Equisect prices real, named PE funds and secondaries — not synthetic peer baskets
Models trained and held-out validated against our own proprietary warehouse of over 1,100 cleaned funds. No black box: every model's out-of-sample honesty check is published right next to its price.
1,100+ named funds · 40k+ fund-quarters modeled · every price checked against real, held-out history
Built to survive scrutiny
Before you type in a client's positions, here's exactly what happens to the data and who's accountable for the number.
Your data stays yours
IC-memo drafting runs on a local, self-hosted model — nothing you enter is ever sent to a third-party AI API. See our privacy policy.
Every export is attributable
Every CSV and PDF carries your account, your plan, and a generation timestamp — reproducible for your own file, not a disposable screenshot.
The warehouse is ours, not licensed
Built and maintained in-house from over 2,500 distinct funds — never shared with, or resold by, a third-party data vendor.
The math is published, not claimed
Deflated Sharpe, bootstrap confidence intervals, FDR correction, and shape-fidelity scoring — read the full methodology before you trust a single price.
Questions before you commit real capital? Talk to our team — not a bot, a reply from someone who can walk through the methodology with you.
The problems we solve
Every card below is a question we've heard from a real desk, and the part of Equisect built to answer it.
"There's no market quote for this secondary."
Enter a snapshot, upload a fund's history, or look one up in our warehouse — get a fair price as a % of NAV at your required return, with a P25–P75 band, not a single unexplained number.
Price a fund → Data"GP-reported data is inconsistent and full of restatements."
We start from over 2,500 distinct funds spanning vintages back to 1976 and clean every one before a model sees it: reconciling duplicates, correcting restatements, matching each fund's history against itself point-in-time.
See the warehouse → Scale"I need a whole portfolio priced, not one fund at a time."
Upload a fund's — or a whole portfolio's — quarterly call, distribution, and NAV history as a CSV. Get each fund's forecast plus a NAV-weighted rollup: concentration, vintage ladder, liquidity profile, cashflow calendar.
Upload a portfolio → IC-ready"Turning this price into something my IC will actually approve."
Generate an IC memo with AI-drafted narrative sections — every stated figure checked against the underlying price before you see it — plus your own manual diligence fields, exported as a branded PDF.
See the IC memo → Honesty · technical"How do I know this model isn't just overfit?"
Every model we ship runs through the same battery quant funds use before trusting a strategy: Deflated Sharpe Ratio, bootstrap confidence intervals, and Benjamini–Hochberg FDR correction for how many variants we actually searched.
Read the methodology → Shape fidelity · technical"Does the price actually trace a real J-curve, or just the right return?"
A model can score well on return and still miss the shape of the cashflow path. We score every model on DTW, Fréchet distance, lag-optimal cross-correlation, and Wasserstein distance against real fund histories — published, not just claimed.
See the comparison →How it works
Four steps from a fund name to a defensible price.
Choose your data source
A manual snapshot, an uploaded fund/portfolio history, or a real fund already in our warehouse — no data entry needed for the last.
We price it against the warehouse
Three or more independent models fit on the same cleaned data: a cohort-curve roll-forward, a Bayesian public-market-factor model, and a quantile cohort or ledger model, each producing its own percentile band.
Every model publishes its own verdict
Deflated Sharpe, bootstrap confidence interval, FDR correction, and shape-fidelity scoring — right next to the price, not in a footnote.
Export a price, or a full IC memo
Download the fair price and projections as CSV, or generate an AI-drafted, figure-verified IC memo as a branded PDF.
A proprietary data warehouse, checked against reality
Raw PE data is notoriously messy — inconsistent reporting, restatements, survivorship gaps. We start from over 2,500 distinct funds spanning vintages back to 1976 and clean every one of them before a model ever sees it: reconciling duplicates, correcting restatements, matching each fund's history against itself point-in-time. What survives is the warehouse below.
Who it's for
Built around the desks that actually need a defensible number, fast.
Secondary buyers & GP-led deal teams
- Bid a real, named position against a warehouse-fitted fair price
- Stress the price across required return and wind-down age before you commit
- Attach an IC memo instead of building one from scratch
Fund-of-funds & LP portfolio teams
- Upload a whole book and get a NAV-weighted rollup, not fund-by-fund spreadsheets
- See concentration, vintage ladder, and liquidity profile in one place
- Track the same honesty verdict across every fund in the book
Family offices & wealth advisors
- Price a client's illiquid PE stake without an internal quant team
- Get a percentile band, not a single point estimate to defend
- Start on Snapshot, no fund history or warehouse access required
For family offices: the whole book, not just one fund
Pricing one GP-led secondary is a single decision. Knowing what the entire illiquid sleeve calls and returns next year is the harder job — and the one that actually drives your treasury planning.
- NAV-weighted portfolio rollup — one blended fair price across every fund you upload, not a spreadsheet you build yourself
- Concentration & vintage ladder — see exposure by manager, asset class, and vintage year before your next allocation decision
- Liquidity profile & cashflow calendar — peak call, breakeven year, and portfolio IRR, so treasury planning isn't a guess
Request a demo on a real fund
No sandbox data, no mockups — the same warehouse every paying account prices from.
3 real, named funds. Set up around how you'd actually use it.
- Blackstone, TPG, CVC, and KKR funds already in our warehouse
- 3 fund lookups, at your own required return
- Reviewed by our team, usually same-day
Prefer to talk it through first? Talk to sales — same team that reviews demo requests.
What a miss actually costs
A Starter seat is a fixed, small number. A pricing miss on one real secondary position usually isn't.
Illustrative, not a guarantee: 5% × $10,000,000 = $500,000 — one plan year costs a fraction of the swing on a single position priced 5 points off. Smaller positions scale the gap down, larger or bigger misses scale it up further. Not investment advice.
Three ways to get a price
See Pricing for the full tier breakdown.
1. Single snapshot
toolType in a fund's current state (age, NAV, paid-in, distributed, unfunded) and price it several ways — Takahashi–Alexander, Equisect Bayesian, and Equisect Cohort — all fitted on warehouse data, each with its own out-of-sample deflation verdict.
- No fund history needed — just today's numbers
- Compare three independent models side by side, four on Team / Institutional and above
- Included on every plan, including Starter
2. Fund or portfolio file
toolUpload a fund's (or a whole portfolio's) own quarterly call/distribution/NAV history as a CSV. Get each fund's Equisect Ledger forecast plus a NAV-weighted portfolio rollup.
- For funds not already in our warehouse
- One file, any number of funds
- Multi-fund uploads also get an Equisect Bootstrap range
- Team / Institutional plan and above
3. Warehouse fund lookup
toolPick a real, active fund already in our warehouse. Get its forecast net cashflow and terminal residual value, and a fair price as a % of NAV at your required return.
- No data entry — we already have the history
- Quantile bands calibrated to their actual held-out coverage
- Included with Enterprise; a separate add-on for Team / Institutional
Research and software, not investment advice.