The paradox is elegant and slightly disturbing: the person who wins a competitive bidding contest is often the person who overpaid. Not because they made a mistake in judgment, but because winning itself is evidence of overestimation.
This is the winner's curse — one of the most important and underappreciated concepts in decision theory, economics, and strategy. It explains patterns that puzzle people across a remarkable range of domains: why so many corporate acquisitions destroy shareholder value, why IPO investors on average underperform, why talent bidding wars tend to leave the winner feeling cheated, and why the oil companies that consistently won offshore drilling auctions consistently earned the worst returns.
Understanding the winner's curse is not about becoming a more cautious bidder. It is about understanding what information is embedded in the act of winning, and how to incorporate that information into better decisions.
The Origin: Oil Leases in the Gulf of Mexico
The winner's curse was first formally identified and named by three petroleum engineers — Richard Capen, Robert Clapp, and William Campbell — in a 1971 paper published in the Journal of Petroleum Technology. The paper, "Competitive Bidding in High-Risk Situations," documented an anomaly in offshore oil lease auctions conducted by the US government.
In these auctions, multiple oil companies bid on the rights to explore and extract oil from specific tracts of the Gulf of Mexico. The true value of any given tract — how much oil it contained — was genuinely unknown. Each company conducted independent geological surveys and produced its own estimate of the tract's value, then bid accordingly.
The engineers observed that companies winning these auctions consistently earned below-normal returns on their investments. The companies that lost bids tended to avoid the worst outcomes. This seemed backward: shouldn't the most successful bidder — the one willing to pay the most — be the one with the best information or the most efficient operations?
The explanation they proposed was statistical rather than judgmental: when many independent parties estimate an uncertain value and the estimates vary around the true value, the highest estimate is almost certain to be above the true value. Whoever submits that highest estimate wins the auction — and wins an asset they overpaid for.
The winner's curse, in other words, is not a bias or a mistake. It is a predictable consequence of competitive bidding on uncertain-value assets, even when all bidders are perfectly rational.
A subsequent study by Edward Capen, Robert Clapp, and William Campbell (1971) calculated that if oil companies had been fully rational and corrected for the winner's curse, they should have bid roughly half their naive estimates on contested tracts. The fact that actual bids were far higher suggests that even sophisticated professional bidders systematically failed to account for the statistical logic they were caught in.
The Theory: Why Winning Is Bad News
Consider a simplified example. A jar contains an unknown number of coins. You and nine other people each estimate independently, then bid based on your estimate. Your estimates might look like this:
| Bidder | Estimate |
|---|---|
| A | $14.20 |
| B | $17.80 |
| C | $19.50 |
| D | $22.10 |
| E | $25.00 |
| F | $26.30 |
| G | $28.00 |
| H | $30.50 |
| I | $33.20 |
| J | $37.80 |
Suppose the jar actually contains $25.00. The average estimate is roughly $25.44 — the group's average is accurate. But the winner (J, who bids $37.80) has systematically overpaid by $12.80.
This is not because J is foolish. The estimates vary around the true value as you would expect from good-faith independent estimation. The problem is that the auction selects specifically for the highest estimate. The winner's information is correct on average — but the act of winning tells you that your estimate was probably above average, which in expectation means above the true value.
"In a common-value auction, winning conveys bad news: it means your estimate was higher than everyone else's. A rational bidder should shade their bid downward to account for this information. The magnitude of the required discount increases with the number of bidders." — Robert Wilson, 2020 Nobel Laureate in Economics
This reasoning — sometimes called the winner's curse correction — means that sophisticated bidders in common-value auctions should never bid their honest estimate. They should bid below it, with the discount increasing as the number of competitors increases.
Economists Robert Wilson, Paul Milgrom, and their colleagues formalized this in auction theory. Wilson and Milgrom shared the 2020 Nobel Prize in Economics partly for this work on how bidders should behave in auctions and how auction design affects outcomes.
The Mathematics of the Correction
If bidders have symmetric information and errors are normally distributed around the true value, auction theorists have derived the optimal bid discount. With N bidders, a rational bidder should reduce their estimate by approximately the expected maximum deviation of a sample of size N from the population mean. As N grows, this discount grows — because more bidders means a more extreme winning estimate in expectation.
In practice, few bidders apply anything close to this correction. A study by Max Bazerman and William Samuelson (1983) in the Journal of Economic Behavior and Organization ran controlled experiments with MBA students estimating and bidding on jars of coins. The true value averaged $8.00. Winning bids averaged $10.01 — a 25% overpayment, despite participants knowing they were in an experimental auction and having full information about the setup. The winner's curse persisted even under conditions designed to teach it away.
Common Value vs Private Value
The winner's curse is strongest in common-value auctions — contests where the item's value is approximately the same to all bidders and is uncertain at the time of bidding. Oil drilling rights are the canonical example: the oil is worth roughly market price regardless of who extracts it, and the amount of oil is unknown.
It is weaker or absent in private-value auctions — contests where the item's value genuinely differs between bidders and each bidder knows their own value with certainty. Antiques auctions can approach this model: if one collector has a deep personal attachment to a particular item, their private value genuinely exceeds others', and paying above the average estimate is not necessarily overpaying.
Most real-world competitive situations involve a blend of both. An acquiring company values a business partly on its standalone market value (common value, susceptible to the curse) and partly on synergies specific to their business (private value, where overpaying relative to others may be rational).
| Auction Type | Winner's Curse Strength | Example | Appropriate Response |
|---|---|---|---|
| Pure common-value | High | Oil lease rights, spectrum licenses | Large bid discount; consider abstaining |
| Predominantly common-value | High-moderate | Corporate acquisitions | Separate common/private value components; discount the common value estimate |
| Mixed-motive | Moderate | Real estate, talent recruitment | Identify specific private value clearly before bidding |
| Pure private-value | Low to none | Personal collectibles, subjective art | Bid near private valuation |
Understanding which component dominates in a given situation is the first step toward managing the winner's curse.
The Winner's Curse in Mergers and Acquisitions
Corporate acquisitions are among the most costly arenas for the winner's curse. The evidence is damning and consistent.
Studies spanning multiple decades find that, on average:
- Acquirers pay premiums of 20-30% above pre-announcement market price for target companies
- Over 50-60% of acquisitions destroy value for the acquirer's shareholders over the subsequent three to five years
- Target shareholders receive most of the value created by acquisitions; acquiring shareholders often give it away
A landmark study by KPMG in 1999 found that 83% of mergers failed to create value for the acquirer's shareholders. McKinsey research, updated regularly through studies from researchers including Mckinsey's Marc Goedhart and Tim Koller, has consistently found that acquirers underperform their industry peers in the years following acquisitions by a margin of several percentage points annually.
A meta-analysis by Gregor Andrade, Mark Mitchell, and Erik Stafford (2001) in the Journal of Finance examined 1,000+ acquisitions from 1973-1998 and found that while target shareholders earned an average abnormal return of 16%, acquirer shareholders earned essentially zero — meaning the full value created by the combination was captured by sellers, not buyers. In competitive auctions (where the winner's curse is strongest), acquirer returns were negative on average.
The winner's curse mechanism is clear in competitive M&A processes: multiple potential acquirers are presented with the same information package by the target's bankers. Each estimates synergies independently. The one with the most optimistic synergy estimates — or the most acute pressure to do a deal — bids the most. They win. Their synergy estimates, being the highest among competitors, are the most likely to be overestimates.
"The best deals in M&A are often the ones you don't do. When you find yourself in a competitive process, the very fact that others are willing to pay a price close to yours should make you question your assumptions. The rational response to winning a bidding war is to wonder why you won." — Warren Buffett, Berkshire Hathaway shareholder letters
Berkshire Hathaway's approach is instructive. Warren Buffett explicitly avoids competitive auction processes, preferring to buy companies privately where he can establish his own valuation without the curse of competing estimates pushing his price to the most optimistic level. This discipline — which requires the patience to forgo many opportunities — has been a defining feature of Berkshire's acquisition success.
The "Deal Fever" Dynamic
A specific pathology amplifies the winner's curse in M&A: what practitioners call deal fever. Once a competitive acquisition process begins, organizations develop institutional momentum toward completion. Deal teams are assembled; advisors are hired; press releases are drafted. The sunk costs in the process create psychological pressure to complete it. Executives who have publicly announced interest face reputational costs from walking away.
Richard Roll's "hubris hypothesis" (1986), published in The Journal of Business, proposed that managerial overconfidence — not just statistical winner's curse effects — drives acquisition premiums. Roll argued that acquiring executives genuinely believe their estimate of target value exceeds the market's estimate, not because they have better information, but because they have excessive confidence in their own judgment. The winner's curse and hubris frequently reinforce each other.
IPOs and the Winner's Curse
The winner's curse has a specific and well-documented form in initial public offerings. IPO shares are allocated by the underwriting bank to investors who indicate interest. Informed investors (institutional funds with deep research capacity) participate selectively in IPOs they believe are underpriced. Uninformed investors (retail buyers, less sophisticated institutions) participate broadly.
The result, first modeled by Kevin Rock in his 1986 paper "Why New Issues Are Underpriced" in the Journal of Financial Economics, is an adverse selection problem for uninformed investors. When an IPO is underpriced (a good deal), sophisticated investors crowd in and there are not enough shares to go around — uninformed investors receive only a small allocation. When an IPO is overpriced (a bad deal), sophisticated investors stay away — uninformed investors receive their full requested allocation.
Uninformed investors win the most shares in the worst IPOs and the fewest shares in the best IPOs. Their overall return, despite participating in many IPOs with strong average performance, can be negative.
This is a variant of the winner's curse: receiving a full allocation is evidence that sophisticated money did not want the deal.
Rock's model was empirically validated by subsequent studies. Jay Ritter (1991, Journal of Finance) documented that IPOs underperform the broader market by approximately 15-20% over the three years following their offering — consistent with the prediction that many IPOs are overpriced and that investors who receive full allocations hold underperforming securities.
The practical implication: investors who receive full allocations in heavily subscribed IPOs should treat that as a warning sign, not a windfall. The pattern "I got everything I asked for" is the IPO equivalent of winning an auction — it means everyone else passed.
Talent Bidding Wars
When multiple employers compete for a single candidate, the hiring equivalent of the winner's curse operates. The company that offers the most must believe the candidate's value to them exceeds all other employers' assessments. If competitors have broadly similar information about the candidate's abilities, the winner is paying a premium above market value.
This does not mean competitive hiring is irrational. A candidate may genuinely create private value for one employer that exceeds their general market value — through a uniquely relevant skill set, an existing relationship with key clients, or cultural fit that reduces management costs. When private value is real and large, winning a bidding war can be rational.
The danger is when competitive pressure causes organizations to bid for talent based on social proof ("if our top competitor wants this person, they must be excellent") rather than independent assessment of the candidate's fit and value.
Research on executive compensation consistently finds that companies that engage in the most competitive CEO searches — typically after high-profile departures requiring public searches — pay significantly above-market compensation and experience worse subsequent performance than companies that promote internally or hire through quieter processes. A study by Rakesh Khurana published in his 2002 book Searching for a Corporate Savior documented how the "market for celebrity CEOs" produces exactly the winner's curse dynamic: the winner of a high-profile CEO search has typically outbid competitors who had equally good information and lower valuations.
Other Domains: Spectrum Auctions, Real Estate, eBay
The winner's curse appears wherever competitive bidding meets uncertain value:
Spectrum auctions: Governments selling mobile spectrum rights have observed consistent winner's curse patterns. Companies that won major spectrum auctions in the early 2000s, including 3G licenses in Europe, paid amounts that proved unrecoverable when projected revenues failed to materialize. Several acquirers went bankrupt or had to write down enormous sums. The British 3G auction alone raised over 22 billion pounds — an amount that analysts later concluded was roughly three to five times the licenses' economic value.
Real estate: In competitive property markets where multiple offers are common, buyers who "win" a bidding war frequently pay above comparable sales prices. The pressure of the competition and the fear of losing the property drives bids above independent valuations. A study by David Genesove and Christopher Mayer (2001, Quarterly Journal of Economics) on Boston condominium markets found systematic evidence of bidding premiums in competitive offer situations that persisted in subsequent resale performance — houses won in bidding wars resold at lower relative premiums than comparable properties.
eBay and online auctions: Research on eBay's auction format found that when many bidders compete on identical items sold by different sellers simultaneously, the winning prices on a given item are systematically higher when there are more bidders — consistent with the winner's curse prediction. A study by Patrick Bajari and Ali Hortacsu (2003, RAND Journal of Economics) confirmed winner's curse patterns in eBay coin auctions, a near-perfect common-value setting.
Free agent sports contracts: Teams signing star athletes in competitive free agency consistently overpay relative to subsequent performance. Studies of NBA and NFL contracts found that players signed in the most competitive free agency markets (most bidding teams, highest media attention) significantly underperform expectations relative to their contract value more often than players signed with less competition. Research by David Berri and Martin Schmidt, summarized in Stumbling on Wins (2010), found that NFL teams consistently overvalue players they competed hardest to acquire.
How to Avoid the Winner's Curse
The winner's curse is not entirely avoidable — it is a structural feature of competitive bidding — but its damage can be substantially reduced.
Distinguish Private Value from Common Value
Before entering a competitive process, explicitly identify: what is the common value of this asset (roughly the same to all bidders) and what is my private value (what makes this uniquely valuable to me specifically)? Only the private value premium justifies bidding above market consensus.
Write this down before the auction begins, with specific numbers: "The common value of this company is approximately $X based on comparable transaction multiples. Our synergy case adds $Y of private value. Our maximum rational bid is therefore $X + $Y." This exercise forces the separation before competitive dynamics obscure it.
Bid Below Your Estimate by an Amount Proportional to Competition
In common-value auctions, the appropriate discount on your estimate increases with the number of competing bidders. With 10 competent bidders, you might shade your bid 15-20% below your estimate. With 50 bidders, the required shade is larger. This is uncomfortable but mathematically correct.
The formula is not mechanical — the right discount depends on how variable you expect estimates to be. With highly uncertain assets (oil exploration rights, early-stage companies), the spread of estimates will be wide, and the winner's curse is severe. With assets that can be valued more precisely (real estate in a liquid market, businesses with stable earnings), the spread is narrower, and the required discount is smaller.
Set a Maximum Before the Auction Begins
Pre-commitment is powerful precisely because it resists the in-the-moment pressure that drives overbidding. Decide the maximum you will pay before the auction begins, write it down, and treat it as inviolable. The competitive heat of a live bidding process is specifically the condition under which the winner's curse is amplified.
Research by behavioral economist Dan Ariely and colleagues (2008) demonstrated that competitive arousal — the heightened emotional state produced by a live bidding contest — causes people to bid above their pre-stated maximum even when they are aware of this tendency. The only reliable countermeasure is a written, pre-committed ceiling established in a calm, deliberative state.
Ask Why You Are Winning
If you are winning a competitive process, ask: why are my competitors not willing to pay what I am willing to pay? Do they have information I lack? Are they less interested because they have better alternatives? This question is uncomfortable but crucial. If you cannot articulate a convincing answer based on private value, the winner's curse may be at work.
This is sometimes called the Sherlock Holmes test for competitive processes: when you find yourself with the winning bid, the correct response is not celebration but investigation. The absence of competing bids at your price is evidence that needs explaining.
Consider Not Participating
For competitive processes where common value dominates and you have no clear private value advantage, the best strategy may be to decline participation entirely. This is Warren Buffett's approach to competitive M&A auctions. In many markets, the expected value of participating in a competitive auction is negative for the buyer.
The willingness to walk away from competitive processes requires both analytical clarity (understanding which processes favor buyers and which do not) and organizational culture that supports passing on opportunities. In many organizations, "we chose not to compete" is treated as failure; analytically, it is often the optimal choice.
The Winner's Curse and Risk Management
The winner's curse has a specific implication for organizational risk management: avoid processes that select for the most optimistic forecast.
Most capital allocation processes in large organizations involve competing proposals, with resources going to the most compelling case. The most compelling case is typically the one with the most optimistic projections. This selects systematically for overconfidence, creating a winner's curse dynamic in internal resource allocation.
Researcher Bent Flyvbjerg (2003) documented this exact pattern in large infrastructure projects: the projects that receive funding are those with the most optimistic projections, and those optimistic projections are systematically wrong. Across thousands of infrastructure projects globally, the average cost overrun is 45% and the average benefit shortfall is 20-50%. The selection process — funding the most optimistic proposals — produces the winner's curse in every capital allocation cycle.
Techniques that counteract this include requiring proposals to include explicit downside scenarios, giving additional weight to track records versus projections, using reference class forecasting (what is the base rate for projects like this?), and creating institutional rewards for accurate forecasting rather than optimistic forecasting.
A practical technique from Daniel Kahneman's work: require that any investment proposal include an "outside view" estimate — what has happened historically to similar projects? — alongside the inside view analysis. The outside view anchors discussion and prevents the selection process from running purely on optimistic projections.
Summary
The winner's curse is a predictable consequence of competitive bidding on uncertain-value assets: when the highest estimate wins, the winner has almost certainly overestimated. The curse is strongest in common-value settings, when there are many bidders, and when competitive pressure removes the deliberative space needed for careful valuation.
Its fingerprints are visible across corporate acquisitions, IPO markets, spectrum licenses, real estate bidding wars, sports free agency, and talent recruitment. The consistent pattern — winners overpay and underperform — is not a failure of intelligence but a failure to account for what winning means.
The core countermeasures are discipline (pre-commitment to bid limits), analysis (separating private value from common value), and sometimes courage (walking away from competitive processes you cannot win on rational terms). The companies and individuals who manage it best share one habit: they ask hard questions about why they are winning, before the prize is in hand.
The deeper insight is epistemological: in any competitive information environment, the position you are in is itself evidence about the quality of that position. Winning tells you something. Receiving a full allocation tells you something. Being the only bidder willing to pay this price tells you something. The winner who ignores that information pays the curse. The winner who heeds it pays a rational price.
Frequently Asked Questions
What is the winner's curse?
The winner's curse is the tendency for the winning bid in a competitive auction to exceed the true value of the item being won. It occurs because when many bidders each independently estimate value and the estimates vary around the true value, the highest estimate — which wins — is likely to be an overestimate. The winner, by definition, is the most optimistic bidder.
Where does the winner's curse come from?
The winner's curse was first identified and named by petroleum engineers Richard Capen, Robert Clapp, and William Campbell in a 1971 paper about oil lease bidding in the Gulf of Mexico. Companies that won bids on offshore drilling rights consistently earned below-normal returns, while companies that lost bids tended to avoid the worst outcomes. They attributed this to systematic overestimation by the winning bidders.
How does the winner's curse apply to mergers and acquisitions?
In M&A, the acquirer who wins a competitive auction for a target company is statistically likely to have overestimated the target's synergies or standalone value. Academic research consistently finds that acquisition premiums average 20-30% above market price, and that over half of acquisitions destroy value for the acquirer's shareholders. The competitive pressure of a bidding process amplifies the winner's curse dynamic.
Does the winner's curse apply to job offers and talent bidding?
Yes. When multiple employers compete for a single candidate, the company that offers the most must believe the candidate's value to them exceeds everyone else's assessment. If competing employers have roughly similar information, the winner is likely paying a premium above the candidate's market value. This does not mean the hire is wrong — private value may genuinely exceed market value — but it suggests caution about bidding wars.
How can you avoid the winner's curse?
The main strategies are: bid below your independent value estimate by an amount proportional to the number of competing bidders; separate your private value (what the item is worth specifically to you) from common value (what it is worth to anyone); and set a maximum bid before the auction begins and commit to not exceeding it. In M&A, this means basing bids on conservative synergy estimates and walking away from competitive processes when the price climbs above pre-set thresholds.