Private markets valuation where agents compute three methods, argue against every mark and cite each source, and your people prepare, check and approve
Every quarterly mark backed by three methods and its evidence, signed off by two people.
See one case, screen by screen ↓Private markets valuation is the quarterly work of setting a fair value mark for each unlisted holding — buyout and growth companies, infrastructure, real estate — usually a Level 3 fair value measurement under ASC 820, from the company's accounts, peer multiples, a DCF and calibration to the entry price or latest round. Private Markets Valuation adds seven agents that read company packs in their own languages, compute the three methods independently, argue against the result and seal the evidence, while the analyst prepares, the Head of Valuation checks and the valuation committee approves.
AI in asset management: a buyer's guide to agentic AI for investment managers →
Updated
Private markets valuation is a quarterly grind: hundreds of unlisted holdings — buyout and growth companies, infrastructure assets, real estate — each marked to fair value from a company pack that arrives as a PDF, often in Italian, German, Japanese or Portuguese. Analysts refresh peer sets by hand from a market data terminal, roll DCFs in personal workbooks, skip calibration to the entry price when "nothing happened", and assemble the valuation committee pack the night before.
The risk sits where examiners now look. The SEC's FY2026 examination priorities name the methods and controls behind the fair valuation of illiquid assets, and the December 2025 IPEV Valuation Guidelines expect calibration at every measurement date. A stale peer set, a plan that keeps missing and is never haircut, a deal team leaning on its own mark, or evidence rebuilt by hand for the auditor can each turn a Level 3 mark into a finding.
Estimated split for one analyst covering about 40 positions in a quarter, by hand and with the solution.
Seven agents read the packs, compute three methods independently and argue against the result; an analyst prepares, the Head of Valuation checks, the committee approves.
Quillfield Capital marks 304 positions in 38 funds for the quarter ended Sep 30, 2026. Ferrante Imballaggi S.p.A., an Italian packaging company held 89.5% by Quillfield European Buyout Fund IV, is first on the analyst's list. Here is the private equity valuation process for that one mark — made, challenged and approved — screen by screen, in the working solution.
Nadia covers 41 European buyout and infrastructure positions. Her home shows four numbers: 304 positions marked, $41.62B, −0.9% vs June; 212 within tolerance; 23 challenged with 27 challenges still open; 8 awaiting company data, rolled forward under policy §7.1. First in Needs you: Ferrante Imballaggi, −6.0%, to committee, two open challenges — plan missed, peer outlier — and 2h 18m to the pack cut-off.
What the agents did overnight: “Challenger agent — Raised 29 challenges on 23 positions; 2 cleared as pass or info.”
The Company Data agent read 296 company packs — 4,812 pages in Italian, German, French, Spanish, Japanese, Portuguese, Swedish and more — and pinned 3,911 figures to their pages; 8 packs had not arrived, so it sent the requests. The Comparables agent refreshed 211 peer sets, the DCF agent rolled 251 DCFs from company plans, the Calibration agent calibrated 268 positions, and weights came from Valuation Policy v7.2. Out came 212 within tolerance, 61 for review, 23 challenged and 8 awaiting data.
The last step of the run: “People decide — Analysts prepare · Head of Valuation checks · committee approves.”
Ferrante's mark opens as three dials: comparables EUR 212.0M at a 30% weight, DCF EUR 198.0M at 45%, calibration EUR 205.1M at 25% — the weights of policy §4.2 for a mature buyout. The methods span EUR 198.0M–212.0M and propose EUR 204.0M, −6.0% on June's EUR 217.0M; the reporting value is $237.4M at EUR/USD 1.164. Beside the dials, each method agent states its result in the conversation, with citations.
Calibration agent: “At entry Quillfield paid 10.95x against a peer median of 11.2x. Holding that 2.2% discount, today's multiple is 8.9x: EUR 205.1M.”
Every cited number opens its source. The Relazione trimestrale al 30 settembre 2026, received through the company portal on Oct 9, sits in Italian beside its English reading, with the cited lines highlighted: LTM revenue EUR 286.5M, LTM adjusted EBITDA EUR 38.4M (June: EUR 37.9M), adjusted EBITDA margin 13.4%, and EUR 1.2M of reorganization costs taken out as non-recurring.
The Company Data agent never changes a figure the company reported; below its confidence threshold it hands the item to the analyst.
Seven listed packaging peers, median EV/LTM EBITDA 9.1x against 9.9x in June. Each shows why it is in the set — Corvana Packaging: paper-based food trays, same end markets, larger scale — its multiple now and last quarter, and an Exclude button. Two companies were screened out by the agent: Ardmore Plastics and Vesuvio Cartiere. Kestrel Paperboard, at 11.3x, is flagged as an outlier.
Comparables agent: “Kestrel Paperboard trades on press reports of an approach. Policy §3.4 excludes a peer only on a confirmed event, so I kept it and flagged it.”
The DCF rolls the board-approved plan: free cash flow for 2027–2031 from the Piano industriale, a 2.85% EUR 10-year risk-free rate and 0.92 unlevered beta from the market data vendor, a 5.50% equity risk premium from policy §4.3. WACC 10.25% and terminal growth 2.00% give EUR 198.0M; ±0.5% on WACC spans EUR 180.5M–217.8M. Moving a slider is a sensitivity; applying it is an override that asks for a reason code.
The DCF agent never invents a forecast: it uses the company plan or the analyst's override.
“I disagree on one point and flag one more.” The plan has missed EBITDA three quarters running (Q1 −1.8%, Q2 −3.4%, Q3 −2.1%); a 2% haircut on forecast cash flow would take the DCF to EUR 192.0M and the mark to EUR 201.3M. And Kestrel Paperboard has traded up 14% on takeover talk since Sep 12; excluding it lowers the mark by about EUR 2.4M. Nadia keeps the forecast with reason code KP-02, “Later information covers it (cited)”.
“A challenge cannot be dismissed without a reason code. Unanswered challenges escalate to the Head of Valuation at the pack cut-off, and the committee sees every answer.”
Luca Bernardi, the deal partner, added information without touching the mark: a price revision with the largest customer, Gruppo Alimentare Sorrentino. The DCF agent answers: +4.5% from Oct 1 on 19% of revenue, about EUR 2.5M of EBITDA a year, more than the Q3 shortfall against plan. Nadia keeps Kestrel too, under KP-03, because the approach is press talk and policy §3.4 keeps the peer. The Challenger accepts both answers into the committee pack.
DCF agent: “It supports keeping the forecast; the decision is yours.”
Nadia prepares the mark at EUR 204.0M with PR-01, “Methods agree within policy; challenges answered”. Tomás Reyes, Head of Portfolio Valuation, checks it with CK-01, “Reasoning and sources checked”. A move over 5% goes to the committee, where Helen Ward, CFO, Private Markets and committee chair, approves it with AP-01, “Approved: methods converge; dissent answered”, with 4 of 5 voting members present.
On the Prepare screen: “You cannot check or approve it yourself.”
The Pack agent has grouped 304 marks into 26 for discussion — material moves, open challenges, overrides — and 278 on the consent agenda, approved in one resolution once checked. Ferrante now shows Approved · AP-01. Then Helen signs the quarter: the approved marks are sealed and sent to fund accounting for the Sep 30 NAV, anything undecided carries to the adjourned session on Oct 23, and the Pack agent drafts the minutes.
“Quorum: 4 of 5 voting present (3 needed).”
The same quarter, seen by five of the people who carry it — what valuation season looked like, and what it looks like now.
Seven agents read the packs, compute three methods independently and argue against the result; an analyst prepares, the Head of Valuation checks, the committee approves.
Reads each portfolio company's quarterly pack, accounts, plan and compliance certificate in the original language; extracts LTM revenue, EBITDA and adjustments, net debt and KPIs; pins each figure to its page; requests late packs from the company CFO in the company's language.
Builds and maintains each peer set, pulls multiples from the market data vendor, explains why every peer is in the set and flags outliers.
Rolls the DCF from the company's plan, rebuilds WACC, runs sensitivities and tracks the plan against actuals.
Calibrates to the entry price or the most recent transaction and rolls the calibrated multiple forward; reads share-class rights for recent rounds.
Argues against every converged mark with last quarter, the peer index, co-investor marks, independent adviser ranges, plan misses and the backtesting of exits.
Drafts the valuation committee pack by exception, the minutes from the decisions, and the quarter reports.
Seals every input, statement and decision with its source and a hash, and builds the auditor and regulator packs.
Maker, checker, approval, evidence. The agents propose; a named person decides.
Ask about any mark, method, fund or source, or say what to try. Every answer cites its page, and nothing changes without your approval.
Why is Ferrante Imballaggi down 6%?
From EUR 217.0M to EUR 204.0M. Peer multiples fell 0.8x to 9.1x: −EUR 19.8M. LTM EBITDA rose to EUR 38.4M: +EUR 3.0M. Net debt fell to EUR 112.6M: +EUR 3.8M.
What if we drop Kestrel?
Without Kestrel the peer median is 8.95x and the mark about EUR 2.4M lower. Policy §3.4 excludes a peer only on a confirmed event; the approach is press talk so far. I have not changed the set: it is your call on the peer screen.
Does the price increase support the forecast?
Luca Bernardi added the price revision with Gruppo Alimentare Sorrentino: +4.5% from Oct 1, on 19% of revenue. That is worth about EUR 2.5M of EBITDA a year, more than the Q3 shortfall against plan. It supports keeping the forecast; the decision is yours.
Show marks above the adviser's range
One: Brightwater Dental Partners at $655.0M, 1.1% above the independent adviser's top of $648.0M. The Challenger raised it; Marcus Bell has recorded an override on terminal growth.
13 screens from the working solution, on its sample data. Pick one to see it large.
Method weights, tolerances and committee materiality are settings in the valuation policy, versioned, and changed only with a second person's approval.
| Setting | Default | Choose from |
|---|---|---|
| Method weights, buyout | Market 30% · income 45% · calibration 25% | Set per strategy: growth 50 · 15 · 35, infrastructure 10 · 80 · 10, real estate 20 · 70 · 10, secondaries 20 · 80 · 0 (secondary bids · adjusted NAV · calibration) |
| Within tolerance | Move under 3% and all methods within ±7% of the mark | Proposed by the Head of Valuation, approved by the committee chair |
| Committee materiality | Move of 5% or more, or $25 million or more | Plus any override and any open challenge |
| Peer exclusion | Only on a confirmed event | Announced offer, delisting, restatement |
| Calibrated multiple | Peer median × entry multiple ÷ peer median at entry | Rounded to 0.1x |
| Late company data | After 15 days | Request in the company's language, roll forward with market movement, flag to the committee |
| Confidence to proceed | Company Data agent 90%; method agents 85% | Below the threshold an agent hands the item to a person |
| Challenge answer | 24 hours after it is raised | Unanswered at cut-off, it escalates to the Head of Valuation and the committee chair |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample quarter · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (an industry valuation survey of 80 US managers — about half finish within 20 business days after quarter end, about 60% use third-party valuation firms; SEC staff statement on the valuation of private assets, Sep 28, 2026 — 54% of interval fund assets were Level 3 at year-end 2025). Context: SEC Division of Examinations, FY2026 Examination Priorities (Nov 17, 2025); IPEV Valuation Guidelines, December 2025 edition. People, firms, funds and companies named on this page are fictional — characters and sample data in the working solution.
Private markets valuation is the quarterly work of setting a fair value mark for each unlisted holding — buyout and growth companies, infrastructure, real estate — usually a Level 3 fair value measurement under ASC 820, from the company's accounts, peer multiples, a DCF and calibration to the entry price or latest round. Private Markets Valuation adds seven agents that read company packs in their own languages, compute the three methods independently, argue against the result and seal the evidence, while the analyst prepares, the Head of Valuation checks and the valuation committee approves.
Three on every position, computed independently from cited sources: comparables (the peer median EV/LTM EBITDA), a DCF rolled from the company's own plan, and calibration to the entry price or the latest round. The valuation policy weights them by strategy — for a mature buyout, market 30%, income 45%, calibration 25% — and the mark shows the range they span.
The Calibration agent holds the entry discount to peers and rolls the calibrated multiple forward each quarter; on the sample quarter it calibrated 268 of 304 positions to entry or the latest round. It never treats a recent round as fair value without calibration: for a growth company it reads the share-class rights and values the fund's class, not the headline round price.
By exception. The Pack agent puts material moves, overrides and open challenges up for discussion — 26 of 304 in the sample quarter — and the rest on a consent agenda. The chair approves each mark with a reason code and quorum, then signs the quarter, and the approved marks go to fund accounting for the NAV.
Yes. The Company Data agent reads each pack in its original language and shows it beside an English reading, with every figure used in a mark pinned to its page. When a pack is late it drafts the request to the company CFO in the company's own language.
No. The agents compute, argue and draft; they never prepare, check or approve a mark. The analyst prepares with a reason code, the Head of Portfolio Valuation checks, the committee approves, and no one can check or approve a mark they prepared. Deal teams can add information but never edit a mark.
The auditor picks a read-only sample and sees each mark's methods, original-language sources, decisions with reason codes, challenges and answers, and its seal, and can ask the agents. The compliance team can export a regulator pack of every mark and approval from the hash-chained audit trail, watermarked and logged. The SEC's FY2026 examination priorities name the valuation of illiquid assets.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your valuation policy, fund list and a past quarter of company packs and approved marks — and it goes live once every quality gate has passed. Setup connects your systems, loads the policy, maps funds and positions, and runs the agents in parallel on last quarter before go-live.
Every KPI cited to its page, every covenant tested the day the pack lands.
1-min demoPrivate Equity Deal ScreeningEvery CIM tested against the fund’s own limits in minutes, with every page cited.
1-min demoVirtual Data Room Due DiligenceEvery document read and every red flag cited to its page, before you sign.
1-min demoWe'll run Private Markets Valuation on a past quarter of your own company packs, peer sets and approved marks.