Why this matters now: The private market valuation disconnect
Private company valuation mark due diligence is the structured audit conducted by limited partners and secondary investors to verify whether unrealized portfolio asset marks reflect true executable exit prices or artificial accounting conventions. For most venture capital and private equity funds, the vast majority of portfolio holdings are classified as Level 3 under the ASC 820 fair value hierarchy, meaning reported Net Asset Values rely almost entirely on unobservable inputs and internal mark-to-model calculations rather than market clearing prices. As the broader private markets liquidity drought stretches fund lifespans toward 15 to 20 years, institutional allocators face an acute valuation disconnect where headline valuations sharply diverge from secondary market realities.
The tension is particularly pronounced in venture-backed artificial intelligence companies. While mega-rounds establish multi-billion-dollar headline valuations with senior liquidation preferences and structured downside terms, the underlying common shares often change hands at steep discounts in secondary markets. Without rigorous verification of revenue quality, inference cost burdens, and liquidation waterfall seniority, allocators risk underwriting fund performance on paper gains that cannot be converted into cash distributions.
- Prolonged liquidity drought: Extended realization timelines leave limited partners holding aging net asset values without cash liquidity, forcing a shift from evaluating TVPI to demanding verifiable DPI milestones.
- Severe secondary discounts: Venture and growth LP portfolios priced at roughly 78% of NAV in the first half of 2025, materially below buyout portfolios at 94%, and single-name late-stage bids can be far weaker still.
- AI valuation volatility: Surging compute capital expenditures and high customer churn challenge the sustainability of triple-digit revenue multiples anchored to prior fundraising rounds.
- Asymmetric governance: Fragmented information rights between lead sponsors and passive co-investors obscure underlying cap table dilution, liquidation seniority, and stale valuation inputs.
The main practical framework: Level 3 assets and mark-to-model
Under accounting frameworks established by FASB ASC 820 and IFRS 13, investments are classified into a three-tier fair value hierarchy. Level 1 assets possess quoted prices in active markets for identical assets, Level 2 assets rely on observable inputs other than Level 1 prices such as market-corroborated data and yield curves, and Level 3 assets rest on unobservable inputs based on management's own assumptions, internal models, revenue multiples, and discount rates. Because private equity and venture capital funds hold predominantly Level 3 securities, reported portfolio marks represent subjective estimates of what hypothetical market participants would pay in an orderly transaction rather than actual transacted liquidity.
Historically, general partners relied heavily on the price of the most recent financing round (the backsolve method) to justify carrying values. However, as private market fundraising cycles stretch beyond 24 to 36 months, managers are forced to transition toward guideline public company (GPC) multiples and discounted cash flow (DCF) models. In practice, secondary pricing provides a hard reality check: average pricing for all LP portfolios reached 90% of NAV in the first half of 2025, up from 89% in 2024, with venture and growth stakes sitting at the weaker end of that market. Those markdowns reflect the cost of illiquidity, minority status, and skepticism toward unobservable modeling assumptions.
| Fair Value Hierarchy Tier | Input Observability | Primary Valuation Methodology | Due Diligence Focus for LPs |
|---|---|---|---|
| Level 1 | Directly observable quoted market prices | Unadjusted exchange quotes (public equities) | Trading volume, market depth, and lockup expiration schedules |
| Level 2 | Indirectly observable market metrics | Corroborated yield curves and comparable transaction multiples | Basis differentials, index correlation, and transaction recency |
| Level 3 | Unobservable inputs and internal assumptions | Mark-to-model, DCF, backsolve, and guideline public company multiples | Multiple selection bias, revenue quality, and calibration rigor |
To establish an objective baseline, institutional LPs must evaluate how managers apply calibration principles recommended by the International Private Equity and Venture Capital (IPEV) valuation guidelines. Calibration requires fund managers to benchmark their ongoing multiple assumptions against the initial entry multiple and subsequent operational progress, preventing arbitrary upward adjustments without corresponding fundamental improvements.
What investors and LPs are really testing
When conducting valuation mark due diligence, sophisticated allocators dissect the fundamental quality of the underlying company rather than accepting top-level enterprise valuations at face value. For venture-backed artificial intelligence startups, this starts with an exhaustive audit of AI valuation premiums to determine whether recurring revenue streams can support high valuation multiples once infrastructure costs and compute expenses are accounted for.
A primary area of scrutiny is gross margin sustainability. Traditional software businesses have long run structurally higher gross margins than AI-native ones: one CFO-led analysis puts SaaS gross margins in the 70% to 85% band against 40% to 60% for AI companies, because every query carries recurring compute, inference, and model licensing costs. Allocators must verify whether reported annual recurring revenue reflects durable enterprise software contracts or ephemeral experimentation budgets that will churn when pilot projects conclude.
Equally critical is the preference stack and liquidation waterfall. In later-stage funding rounds, new capital often enters via senior preferred shares with cumulative dividends, redemption rights, or liquidation multiples. In an exit scenario below headline valuation marks, senior preferred holders absorb the entire payout, leaving common equity and junior preferred classes completely wiped out. Allocators must evaluate where their fund's specific share class sits within the capitalization structure to measure downside protection and understand the realistic path toward DPI vs TVPI performance.
- Unit economics and inference gross margins: Deconstruct the true cost of goods sold, isolating compute hosting fees and foundational model API dependencies from pure software margins.
- Contract durability and retention: Analyze net revenue retention (NRR) and logo churn to verify whether customer expansion is driven by durable software workflows or short-term token consumption.
- Liquidation waterfall modeling: Stress-test exit distributions across various realization thresholds to quantify the haircut junior equity holders suffer under senior preference stacks.
- Runway and capital efficiency: Measure cash burn rates against cash-on-hand to anticipate whether imminent down-rounds or structured recaps will dilute existing fund marks.
What companies and funds are expected to show
To satisfy institutional limited partners, general partners and portfolio companies must maintain rigorous valuation governance supported by clear documentary evidence. The days of relying solely on informal valuation committee notes or rubber-stamping previous primary rounds are over. Allocators expect complete transparency regarding the inputs, assumptions, and peer peer-group selections used to construct quarterly marks.
Fund managers should provide comprehensive valuation committee packages that detail the specific methodology applied to each Level 3 asset. If a guideline public company method is used, the GP must articulate why the selected public peers are genuinely comparable in growth rate, operating margins, and market positioning. Furthermore, managers must document any applied illiquidity or minority discounts, particularly when underlying market transactions signal softening demand.
- Valuation committee documentation: Comprehensive meeting minutes, valuation policies, and written rationale for any departures from calibrated market multiples.
- Standardized performance reporting: Transparent quarterly disclosures breaking down gross and net asset values, unrealized multiples (RVPI), realized gains (DPI), and total value to paid-in capital (TVPI).
- Observable transaction corroboration: Analysis of recent secondary market transactions, secondary tender offers, or third-party primary equity raises within the sector.
- Third-party valuation opinions: Independent valuation reports from accredited third-party appraisal firms for all Level 3 holdings exceeding materiality thresholds.
A red-flag table for AI startup valuations
Detecting valuation anomalies early prevents institutional investors from making follow-on commitments or secondary purchases based on illusory asset values. When evaluating AI startup marks, specific quantitative disconnects and governance gaps signal elevated valuation risk.
| Valuation Signal | Observed Discrepancy | Underlying Risk | Verification Diligence Test |
|---|---|---|---|
| Stale Round Anchoring | Carrying value still held at a boom-era peak primary round despite sustained contraction in listed peer multiples | Overstated NAV and delayed down-round recognition | Recalibrate EV against current median public SaaS and AI revenue multiples |
| Severe Secondary Divergence | Secondary bids or closed transactions clearing far below reported NAV, well outside prevailing market pricing for venture interests | Illiquidity shock and imminent common share impairment | Cross-reference cap table preference stack against secondary transacted clearing price |
| Multi-Manager Mark Disparity | Identical private asset marked at divergent multiples across different fund managers | Subjective model manipulation or non-standardized valuation policy | Benchmark marks across institutional holdings using standardized cap table waterfalls |
| Compute Capitalization | R&D compute expenses capitalized onto balance sheet to inflate EBITDA | Artificially elevated operating margins masking unsustainable cash burn | Reconcile GAAP cash flow from operations against reported EBITDA adjustments |
| Customer Concentration | A large share of ARR concentrated in non-arms-length affiliate or seed investors | Inflated revenue quality masking artificial top-line growth | Audit customer contracts for reciprocal vendor credits and non-recurring pilot clauses |
A data-room and evidence checklist
To conduct an uncompromised valuation mark audit, limited partners and secondary diligence teams must request a specific set of primary source documents. Relying exclusively on general partner summary slide decks creates information blind spots that obscure structural liabilities.
- Detailed capitalization table: Complete post-round cap table showing share classes, conversion ratios, participation rights, cumulative dividend accruals, and liquidation preferences.
- Audited annual financial statements: Financial disclosures with detailed Level 3 fair value footnote disclosures and external auditor valuation testing notes.
- Valuation policy and governance charter: The written policy defining valuation frequency, committee composition, third-party appraisal triggers, and override protocols.
- Historical mark adjustment log: A chronological audit trail tracking every upward and downward adjustment of the asset carrying value since initial fund entry.
- Secondary transaction and transfer register: An exhaustive log of all approved, pending, and rejected secondary share transfers, tender offers, and company buybacks over the prior 24 months.
- Revenue bridge and cohort schedules: Granular contract-level recurring revenue schedules separating new customer additions, expansion, contraction, and churn.
How to use this in your next diligence workflow
Institutional limited partners and secondary underwriters should embed valuation mark verification directly into their recurring underwriting and portfolio monitoring cycles. By combining structured document ingestion with automated waterfall modeling, investment teams can quickly cut through subjective mark-to-model assumptions to uncover true economic exposure.
This is where modern deal intelligence platforms transform manual due diligence. Plausity provides institutional deal teams with an advanced AI-Analysis Engine that automatically ingests virtual data rooms, cross-references hundreds of cap table agreements and valuation memos, and surfaces hidden liquidation liabilities within minutes.
Using integrated Findings & Risk Intelligence, investment professionals can instantly detect discrepancies between headline marks, preference stack seniority, and underlying revenue metrics. Rather than spending days manually recalculating waterfalls in disconnected spreadsheets, diligence teams generate clear, IC-ready findings that empower investment committees to challenge stale marks, renegotiate secondary pricing, and protect institutional capital with complete confidence.
How Plausity supports the workflow
Plausity is an AI-native due diligence and deal intelligence platform. For LPs and secondary buyers evaluating private-company valuation marks, Plausity helps convert capital-account statements, valuation policy documents, secondary bid/ask data, and preference-stack waterfalls into a structured, source-backed evidence base that stays traceable through the valuation committee.
LP teams can use AI-powered diligence analysis to cross-reference last-round marks against executable-price evidence, revenue quality, and liquidation waterfalls, then organize valuation risks with findings and risk intelligence before signing off on quarterly NAVs. Plausity is a document-and-workflow layer, not a substitute for professional judgement: it does not independently provide legal, financial, tax, commercial, or technical advice, and it does not guarantee investor decisions, valuations, or diligence outcomes.
