The arithmetic behind the line
Early-stage returns are a power law: in any honest sample, most cheques go to zero or thereabouts, a modest few return capital, and a tiny number pay for everything. Portfolio math under a power law is unforgiving about breadth — you need enough independent positions for the tail to plausibly show up. Fifteen to thirty cheques is the working consensus for a real chance at the distribution's good side. A 5% allocation of net worth across ~20 positions lands each cheque near 0.25–0.5% — the house states the ceiling as 1% per cheque, 5% for the whole book, which is already the aggressive end of sensible.
The follow-on trap
The cheque is not the commitment. Every successful position will ask again — pro-rata rights exist to be exercised — and a book with no follow-on reserve faces a miserable choice at exactly the moment of success: dilute out of your winners or overshoot your allocation. The convention: hold reserve equal to the initial cheque (1:1), which means a $50,000 cheque is a $100,000 commitment, and the 1% line should be read against that doubled figure. Dropping the reserve halves the commitment and buys the dilution seat — a legitimate choice, made honestly or not at all.
When the line bends
Genuine edge — operating expertise in the exact space, real deal flow, information the market lacks — justifies concentration the way it does everywhere. Liquidity justifies less than people think: the money is gone for 7–12 years regardless of net worth. And "I can afford to lose it" is a statement about solvency, not sizing; the line exists so the book can survive being right about the asset class's math, not just wrong about one company.
The house position
The Angel Cheque, Sized reads any cheque against net worth, current exposure and reserve policy, and prints the full commitment the friendly number implies. Illiquid paper concentration belongs in the same conversation as public-stock concentration — instrument No. 06 — and the two lines should be read together.