Definition
What is a contested valuation?
A contested valuation is a situation in which the parties to a transaction — shareholders selling out, beneficiaries receiving a bequest, spouses dividing equity, or a company running a tender — hold material, well-reasoned disagreements about the fair value of the same asset. It is not the same thing as a disagreement about a sale. In a contested valuation, the number itself is the matter. The deal would be simple if everyone agreed on the price.
Every party has a reason for the number they hold. Most of those reasons are reasonable on their own terms: a stale 409A, a current market reset, a control premium applied correctly, a minority discount applied correctly, a defensible projection, a different defensible peer set. The issue is not that one side is wrong. It is that both sides have constructed their own coherent story, and the stories do not meet.
Where contested valuations come from
The most common scenarios.
Four shapes account for the majority of contested valuations we see. None of them are exotic. They are the ordinary consequences of private markets, opaque cap tables, and human relationships that do not always end neatly.
Co-founders splitting a cap table
Two co-founders, two visions of the company. One anchors on the last 409A, the other on a forward-looking Series A multiple. The cap table is no longer the asset; the valuation is.
Estates, trusts, and probate
A private company sits in a trust. Two appraisal firms, two letters, two numbers. The IRS has its date-of-death value; the marketplace reset is six months old or six months ahead. Beneficiaries and executors need a single defensible figure to administer the trust cleanly.
Divorce and equitable distribution
Restricted stock and carried interest are on the schedule. Spouses hire separate valuation experts, and the resulting reports diverge in ways that read like competing briefs. Equity division is hard enough without the number itself being a fight.
Secondaries and tender offers
A company launches a structured tender or a tender-style secondary. Existing investors hold at last-round marks; the company wants to reset to a current private-market multiple. Two reasonable numbers, no shared anchor.
Where the numbers fork
Why independent appraisals diverge.
Two reputable firms, working from the same data room, will routinely produce numbers 30–60% apart on the same private-share lot. The disagreement looks irrational from outside and feels inevitable from inside. Four mechanisms do most of the work:
Methodology
A DCF at one team’s WACC and a revenue multiple at another team’s peer set produce numbers that are both defensible in isolation — and irreconcilable in the joint report.
Valuation date
A stale snapshot pulled from a prior 409A is not the same animal as a refresh that incorporates the last quarter’s bookings or the last round’s reset. Each side has a good reason for picking the date it picks.
Control premium vs minority discount
One side applies a control premium because the lot is being transacted at scale. The other applies a minority discount because the lot is non-controlling. Same share class, opposite direction.
Illiquidity and pick of comparables
Discounts for illiquidity, key-person risk, and customer concentration are applied selectively. The peer set each side picks is the one that supports its number. Both sets look reasonable; both push the answer a different way.
None of these are mistakes. They are choices that look defensible when read in isolation. The fix is not to choose a winning expert; it is to surface every choice, in one place, in front of both sides, with reasoning attached.
Paths to a number
Litigation, arbitration, or mediation.
Three conventional paths to a usable number. Each has a place; none is universally right. The right choice depends on how much time, money, and adversarial oxygen the matter can absorb.
| Axis | Litigation | Arbitration | Mediation |
|---|---|---|---|
| Time to a number | 12–36 months | 4–9 months | 2–4 weeks |
| Cost structure | Counsel + experts + court time; open-ended | Fixed arbitrator fee + counsel + one expert | Set-up fee + per-bidder; capped by room size |
| Transparency of reasoning | Findings of fact and conclusions of law; sealed exhibits; rationale often opaque to non-counsel | Award with reasoning, usually private; outside counsel redacts | Every bid and counter timestamped; AI-neutral rationale in plain language; exportable record |
| Adversarial posture | Maximum — discovery and motions are weapons | Higher than mediation; counsel still positional | Minimum — parties engage on the basis, not the outcome |
| Defensibility post-hoc | Court record; jury or judge ruled; persuasive but slow | Final by contract; appeal narrow | Settlement by agreement; record attachable to a fairness opinion |
The Velmark approach
The case for a mediated midpoint.
A mediated midpoint is not a compromise. It is a number produced by a structured auction in which every bidder posts their amount AND the reasoning behind it, in a sealed compartment, under agreed rules. The midpoint that lands is the median of the highest-conviction positions in the room, after an AI neutral has scored every bid on comparability, recency, conviction, and risk.
The structure is what produces defensibility. Sealed compartments prevent anchoring on a single price before the full field is visible. The basis attached to every bid prevents the parties from retreating to “low” or “high” without a story. The neutral scorer ranks reasoning, not just prices, so a bid with a better peer set and a tighter discount can outrank a higher number with shakier methodology. The median rule concentrates weight on the strongest positions, not the loudest outliers.
Both sides agree to the process because both sides write the rules. Both sides can read the record end to end. The number is not handed down; it is the median of the best-reasoned positions both sides submitted to. That is the difference between a price a court can sign off on and a price neither side will sign at all.
In a Velmark room, the audit trail is timestamped, sealed, and export-ready. Every offer, every counter, every comparable, and the AI neutral’s written rationale attach cleanly to a fairness opinion or a divorce filing. The room runs in days, not quarters. Bidders are counterparty-checked before they are invited. The whole exercise is the kind of record a judge, an arbitrator, or a beneficiary’s counsel can read in an afternoon.
The number that comes out is not theoretical. It is mediated.