What our Buyout price says at year 4, 5, and 6 — checked against the dealer desks
Every H1 and full-year, secondaries intermediaries survey the buy side and publish an average price for LP-led Buyout stakes. We ran our own models at the exact same young ages and lined the two up.
A second kind of check
Our overfitting battery checks a model against its own held-out history — the same data, withheld and re-tested. That's necessary but it's an internal check. A different, external one is whether the price a model actually outputs lands anywhere near what real secondaries desks report the broad Buyout market is clearing at, published twice a year by the intermediaries who broker these deals (Jefferies, Campbell Lutyens, and others). We keep a running, cited dataset of those published figures — this checks our Buyout output against it, specifically at the youngest ages a fund is priced.
The published market average
Campbell Lutyens, FY 2024 LP-led Buyout
Jefferies, H1 2025 LP-led Buyout
simple average of the two
Our own output, year by year
We priced the same representative Buyout snapshot — NAV, paid-in, distributed, and unfunded held fixed — at ages 4, 5, and 6, once per model, at our platform default 15% required return. Each cell below is the fair price, with the model's own P25–P75 Monte-Carlo band in parentheses — the same band every live result on Equisect ships with, not a point estimate dressed up as a single number.
| Model | Age 4 | Age 5 | Age 6 |
|---|---|---|---|
| TA (baseline) | 97% (59–113%) | 93% (56–107%) | 102% (63–117%) |
| Equisect Bayesian | 72% (44–107%) | 74% (46–108%) | 83% (57–114%) |
| Equisect Cohort | 65% (41–96%) | 62% (40–97%) | 68% (44–110%) |
All figures are fair price as a % of the fund's current NAV, banded to the model's own P25–P75 Monte-Carlo range. Every band here comfortably contains the published survey average of 91.5%, even where a model's own fair-price point estimate doesn't.
What a lower required return does to the price
15% is our platform default, not a fixed constant — required return is a number you set, and it's the single biggest lever on the fair price. We re-ran the identical snapshot at a 10% required return, a hurdle closer to what a more return-tolerant buyer might underwrite a stable Buyout position at.
| Model | Age 4 | Age 5 | Age 6 |
|---|---|---|---|
| TA (baseline) | 113% (69–131%) | 107% (65–122%) | 114% (71–132%) |
| Equisect Bayesian | 84% (50–126%) | 85% (53–124%) | 93% (63–128%) |
| Equisect Cohort | 81% (53–120%) | 77% (51–117%) | 81% (54–127%) |
Dropping the required return from 15% to 10% lifts every single price — by 10 to 16 points depending on model and age. TA moves from clustering around the survey average to overshooting it at every age. Equisect Bayesian's age-6 price (93%) lands almost exactly on the survey average (91.5%). Equisect Cohort closes most, but not all, of its gap — it's still the cheapest of the three at either required return.
Reading the gap
At our 15% platform default, the TA baseline lands close to the published range at every age we tested — within 6 points of the two-source average, and almost exactly on it at age 5 (93% vs. 91.5%). Equisect Bayesian, the one we actually ship as the validated price, is meaningfully cheaper across all three ages — 8 to 19 points below the survey average. Equisect Cohort — the one with the cleanest overfitting verdict of the three — prices cheaper still, 24 to 29 points below survey, the widest gap of any model here. None of that is a bug we're papering over: each model prices to a required return an investor should demand given the fund's own risk and factor exposure, not to whatever the broad market happens to be clearing at in a given half. A dealer-desk average blends every seller's motivation — some distressed, some patient — into one number; our models price one specific return hurdle each, and Equisect Cohort's own dispersion-driven band is the widest of the three, which is exactly why its point estimate sits furthest from a single blended market number. The size and consistency of these gaps is worth sitting with rather than averaging away, which is exactly why we're publishing all three instead of just the model that happens to look best.
What to do with it
Treat the published survey average as a market-clearing sanity check, not a target any one model should hit. If a model's fair-price point sits close to the survey number, that's one more data point that the price is reasonable. If it doesn't, check whether the survey number at least falls inside that model's own P25–P75 band before concluding anything is wrong — here, it does, for all three models, at every age. A wide gap between a model's point estimate and the published market average is a prompt to look at required-return assumptions and your own read on the deal's risk, not proof either number is simply wrong.
Research and software, not investment advice. Survey figures are published third-party market data (Campbell Lutyens Secondary Market Overview FY 2024; Jefferies Global Secondary Market Review H1 2025), reproduced here for comparison, not verified independently by Equisect.