Can You Rebuild the Mark You Reported?

An LP challenges a portfolio mark — a company the fund carries at a value the LP reads as stale, or too generous. The GP is confident in the number. What decides whether the fund can defend it is not whether the mark is right in some abstract sense but whether the fund can show how it got there: the technique it used, the assumptions behind it, the evidence it rested on, laid out in enough detail that a reviewer can follow the judgment and weigh it. A reviewer who can follow the build may still reject the assumptions or the conclusion. A reviewer who can't follow it has nothing to weigh at all, and the mark is exposed no matter how sound it was.

What defending a mark actually requires

A fair-value mark is an estimate, not an observation. Under ASC 820, the U.S. GAAP standard for fair-value measurement, fair value is an exit price — what the fund would receive to sell the asset in an orderly transaction between market participants — and for a private holding that price has to be built rather than looked up. ASC 820 does not fix a single method; it asks for one or more valuation techniques consistent with the market, income, or cost approach, chosen to fit the asset. A technique sits beneath its approach — multiples of comparable companies or a recent transaction price under the market approach, a discounted-cash-flow model under the income approach, current replacement cost under the cost approach — and a single mark can draw on more than one. Whichever it uses, the inputs are meant to reflect what a market participant would assume, and hard-to-value private holdings often rely significantly on unobservable inputs and on judgment.

None of that is settled by being able to reproduce the number. A fund can reproduce a flawed method perfectly — apply the wrong multiple consistently, carry an assumption no market participant would accept, and arrive at the same defective figure every time. Reproducibility is a necessary condition of a defensible mark, not the whole of it. What it buys is reviewability: a build that can be reconstructed is one a reviewer can test against the standard, and a build that can't be reconstructed forecloses that test before it starts.

What a challenge tests

When an LP or an auditor questions a mark, they do not necessarily need to arrive at the same number. Fair value for an illiquid asset is a considered estimate, and reasonable judgment applied to the same evidence can land on a different figure within a defensible range. What the challenge tests is the reasoning: whether the approach fits the asset, whether the assumptions are supported, whether the evidence was current, whether the method was applied the way the fund says it was. The mark is questioned through its build, and the fund answers by walking the reviewer back through it.

The control that makes a mark reviewable

That is why a defensible valuation is a documented one, and documented in a particular way. The support has to be contemporaneous — assembled when the mark is struck, not reconstructed from memory once it is questioned. It records the technique chosen and why it fits the asset, the inputs and where they came from, the assumptions stated plainly, and, where the method turns on comparables or on calibration to a recent transaction, which ones were used and which were considered and set aside. It carries dates. And it passes through a review and an approval, so the judgment is not only written down but checked before it becomes the number of record. None of that guarantees the mark is right; it makes the mark reviewable, which is the precondition for ever showing that it is.

The useful implication runs backward from the dispute. Whether a mark can be defended is mostly determined before anyone questions it — by whether the reasoning was recorded, and reviewed, when the mark was made. Contemporaneous support tends to be more persuasive than an account assembled after a challenge lands, and a mark that passed a genuine review when it was struck starts from firmer ground than one whose basis has to be rebuilt under pressure. The work is less about winning the argument than about having done the valuation well enough, and recorded it well enough, that there is something solid to stand on when the argument comes.