Pillar
Portfolio Construction Discipline
Portfolio construction is the set of commitments a fund makes about how it will deploy capital: how many companies it will back, at what ownership, at what stage, across how much of the market, and how much it will hold in reserve. Every one of them is settled before the first cheque goes out, and each becomes considerably harder to move afterwards.
4 pages
- Why no one can honestly tell you what your fund will returnA return forecast would be worth more than a consistency reading, and it cannot be produced honestly from what a fund knows about itself. What a configuration can support instead, including the 20-slot rule question.
- Should you model your fund forward or check the one you haveA construction model projects forward from assumptions. A construction check reads a declared configuration for internal consistency. Why they answer different questions and why a firm can want both.
- When your fund stops looking like the one you raisedConfiguration drift is the gap between the fund a General Partner declared and the fund it became. Why it is the ordinary condition rather than a failure, and how it is surfaced without guessing at intent.
- Why portfolio construction advice runs out once you have closedThe existing writing on venture portfolio construction is good and is almost entirely addressed to a General Partner who has not closed a fund yet. What the post-close question actually is, and why it is harder.
How to read this pillar
Discipline here means consistency rather than caution. A Conviction-stage firm running a concentrated portfolio at high ownership is exercising construction discipline exactly as much as a Continuity-stage firm running a broad portfolio at moderate ownership, provided that in each case the rest of the configuration supports the choice. The three lifecycle stages occupy different frontiers, and the model does not treat a newer firm as a weaker one. What it looks for is contradiction: a fund declaring a portfolio count, an ownership target, and a fund size that cannot all hold at once is describing a construction that cannot be executed as written.
The same set of decisions is read by three engines, each asking a different question about it. Portfolio Efficiency works at the fund level, Firm Design Congruence at the firm level, and Cross-Fund Concentration across every fund the firm runs, with the Architecture Score integrating the three.
The model evaluates how a fund is built, not how it will perform. It does not rate firms and it does not produce investment recommendations. A construction the model reads as coherent can still return poorly, and one that surfaces tensions can still return well. Market timing, team dynamics, and deal access sit outside what a configuration can express. The reading is an instrument that supports judgment rather than a substitute for it.