Winner's Curse - OrgRiver

When several bidders are estimating a value that’s largely common to all of them, the winner is unusually likely to be the bidder whose estimate was too high, simply because winning selects for whoever guessed highest.

4 min read

What Is It?

The winner’s curse was first identified by petroleum engineers in the 1970s studying offshore drilling rights auctions, where the company that won the bid consistently seemed to overpay relative to what the field actually produced. The mechanism is straightforward once named: when several bidders independently estimate the value of something genuinely uncertain, their estimates will scatter around the true value, some too high, some too low, some close. The highest bidder wins. But the highest bid is, by construction, drawn from the high end of that scatter, which means the winning bid is more likely than any individual bid to be an overestimate, not because the winner is worse at judging value than everyone else, but purely because winning selects for whoever guessed highest. The effect is strongest specifically when the value being bid on is largely common to every bidder, how much oil is actually underground, what a company will ultimately earn, what an asset can be resold for, rather than something whose value genuinely differs from bidder to bidder. An offshore oil lease is the canonical example precisely because the amount of oil underground doesn’t change depending on who buys the lease. Where the value genuinely is bidder-specific, a buyer with real synergies, a company that can use an asset in a way no competitor can, winning isn’t automatically evidence of a bad estimate, it can reflect a real difference in what the thing is actually worth to that particular bidder. Most real organizational bidding situations mix both: some of the value is common to everyone, some of it is genuinely specific to the winner, and the curse applies most cleanly to the common-value share.

Why Does It Matter?

A bidding war for a target company or a contract, where the thing being bid on has a real, largely common value, future earnings, resale value, market position, rewards conviction more than accuracy. The organization willing to pay the most is frequently the one that most overestimated the common-value component, which means winning a competitive process is weaker evidence of a good deal than it feels like in the moment, precisely because the process was designed to surface the highest number, not the most accurate one. This is easy to miss because winning feels like validation. A large gap between the winning bid and the next-highest one is worth investigating rather than celebrating, it may reflect real value available only to the winner, but it may also mean the winner’s estimate was simply the outlier. Competitive hiring is a genuinely mixed case and deserves more care than treating it identically to an oil lease. A candidate’s general underlying capability has something like a common value that every employer is independently estimating, and the winner’s curse can apply to that part. But a candidate’s fit and productivity are also partly employer-specific, the same person really can be worth more to one company than another because of different technology, team, or role, which means winning a competitive offer isn’t automatically evidence of overpaying the way it more cleanly is for an oil field. The organizational habit this exposes is treating “we won” and “we got it for the right price” as the same fact. They’re often not, at least not automatically. The very fact that a bid or offer won a competitive process for something with real common value is a signal worth discounting, not confirming.

What Changes Once You See It?

You start treating a wide gap between the winning bid and the next-highest one as a signal worth investigating rather than a triumph, since it raises the possibility that the winner’s estimate was the outlier, not the most accurate one. You start “shading” a competitive bid downward when the underlying value is largely common and genuinely uncertain, recognizing that the conditions under which you win are also the conditions under which your own estimate is most likely to have been the optimistic one. You also get more comfortable losing a bidding war on purpose once the price crosses what your own best estimate says the thing is worth, since losing a bidding war for an overpriced asset isn’t actually a loss.

Common Misunderstandings

It isn’t a claim that the winning bidder is worse at estimating value than the losing bidders. Any individual bidder’s estimate could be equally accurate on average, the curse comes from the selection effect of winning, not from any particular bidder’s skill.
It isn’t the same as simple overpaying through carelessness. The curse can strike careful, well-informed bidders just as easily, since it’s a structural property of competitive bidding under uncertainty, not a failure of diligence.
It doesn’t apply when the value being bid on is actually known with confidence rather than genuinely uncertain. The curse specifically requires real uncertainty about value, which is why it shows up hardest in acquisitions, talent markets, and anything without a clean, observable market price.
It isn’t a reason to avoid competitive processes altogether. It’s a reason to build the expected overestimate into how much weight you give your own winning bid.
It isn’t a claim that winning is always evidence of overpaying. Where the value genuinely is bidder-specific, real synergies, a genuine unique fit, winning can reflect a true difference in worth rather than an estimation error. The curse applies most cleanly to the portion of value that’s common to every bidder, not to value that’s genuinely specific to the winner.

Diagnostic Question

If we win, what should the fact that every other bidder valued this less than we did cause us to reconsider about our own estimate?

Explore Further

Field Notes

None yet.

Related Field Guide

Origin

First identified by Edward Capen, Robert Clapp, and William Campbell, “Competitive Bidding in High-Risk Situations” (1971), Journal of Petroleum Technology, from patterns observed in offshore oil-lease auctions; later formalized and extended in auction theory and behavioral economics.