What the DLOM is, and what it is not
The discount for lack of marketability — DLOM — is a percentage reduction applied to the value of a privately held interest to reflect the fact that the holder cannot sell it quickly, easily, or on a public exchange. It is not a penalty. It is not a negotiating tactic. It is a structural feature of private ownership: the asset is real, but the market for it is thin, slow, and often closed entirely to minority holders.
Courts have recognised DLOM for decades. The IRS accepts it as a legitimate input under Revenue Ruling 59-60. Valuation firms apply it routinely in estate filings, divorce proceedings, and dissenting-shareholder actions. The concept is uncontested. What is contested is the number.
Appraisers using defensible, peer-reviewed methodologies on the same interest will produce DLOM estimates ranging from 15% to 45%. That spread is not an error. It is a direct consequence of the underlying uncertainty: private shares do not trade, so any DLOM estimate is a model output, not a market observation. Different models, different inputs, different answers.
Why the range is so wide
Three mechanisms drive most of the divergence in DLOM estimates.
The choice of method. The two most common approaches are the restricted-stock studies and the put-option models (Finnerty, Chaffe, Longstaff). Restricted-stock studies look at the discount at which registered shares trade relative to public equivalents during a lock-up period. Put-option models price the cost of a hypothetical put that would allow the holder to exit at will. Both are legitimate. Both produce systematically different numbers, and neither is the "right" method in the abstract — the right method depends on the facts of the holding.
The assumed holding period. Every put-option model requires an estimate of how long the holder would need to wait for a liquidity event: an IPO, an acquisition, a secondary tender, or a redemption. One appraiser assumes three years based on the company's stated timeline. Another assumes six years based on the sector's historical M&A data. The DLOM output moves materially with that input. Both assumptions can be defended; neither can be verified.
The chosen peer set. Restricted-stock studies are averages of historical transactions, and the peer set used to compute the average matters. Selecting transactions from high-growth tech companies produces a lower average discount than selecting from mature, capital-intensive businesses. Both selections are defensible if the underlying company fits the profile. Both push the DLOM in opposite directions.
Where the dispute lands in practice
In a contested secondary, the DLOM disagreement typically looks like this: the buyer's appraiser applies a 15–20% discount (narrow range, liquid-adjacent sector, short assumed hold); the seller's appraiser applies a 30–40% discount (wide range, restricted class, uncertain exit timeline). The same share class, the same data room, a 20-point spread in the discount, and therefore a valuation gap that can be 25–40% of the total transaction value.
Neither number is fabricated. Both can be defended in a deposition. That is precisely why the disagreement is so difficult to resolve without a structured process: there is no clean external reference to appeal to, and a joint appraiser simply produces a third number that both sides are equally free to reject.
The same dynamic appears in dissenting-shareholder appraisals, in estate proceedings where the IRS challenges the DLOM applied on a Form 706, and in divorce proceedings where restricted equity is on the marital schedule. In each case, the DLOM is one input among several — alongside control premiums, minority discounts, and the baseline enterprise value — but it is often the input that accounts for most of the disagreement between the parties' final numbers.
What a mediated room does with the DLOM
A Velmark room does not adjudicate which DLOM is correct. It does something more useful: it surfaces every assumption behind every number, in a sealed compartment, in front of both sides, and lets the field of competing analyses resolve itself into a median that both sides have contributed to.
Each bidder in the room posts not just a price but the basis for it: the DLOM applied, the method used, the assumed holding period, the peer set, the control or minority adjustment. The AI neutral scores every position on the consistency of its assumptions, the recency of its data, and the quality of its comparables. A bid with a well-supported DLOM and a clear peer set ranks higher than a bid with an extreme discount and thin reasoning — even if the extreme bid is numerically between two better-supported positions.
The median that emerges from the room is not a compromise. It is the central tendency of the best-reasoned positions submitted. It carries an audit trail: every assumption is recorded, every scoring rationale is written, and the record is export-ready for attachment to a fairness opinion, a court filing, or a beneficiary's counsel's review.
What this means for counsel and shareholders
If you are advising a minority holder in a contested secondary, the DLOM is not a fixed input you can simply stipulate away. The range is real, the disagreement is legitimate, and no single appraiser can close it. What you can do is change the process: rather than a battle of experts in which the tribunal picks a winner and both sides lose the cost of the fight, a structured room produces a mediated number in days, with reasoning attached, that both sides have formally participated in.
If you are the company running the tender, the same logic applies in the other direction: a DLOM that is too low is a precedent you will face again on the next secondary. A mediated room produces a number with a clear audit trail, which is a better foundation for the next transaction than a number that one side can credibly say was imposed rather than agreed.
The illiquidity discount is not going away. Private shares are illiquid; the discount is real. The question is whether the number that captures it comes from a process both sides can read, or a process neither side chose.