In Due Diligence, Value You Can’t Prove Gets Priced as Risk

← Back · ACT I · Chapter 4: The Information Asymmetry Collapse

Value governed by a standard is legible in the data room; value that lives in someone else’s method gets priced as a risk, or leaves at renewal.

Diligence has become thorough about the technology and quiet about the thing that actually concentrates risk. The models, the vendors, the contract terms, the data location, the security review, all of it gets examined. What rarely gets examined is the accumulated logic that decides how the business behaves, the thresholds and escalation paths and rules about what runs without a human. That logic is the concentration. It is also the part no standard diligence question is built to find.

The market has started to notice, and it is worth being exact about how much, because the looser version of this claim is easy to find and does not survive contact with the research. EY-Parthenon’s Global Private Equity Exit Readiness Study, published in June 2026 from a survey of a hundred private equity executives and a hundred executives at recently exited portfolio companies, found that 18 percent of general partners now name AI as a challenge in preparing a company for sale, against 7 percent a year earlier. That is a sharp rise off a small base, and it is a finding about preparation rather than about deals collapsing. What buyers are doing, on EY’s account, is probing exposure and asking for evidence, which is the stage before a discount rather than the discount itself. The question underneath that probing is narrow and hard. Does the business own what makes it repeatable, or does that belong to whoever built it and walk out when they do.

Diligence Prices What It Can Verify

A quality of earnings review tests whether the numbers are repeatable. It does not test whether the repeatable thing is owned. When the decision logic sits inside a vendor’s delivery method or a single person’s judgment, the acquirer cannot confirm it survives the sale. So the acquirer does the only rational thing and prices the doubt. There are figures circulating for how large that discount runs, and they are worth leaving alone: the ones in wide circulation come from valuation-software and brokerage firms publishing them alongside the service that fixes the problem, with no sample and no method attached. The size is not the point in any case. The point is that a discount exists, that it is applied to the seller, and that nothing in a standard diligence pack is built to argue it down.

A Standard Is Written Where A Buyer Can Read It

The diagnosis to this point is not original and should not be dressed up as though it were. Advisory firms have been publishing it since at least the start of 2026, and their prescription is always the same one: prepare earlier, document more, run a readiness review, engage the firm publishing the advice. That puts the work on the seller in the final months and produces a documentation effort rather than a property of the business. The alternative is narrower and it starts much earlier. Value governed by a standard requires the decision rules to exist in a form the buyer holds. That is not a courtesy extended to the data room. It is the condition under which the value is recognized in the first place. The logic is legible because unwritten logic does not meet the standard, not because a diligent owner remembered to document it before the sale. A buyer does not have to reconstruct how the business decides. The standard already required it to be readable.

There is a second thing the standard settles. A named owner can leave the seat vacant, because roles change and mandates get redrawn while the logic keeps running with no one accountable for it. A standard has no ownerless decision logic. Custody is not granted to a seat that can empty, it is a default of how the value was governed, so the buyer inherits both the rules and the accountability for them.

Legible Value Does Not Discount At Exit

A target that governs value by a standard does not carry the hidden dependency, so it does not carry the hidden discount. What is repeatable is demonstrably the company’s, transferable on the day the deal closes, and intact when the vendor and the key employee both move on. The seller does not spend the final months of a process trying to prove a negative. The value was owned in a form a buyer could read from the start, and that shows up in the price.

A buyer discounts what it cannot verify. Governing value by a standard is what removes the thing being taken on faith, and it has to be in place long before anyone arrives to check. The same structure answers a question that arrives from the opposite direction, when a duty of oversight has to be discharged rather than assigned, and it is the reason owning the machinery was never the point.

References

EY-Parthenon, Global Private Equity Exit Readiness Study 2026, published 2 June 2026, authors Konstanze Nardi and Ivan Lehon, study here. The figures used above are the study’s own: 18 percent of general partners and 17 percent of portfolio company executives name AI as an exit-preparation challenge, against 7 percent of general partners a year earlier, from a survey of 100 general partners and 100 portfolio company executives conducted between February and April 2026. Opened and read 14 September 2026. The valuation-discount percentages that circulate on this subject are not cited here, deliberately: the sources carrying them are firms selling valuation software and transaction advisory, and none of them publishes a sample or a method.

Keep reading: ACT I · Chapter 4: The Information Asymmetry Collapse