Roll’s move is to take the evidence that acquirers gain little and offer an explanation that requires no villains. Managers need not be empire-building or self-dealing. They need only be confident, individually reasonable, and wrong in a particular direction. Across many valuations of the same target, the estimates scatter; the bid that wins is drawn from the high tail of that scatter. Winning is therefore evidence about your own estimate, and the evidence is unflattering.
For a private buyer the mechanism is easy to dismiss and should not be. It is tempting to say that a proprietary conversation with a retiring owner involves no auction and therefore no winner’s curse. But the curse does not require rival bidders, only a distribution of possible valuations of which yours is one. If your number is the one that justifies proceeding and a more cautious analysis would not, you are standing in the same tail Roll described. This is why the Institute’s harder question is not what the business is worth but what would have to be true for your number to be too high.
Jensen and Ruback, also on this shelf, marshal the evidence that corporate control transactions do create value, and Fuller, Netter and Stegemoller find that buyers of private targets earn positive returns where buyers of public ones do not. Roll is best read as an explanation of bidder behavior rather than a verdict that acquisitions destroy value, and the private-company reader should hold it alongside that private-target evidence rather than instead of it.