Insights
Notes on hard-technology investing.
Writing on private markets, long-horizon capital, and building the future deliberately.
Conviction is a research problem.
Holding through the hard middle looks like temperament. It's mostly the compound interest of work done before the check was written.
ReadA moat you can't A/B test.
Software defends with metrics you can read weekly. Deep tech defends with physics, certification, and accumulated proof — moats that move on a different clock.
ReadCapital intensity is a filter, not a flaw.
The money and years hard technology demands are usually framed as its weakness. They're also the reason the winners face so few challengers.
ReadSecondaries are private markets' price discovery.
An illiquid asset has no ticker. The secondary bid is the closest thing it gets — a real price, from a real buyer, between the marks.
ReadRoll or cash out.
When a continuation vehicle forms, every LP gets the same letter and the same deadline. The choice looks like a liquidity election. It's a fresh investment decision.
ReadThe return is in the middle.
Entry and exit get the attention. The decade between them is where the investment is actually made or lost.
ReadStage capital to physics, not to growth.
A software round is priced off traction already shown. A hard-technology round has to be priced off proof — the milestone the physics actually turns on.
ReadHard technology is always five years away.
The timeline that keeps sliding isn't a sign the bet is wrong. It's the thing you have to underwrite directly, because it will slide again.
ReadWhat a fairness opinion settles, and what it doesn't.
It answers a narrow question about price. The broader questions it gets asked to answer — is this a good deal, is the conflict resolved — were never in its scope.
ReadA continuation fund puts the GP on both sides of the table.
The same firm sells the asset and buys it, prices it and manages it. A fairness opinion and an LP vote are the machinery built to hold that conflict in check.
ReadWhy a good company outlives the fund that bought it.
A fund has a fixed term. The companies inside it don't. The continuation vehicle exists to resolve a mismatch that was there from the first close.
Read