Choosing a valuation method in a contested buyout. A disciplined frame for counsel.

The method is often where the dispute begins

In a contested private-company buyout, the headline disagreement is usually a price. The underlying disagreement is more often a choice of method: what the interest is being valued for, which date controls, what information belongs in the record, and which limitations on ownership should affect the result. A different answer at any of those points can move the conclusion materially without either side having acted irrationally.

That is why a valuation method should not be selected by asking which formula produces the most favourable number. The relevant question is whether the method fits the legal purpose of the assignment, the rights attached to the interest, the information available at the valuation date, and the economic reality the parties are asking the process to measure.

The analysis below is not a universal answer or a substitute for the governing agreement, statute, order, or expert judgment. It is a disciplined frame for counsel and mediators who need to move from competing positions to a price that can be explained, tested, and accepted by the parties.

Start with the standard of value, not the spreadsheet

Standard of value and valuation method are related but distinct. The standard defines the premise of the conclusion; the method supplies the mechanics for estimating it. In a dissenting-shareholder proceeding, for example, the governing law may direct a fair-value analysis with specific treatment of discounts. In a negotiated secondary, the parties may instead be testing fair market value, investment value to a particular buyer, or the price at which a willing buyer and willing seller would transact under stated assumptions. A buyout under a shareholder agreement may carry a contractual definition that controls the exercise even when an expert would ordinarily select a different frame.

The engagement should therefore answer several questions before the expert chooses a model:

Is the assignment asking for fair value, fair market value, investment value, or another defined standard?
Is the interest being valued as a stand-alone minority holding, as part of a control block, or under a contractual purchase formula?
Does the governing instrument require or prohibit minority, control, or marketability adjustments?
Is the conclusion intended to inform a court, a fiduciary decision, a mediation, a buy-sell process, or a negotiated transaction?

Those questions do not dictate a single valuation method. They establish the perimeter within which the methods can be compared. A discounted cash flow analysis, a guideline public-company method, a precedent-transaction method, an option-pricing approach, or an asset-based method may each be appropriate in a different assignment. The method is defensible when its premises match the question being asked.

In a mediation, this separation is particularly useful. Counsel can disagree about the standard without disguising that disagreement as an argument about a growth rate or a multiple. The parties can identify which issues are legal, which are factual, and which are expert judgment. That division makes later settlement movement more intelligible.

Fix the valuation date and the information cutoff

The valuation date is not a timestamp added to the cover page after the model is complete. It determines which facts, forecasts, financing conditions, and market observations belong in the analysis. A valuation that uses a later financing round, a post-date acquisition discussion, or a subsequent operating result may be more informed in one sense while answering a different question in another.

Counsel should distinguish three concepts that are often compressed into one:

The date on which the interest is legally or contractually valued.
The information that was known or reasonably knowable as of that date.
The later information used only to test whether an earlier assumption was reasonable, rather than to rewrite the earlier conclusion.

The distinction matters in a private company because information arrives unevenly. One party may point to a later customer win as confirmation of a forecast; another may argue that the win was not knowable at the valuation date. A later financing can establish that capital was available, but it may also reflect changed terms, a new investor class, or a different risk environment. A subsequent sale can illuminate the market, but it does not automatically prove what the interest was worth before the sale process began.

The clean practice is to maintain an information cutoff and label later material separately. Discovery should identify when forecasts were prepared, when board materials were circulated, when financing terms were negotiated, and when a liquidity event became more than speculative. Expert reports should state which post-date facts are used as a reasonableness check and which, if any, alter the model itself.

That record also prevents a settlement from becoming a debate over hindsight. Parties can negotiate a range knowing which assumptions are historical inputs and which are later evidence. The final price then reflects an intentional resolution of uncertainty rather than an unexamined blend of dates.

Treat minority, control, and marketability adjustments as facts

Discounts and premiums are not interchangeable labels for negotiation leverage. They are attempts to describe specific economic attributes of the interest. A minority discount may address the absence of unilateral control over distributions, strategy, or a sale. A control premium may address the value of decision rights. A discount for lack of marketability may address transfer restrictions, the time and cost of finding a buyer, information asymmetry, and the probability of a liquidity event.

The right question is not whether a discount is large or small in the abstract. It is what feature of this interest the adjustment measures, and whether that feature is already captured somewhere else in the analysis.

Relevant facts can include:

voting rights, board appointment rights, veto rights, and protective provisions;
transfer restrictions, rights of first refusal, drag-along and tag-along provisions;
the company's distribution history and realistic path to liquidity;
the size of the interest, its relationship to other holders, and the ability to form a coalition;
the quality and timeliness of information available to the holder; and
the likely holding period, transaction costs, and available secondary markets.

Interaction is as important as selection. A control analysis and a minority discount may be answering opposite questions about the same rights. A marketability discount may overlap with a long holding-period assumption in an option model or with a low multiple selected because the company is difficult to exit. Applying every adjustment that appears plausible can create double counting. Omitting an adjustment because another method uses a different mechanism can be equally misleading.

Experts should show the bridge from the selected standard of value to each adjustment, identify overlaps, and provide sensitivity rather than burying the judgment in a single percentage. The conclusion can still be one number. The path to it should show which facts moved the number and which uncertainties remain.

Discovery should test the model's inputs

Discovery in a contested buyout should do more than collect documents for a chronology. It should test whether the inputs used by each method are complete, comparable, and tied to the interest being valued. A focused request often includes:

the cap table, capitalization history, option and warrant schedules, and amendments to governing agreements;
transfer restrictions, repurchase rights, buy-sell provisions, and prior offers or indications of interest;
historical financial statements, budgets, forecasts, board materials, and variance analyses;
financing history, term sheets, preferred rights, liquidation preferences, and changes in capital structure;
customer concentration, recurring revenue, contracts, pipeline evidence, and operational dependencies;
the company's liquidity path, including distributions, tenders, redemptions, strategic discussions, and likely exit timing; and
the provenance, date, selection criteria, and adjustments for every proposed comparable.

The provenance of a comparable matters as much as its headline multiple. Counsel should be able to ask why a transaction belongs in the set, what rights were sold, whether the transaction was distressed or strategic, how recent it was, and whether the market conditions were materially different. A public-company multiple without an explanation of size, growth, margin, leverage, or control rights is not a comparator; it is an unsupported anchor.

Discovery should also expose the decision points hidden in a forecast. Was a projected margin approved by the board? Was a customer renewal assumed before a signed commitment existed? Did the financing plan depend on a term sheet that later expired? These questions do not automatically discredit a forecast. They determine how much weight the forecast should carry and whether a sensitivity case belongs beside the base case.

Compare expert reports by reasoning, not headline number

Two expert reports can be polished, internally consistent, and far apart. The proper comparison is not which report sounds more certain. It is where the reports use different facts, premises, or treatments of uncertainty.

Counsel can structure the comparison around four tests:

Assumptions. What standard of value, valuation date, ownership premise, growth path, capital structure, and liquidity horizon does each report use?
Evidence. Which assumptions are supported by contemporaneous records, and which are selected as expert judgment? Are comparable transactions verified and adjusted for rights and conditions?
Sensitivity. How much does the conclusion move when the key inputs change? Does a narrow conclusion depend on a wide range of hidden judgments?
Reconciliation. Where methods produce different results, is the difference explained by the economics of the interest or merely averaged away?

An expert report is more useful when it identifies the inputs that deserve a mediation discussion. A reconciliation schedule can show that the parties agree on enterprise value but diverge on capital structure, or agree on the valuation date but diverge on the probability of a liquidity event. That is a narrower and more tractable dispute than two unexplained final numbers.

The AI-assisted review in a Velmark room should be understood in that limited way. It can normalize submissions, surface provenance gaps, compare stated assumptions, and flag where a position is sensitive to an input. It does not choose the governing legal standard, replace counsel's judgment, or turn a disputed premise into an objective fact. The purpose is to make the reasoning visible enough for the parties to evaluate it.

Settlement patterns that avoid unsupported midpoint math

Settlement does not require the parties to agree that one expert is wholly right and the other wholly wrong. Several patterns can move the case while preserving the distinction between evidence and compromise.

One pattern is an assumption ladder: the parties identify the few inputs that drive most of the spread, agree on factual bounds, and price the cases across those bounds. A second is a bracketed resolution, where each side commits to a range and the mediator tests movement against the strength of the supporting evidence. A third is a staged or contingent payment, used carefully where a defined future event is measurable and the cost of continued uncertainty is material. A fourth is a structured buyout that separates the value of the interest from timing, payment security, or a later reconciliation mechanism.

None of these patterns makes a split-the-difference result inherently wrong. A midpoint can be rational when it reflects a deliberate allocation of model risk and litigation risk. It is unsupported when it is chosen only because two numbers are far apart. The settlement record should say what the midpoint resolves: a valuation-date uncertainty, a discount range, a disputed forecast, the cost of delay, or the parties' shared preference for finality.

How Velmark carries the dispute through the case lifecycle

The method debate should not disappear when the parties enter a price-discovery process. It should be carried forward as a documented set of assumptions and confidential positions. Velmark's five-stage flow is designed for that handoff.

1. Case opened

At case opened, the neutral coordinator records the parties, the contested asset, the unit of value, the round schedule, the minimum bid increment, the refundable deposit, and the comparator universe. This is where counsel can identify the governing standard, valuation date, information cutoff, and any agreed boundaries for the exercise. Later amendments append a documented reason to the case file; they do not silently replace the foundation.

2. Comparables submission and AI surfacing

During comparables submission and AI surfacing, each party submits comparables confidentially. The platform normalizes them by asset class, vintage, and discount class, attaches provenance, and surfaces an AI-assisted view shared with the parties. The dataset is discussed without exposing offer values. This stage is suited to testing whether the proposed comparables actually measure the same economic rights and whether a discount or premium is being counted twice.

3. Sealed bid rounds with masking and transitions

In sealed bid rounds with masking and transitions, parties submit positions into sealed compartments. No plaintext offer is shown to any party until the configured reveal window. Advance, hold, and conclude transitions follow thresholds recorded in the case policy, while rewrites and counters are appended to the round-by-round audit log. The sealed structure separates a party's considered valuation from the pressure to follow the other side's number.

4. Winning bid revealed and audit-ready summary

At winning bid revealed and audit-ready summary, the sealed repositories unmask at the close of the final round and the AI reconciliation is appended to the case file. The record includes the offers, counters, comparables and their provenance, the weighting assigned by the AI neutral, and the written rationale for the resulting price. Counsel can distinguish the winning position from the assumptions that supported it without rerunning a model or relying on an unexplained headline.

5. PDF export and case conclusion

Finally, PDF export and case conclusion turns the record into a counsel-grade document. The PDF preserves the case configuration, submissions, round history, revealed result, reconciliation, and documented amendments. The case is then marked concluded and the bid form is locked from further entry, subject only to a documented coordinator-role reopening. That final state matters in a contested buyout: the price is accompanied by a reconstructible record of how the parties reached it.

A defensible method is one the parties can explain

There is no valuation method that is universally correct for every contested buyout. The defensible choice is the one that answers the assignment's actual question, respects the valuation date, connects adjustments to the rights of the interest, tests the provenance of the evidence, and makes uncertainty visible rather than hiding it in a final percentage.

Velmark does not adjudicate the legal standard or declare one expert's method universally correct. It gives counsel and mediators a controlled procedure for submitting the competing analysis, testing the comparables, keeping bids sealed through the configured rounds, and producing an audit-ready final price with the reasoning attached.