Common-value auction effect
Winner's Curse
In a common-value auction, the highest bidder is disproportionately likely to have received an overly optimistic signal unless the bid adjusts for winning.
E[value | my signal, I win] < E[value | my signal]
Winning is informative because it means competing estimates were lower. Rational bidders shade bids to account for this adverse selection.
All bidders estimate the same hidden value with noise. The winning estimate and realized value remain visible over repeated auctions.
(%)
The animation runs automatically, pauses on the conclusion, and then repeats. The main control changes the scenario rather than scrubbing the timeline.
- CHANGE
- Number of bidders
- WATCH
- winner overestimate
- MEANING
- All bidders estimate the same hidden value with noise. The winning estimate and realized value remain visible over repeated auctions.
The maximum estimate is selected partly because its error is high.
A dot plot of bidder signals highlights the winning bid, the hidden common value, and the bid after conditional-value adjustment.
What it actually says
The curse is a conditional inference problem. Even unbiased estimates become biased after selecting the maximum. More bidders can intensify the selection pressure.
It is most relevant when the item has a shared but uncertain value, such as resource rights or acquisition synergies. In private-value auctions, each bidder may rationally value the item differently.
"A useful law compresses a pattern. It does not erase the conditions that make the pattern true."
How the idea developed
The modern form emerged through observation, argument, and later refinement. The timeline separates the first insight from the version now used in textbooks and practice.[1]
Capen, Clapp, and Campbell describe the effect in oil-lease bidding.
Auction experiments document overbidding and learning.
Procurement, M&A, and spectrum design account for information structure.
How the pattern works
The relation becomes useful only when its mechanism, measurement process, and operating range are visible.
Bidders estimate one uncertain value.
The highest signal tends to contain positive error.
Naive bids ignore the information in winning.
Winning is informative because it means competing estimates were lower. Rational bidders shade bids to account for this adverse selection.
Where it earns its keep
Applications are strongest when the law changes a decision, measurement, model, or experiment rather than merely providing an analogy.
Model conditional value
ApplicationRe-estimate after assuming every rival bid lower.
Use scenario ranges.
Separate synergy from selection optimism
ApplicationCompetitive processes amplify optimistic forecasts.
Apply outside-view checks.
Where it stops working
Experienced bidders can learn, and auction format, information release, affiliation, and private values change the prediction.
"The winner always loses money"
Better: Adjusted bidding can remain profitable."Every expensive purchase is cursed"
Better: The mechanism requires common-value uncertainty and competitive selection.Sources and further reading
Original publications and serious secondary scholarship are prioritized over summaries.
- Capen, Clapp, and Campbell - Competitive Bidding in High-Risk SituationsOriginal petroleum-auction paper.https://onepetro.org/JPT/article/23/06/641/164384/Competitive-Bidding-in-High-Risk-Situations
- Thaler - Anomalies: The Winner's CurseClassic review.https://doi.org/10.1257/jep.2.1.191
- Kagel and Levin - Common Value Auctions and the Winner's CurseExperimental and theoretical synthesis.https://press.princeton.edu/books/paperback/9780691056760/common-value-auctions-and-the-winners-curse