Curse of the winner

from Wikipedia, the free encyclopedia

The curse of the winner (Engl. Winner's curse ) describes the effect that the highest bidder in auctions with incomplete information systematically pays too high a price. The winner's curse belongs to the field of auction theory and behavioral economics .

theory

idea

The idea of ​​the winner's curse first came up in 1971 in a scientific paper by EC Capen, RV Clapp and WM Campbell. The three engineers had observed companies drilling for oil in the Gulf of Mexico. They had hoped for a certain yield and, on the basis of the latter, bought oil drilling rights for an area. Although they found a lot of oil and gas in the area, they did not make as much profit as hoped. Based on this observation, they formulated the following idea: The initial situation is that there is competition in the form of a covert auction for the sale of patent rights. In this situation, common sense tells us the following: The winner of this auction is the one who most overestimates the value of the raw materials to be mined. From this it follows that the winner not only wins - but potentially cannot cover the costs of the auction with later income or reduce his profit. Capen and his colleagues also found that the individual bidders either bid too high or too low, but on average the bid could correspond to the true value of the object to be auctioned.

definition

The oil field auction situation described is a common value auction . This means that the drilling rights are worth the same to all bidders - i.e. have an objective true value, which can be understood here as the market price of oil (and gas) - and the bidders have incomplete information about the value. Bidders' uncertainty about true value is a defining feature of the winner's curse. The bidders can obtain an estimate of the value from an expert . Since it is very difficult to determine the exact value of the oil in the region, the bidders will receive different estimates. Those bidders who received the higher estimates will tend to bid higher than those who received the lower estimates. The bidder with the highest bid will win the auction. In this case, however, the winner will lose and be “cursed”.

When the winner's curse occurs, there is a situation in which the "winners of a [common value auction] [...] often do worse than those who did not, because the winner is usually overly optimistic and bids therefore more for the auction object than it is actually worth ”. Bidders often overlook the fact that common value auctions come under the curse of the winner and can therefore not adapt to them.

There are two versions of this curse: The bid with which the auction is won can be greater than the value of the auctioned area and thus the company will lose. The second (milder) possibility is that the value of the area is below the expert's estimate and the company will be disappointed. The disappointment is that the company is not making a profit - but neither is it making a loss - or is making less profit than it had hoped. In both cases, the bidder was unhappy with the result of the auction, said Richard Thaler .

The winner's curse is measured as the difference between the individual bid and the objective (unknown) true value of the good.

causes

The reason individuals succumb to the winner's curse is to omit relevant information in the bidding process, argued Bazerman and Samuelson in 1983. If a bidder believes his bid will win the auction, it should show him that he is probably overestimated the value of the good compared to the other bidders and should correct it downwards. It is, however, possible that winning gives the bidder a benefit or that the good has a personal value for the bidder and thus shows no remorse, although the winner is curse. Conversely, a winning bidder can pay less than the objective value and experience the subjective curse of the winner. This can happen, for example, if a person wants to pay exactly 85% of the objective value, they would be disappointed if they bought it at 92% of their value. The psychological literature suggests that case one is more likely than case two.

Under what conditions is it likely that the winner's curse will occur? Bazerman and Samuelson identify two factors that will affect the likelihood and extent of the winner's curse. The first factor is the degree of uncertainty about the value of the item being auctioned. The greater this uncertainty, the greater the variance in the estimates of the value of the bidders. For example, if a one dollar bill were auctioned, there would be no uncertainty about the value and there would be no bids over one dollar. For a jar with 100 pennies (the exact number is unknown to the bidders) the uncertainty is higher and there is greater variance , which offers the possibility of the winner's curse. One possibility to prevent the curse would now be to reduce one's own commandment as the uncertainty increases. If a bidder fails to reduce the bid despite uncertainty, it increases the likelihood and extent of the winner's curse. The second factor is the number of bidders. As the number of bidders increases, the range of estimates and bids will also increase. In this case, too, the following applies: If the number of bidders increases, bidders should reduce their bids in order to counteract the increased probability of outbidding.

Bazerman and Samuelson conducted an auction experiment with MBA students at Boston University in 1983 . The item up for auction was a mason jar full of coins. The students were not aware that the coins of each jar were worth eight dollars. Now the students had to place secret bids and the highest bid would win the glass. The average of all bids placed was $ 5.13, which is less than the true value of $ 8, so it was a systematic underestimation. Even so, the average winning bid was $ 10.01, which translates into an average loss to the winner of $ 2.01. Although the bidders are aware that there is uncertainty, the experiment shows that they do not take this into account sufficiently in their bids. The results also show that they do not adjust their bids as the number of bidders increases.

Kagel and Levin conducted another experiment in 1986 in which they found that the winner's curse was more likely to occur in large groups than in small groups because the subjects bid more aggressively in large groups. These results were shown for the first price auction and the second price auction .

If the bidder is not fully liable for the consequences , it can be rational to provoke losses and place very high bids. If renegotiations are possible and profits can be generated through other services, this effect can also occur.

Empiricism

Below are examples of fields in which the winner's curse was and is being observed.

Oil field auctions

The oil field auctions - investigated by Capen, Clapp and Campbell - have already been mentioned above. Years after Capen et al. Having presented their idea, other research teams evaluated the situation. Mead, Moseidjord and Sorensen in 1983 and Hendricks, Porter and Boudreau in 1987 provide empirical evidence for the curse of the winner at the oil field auctions in the Gulf of Mexico.

UMTS auction in Germany

In 2000, UMTS licenses were auctioned in Germany at a price of DM 100 billion . The situation of the bidders was characterized by a high level of uncertainty about the values ​​of the auction objects. From the auctioneer's point of view, the auction should be regarded as a success, as high proceeds were achieved. Overall, it is criticized that the price that came about was too high. This would indicate that the winner's curse was occurring.

Dealing with the winner's curse

Learning effects

In 1994 Garvin and Kagel investigated how individual learning effects from bidders affect auctions in which the winner's curse occurs. Experiments have shown that the winner's curse is more pronounced with inexperienced than experienced bidders. The experienced bidders adapt and thus overcome the worst effects of the phenomenon and consistently make profits. Based on these statements, Garvin and Kagel investigated what underlies learning. They came to the following results: Compared to inexperienced bidders, the curse of the winner is clearly reduced in frequency and severity among experienced bidders. Experienced bidders bid significantly lower than inexperienced bidders. Aggressive bidding by inexperienced bidders ensures that they lose more than 50% of all auction periods. Experienced bidders, on the other hand, achieve profits in 58.4% of all auction periods. With experienced bidders this may be a. therefore better, since more aggressive bidders are more likely to leave the auction and thus a selection will take place. The selection consists in the fact that aggressive bidders are already eliminated in the course of their inexperienced stage. This leaves the non-aggressive bidders. In addition, these have learning effects over the periods. The better result with experienced bidders is therefore not only due to learning effects, but also to the selection described. Garvin and Kagel also found that wins have no impact on a bidder's bid, but actual losses would significantly reduce the bid. Learning can also be done by observing losses at other bidders.

Auctioneer's strategy

The auctioneer benefits from the winner's curse - assuming he is acting as a profit maximizer - as he receives a higher return from the auction. However, if the bidder and the auctioneer have a long-term business relationship, it can be argued that the auctioneer has an interest in the influences that determined the result of the auction. So to reduce the curse of the winner, all bidders can be provided with the same information. The application of the English auction gives the bidders the opportunity to see when the other bidders will exit the auction. If many bidders get out before their own bid is reached, they can question whether they have overestimated the value of the good.

In the UK auction, it is not critical if a bidder is overly optimistic. This will only pay insignificantly more than the second highest bid. From two overly optimistic bidders, however, the winner can be curse. When an auction is held with closed bids, an optimistic bidder is enough to bring out the winner's curse. This would win by a large margin to the other bidders. The auctioneer can therefore influence the probability of a curse on the winner by choosing the type of auction.

Strategy of the bidders

How could a bidder take advantage of the winner's curse, for example in an oil auction? If the bidder can optimally reduce his bid, he can avoid paying too much or even choose not to bid. However, this solution is not very satisfactory. The bidder could share his knowledge with the other bidders and convince them to reduce their bids. If the other bidders can be convinced, this strategy can be profitable for the bidders. A possible profit could show itself in the form of learning effects with regard to the information situation.

In order to be successful as a bidder in an auction in which the curse of the winner occurs, not only the value of the goods to be auctioned has to be estimated. It must also be borne in mind that this estimate and that of all other bidders can be incorrect. Oil companies have responded to the findings on the winner's curse and are removing the winner's curse from their commandments. To do this, they take their highest bids and subtract the expected error - standard deviations of the estimated values ​​are used, which they can calculate based on the historical data available to them. Other areas of the economy often have a harder time if data is not available on this scale.

The loser's curse

The curse of the loser can occur when bidders bid below the optimal bid due to a lack of information. So the loser's curse is exactly the opposite of the winner's curse. In laboratory situations it is possible to create the winner's curse and the loser's curse in such a way that they neutralize each other or one bias dominates the other.

literature

  • Bazerman, MH / Samuelson, WF (1983): I Won the Auction but Don't Want the Prize , The Journal of Conflict Resolution , Vol. 27, No. 4, pp. 618-634
  • Capen, EC / Clapp, RV / Campbell, WM (1971): Competitive Bidding in High-Risk Situations , Journal of Petroleum Technology , Vol. 23, pp. 641-653
  • Clement, R. / Schreiber, D. (2013): Internet economy - Basics and case studies of the networked economy , 2nd edition, Springer-Verlag: Berlin, Heidelberg
  • Garvin, S. / Kagel, JH (1994): Learning in common value auctions: Some initial observations , Journal of Economic Behavior and Organization , Vol. 25, No. 3, pp. 351-372
  • Holt, CA / Sherman, R. (1994): The Loser's Curse , The American Economic Review , Vol. 84, No. 3, pp. 642-652
  • Lindstädt, H. (2001): The auction of the German UMTS licenses - An economic analysis of the bidder behavior, HHL working paper, No. 42, Leipzig University of Applied Sciences : Leipzig
  • Pindyck, R. / Rubinfeld, D. (2009): Microeconomics , 7th edition, Pearson Studium: Munich
  • Ruhnau, T. (2012): Auctions with revenue management in the automotive industry - hybrid distribution for self-regulating vehicle allocation , Springer Fachmedien: Wiesbaden
  • Sieg, G. (2007): Economics - With current case studies , Oldenbourg Wissenschaftsverlag GmbH: Munich
  • Thaler, RH (1988): Anomalies - The Winner's Curse , Journal of Economic Perspectives , Vol. 2, No. 1, pp. 191-202

Individual evidence

  1. ↑ a b c d e Pindyck, R. / Rubinfeld, D. (2009): Mikroökonomie , 7th edition, Pearson Education, Munich, pp 665/666
  2. ↑ a b c d e Thaler, RH (1988): Anomalies - The Winner's Curse , Journal of Economic Perspectives , Vol. 2, No. 1, pp. 192/193
  3. ↑ Capen, EC / Clapp, RV / Campbell, WM (1971): Competitive Bidding in High-Risk Situations , Journal of Petroleum Technology , Vol. 23, pp. 641ff.
  4. ↑ a b Bazerman, MH / Samuelson, WF (1983): I Won the Auction but Don't Want the Prize , The Journal of Conflict Resolution , Vol. 27, No. 4, pp. 622f.
  5. ^ A b Bazerman, MH / Samuelson, WF (1983): I Won the Auction but Don't Want the Prize , The Journal of Conflict Resolution , Vol. 27, No. 4, p. 620
  6. ^ Bazerman, MH / Samuelson, WF (1983): I Won the Auction but Don't Want the Prize , The Journal of Conflict Resolution , Vol. 27, No. 4, p. 631
  7. ^ Thaler, RH (1988): Anomalies - The Winner's Curse , Journal of Economic Perspectives , Vol. 2, No. 1, p. 195
  8. ^ Sieg, G. (2007): Economics - With current case studies , Oldenbourg Wissenschaftsverlag GmbH: Munich, p. 157
  9. ↑ Thaler, RH (1988): Anomalies - The Winner's Curse , Journal of Economic Perspectives , Vol. 2, No. 1, pp. 197f.
  10. ↑ Lindstädt, H. (2001): The auction of the German UMTS licenses - An economic analysis of bidder behavior, HHL working paper, No. 42, Leipzig University of Applied Sciences : Leipzig, p. 1ff.
  11. ↑ Garvin, S. / Kagel, JH (1994): Learning in common value auctions: Some initial observations , Journal of Economic Behavior and Organization , Vol. 25, No. 3, pp. 352ff.
  12. ↑ Ruhnau, T. (2012): Auctions with Revenue Management in the Automotive Industry - Hybrid Distribution for Self-Regulating Vehicle Allocation , Springer Fachmedien: Wiesbaden, p. 31f.
  13. ↑ Thaler, RH (1988): Anomalies - The Winner's Curse , Journal of Economic Perspectives , Vol. 2, No. 1, p. 200
  14. ↑ Holt, CA / Sherman, R. (1994): The Loser's Curse , The American Economic Review , Vol. 84, No. 3, p. 642