The conversion formula depends on the assumed rate of return on capital. At a 5% return (a conservative baseline for diversified assets), a 1% wealth tax consumes 20% of the year's gains before you've realized anything. At a higher growth rate — which describes early-stage equity well — the effective rate looks lower in percentage terms, but the absolute dollars consumed compound against you year over year.

The essay makes the mechanism legible. Your cap table position is a wealth asset — not yet liquid, but mathematically subject to the same conversion logic. If your country or state introduces a mark-to-market or wealth tax on private holdings (a live policy debate in several jurisdictions in 2026), the drag on your eventual post-liquidity outcome is not the headline rate. It is that headline rate converted through Graham's formula into an income-equivalent rate, applied annually from the moment of imposition.

Founders evaluating whether to relocate, restructure their equity, or accelerate a liquidity event should run this math explicitly. Graham's point — left implicit in the essay but obvious once you see it — is that wealth taxes are systematically underestimated by people accustomed to thinking in income-tax terms. The policy arena he's writing for is also the regulatory environment your company will eventually exit into.

The boundary condition Graham makes explicit: this conversion logic assumes capital is productive (generating returns). For illiquid, pre-revenue equity held by founders who are also salaried employees of their own companies, the "rate of return" is not observable yet. The practical takeaway is to run the analysis at the time a liquidity event is on the horizon — not retroactively, when tax structures are already set.

"Yes, You Can Earn a Billion Dollars" — garryslist.org, published 2026-05-08 2

Tan published this in response to a skeptical claim that founding a company cannot realistically produce billionaire-level personal wealth for most founders.

His rebuttal is data-driven: he draws on Paul Graham's two decades of tracking which YC founders become billionaires, arguing the base rate is higher than the critics suggest — and that the critics' argument systematically discourages early-stage founders from targeting large problems.

The structural point Tan makes is that billion-dollar personal outcomes require building companies that create many billions in enterprise value. That sounds obvious, but it shapes which problems are worth tackling. A founder aiming at a $100M outcome will size her problem accordingly. A founder aiming at a billion-dollar personal outcome needs a company capable of generating $5–10B in equity value (assuming realistic dilution across multiple rounds). That selection effect determines the market she enters, the team she builds, and the product decisions she makes — from day one.

Tan's practical argument for early-stage AI founders specifically: the current AI-native company creation window is unusually large. Costs of building are low, the underlying model capabilities are compounding, and the gap between ambitious founders and cautious ones is growing. This makes it the wrong moment to self-limit.

The action implication: if you're currently scoping your company at $50–100M because you find it more "realistic," audit whether that ceiling comes from the actual market dynamics you've analyzed — or from an unexamined prior that ambitious outcomes don't happen to people like you. Tan's point is that the prior is wrong. Graham's data backs it.

The boundary condition Tan doesn't fully address: this argument applies cleanest to founders who have found genuine product-market signal and are now making bet-size decisions. It is less directly applicable to pre-product founders still searching for the right problem — where the more pressing question is finding a true wedge, not sizing the ambition attached to it.

Both pieces push against the same error: founders who underestimate what they're actually doing.

Graham's essay is about underestimating the cost of taxation applied to wealth creation — a technical miscalibration that compounds quietly. Tan's essay is about underestimating your right to aim at large outcomes — a psychological miscalibration that shapes every early decision.

For a founder in 2026 building an AI-native company, the combined lesson is specific: your equity position is a meaningful wealth asset that will be subject to policy environments that are genuinely in flux; and the problem you choose to attack should be sized by what the market can support, not by what feels socially acceptable to claim. Run Graham's conversion math before your next term sheet. Read Tan's essay before you scope your next pivot.