Changed My Mind
Showing the work.
Entries are never edited after publication. Corrections are added as new rows.
| Date | I used to think | I now think | What moved me |
|---|---|---|---|
| October 20, 2026 | Early TVPI was a useful signal of manager quality. | Early TVPI is mostly a signal of what other investors paid. | Comparing year-three and year-ten numbers across a set of funds: the rank order barely held. |
| October 13, 2026 | Seed valuations were primarily driven by market comparables. | Seed valuations are primarily an output of lead ownership targets and round size. | Sitting in on twelve pricing discussions this year. Comparables came up in all of them, but never as the deciding input. |
| October 6, 2026 | Non-lead checks were mostly decided by the company's metrics. | Non-lead checks are mostly decided by the quality and commitment of the lead. | Reviewing two years of our own syndicate decisions: lead quality predicted our yes rate better than revenue growth. |
| September 29, 2026 | Pro rata rights were worth exercising by default in top-quartile companies. | Pro rata should be exercised only when the next round's price still clears our fund return target. | Modeling our last fund: default pro rata into expensive rounds lowered net multiple by 0.3x. |