Friday, February 17, 2006

A Totally Made Up Idea About Auction Psychology

I've been playing around with this idea since Property today and reading Sean's post below. The ascending 'English auction' model seems to be preferred by many auction houses over the descending 'Dutch auction' model. The intuitive explanation I've heard is that bidding on an item somehow vests the bidder with a psychological stake in winning the auction, combined with what you discussed about spiraling revaluations as each bidder's willingness to go higher leads rivals to re-evaluate the objective/common value of the item.

But what about this as an alternative explanation: The English auction doesn't lead bidders to revise their overall subjective valuation. Instead, it diverts attention away from it by redirecting their focus to the bid margins, rather than on the the total bid price, leading them to implicitly and fallaciously treat their previous bids as quasi-sunk costs. They think in terms not of how much they are willing to pay for a certain item, but how much more they are willing to pay above the already revealed preference. It's the whole "if I'm willing to pay $10,000, then really $1,000 more is not a big deal." A familiar cognitive fallacy occures when people weigh total cost of an item against the total utility of the item, rather than against the marginal utility of the item relative to what they already have (like when I buy one new computer after another, thinking about the total utility of the new computer rather than the increase in utility over my current one). What I'm describing here is the converse - in deciding to bid as the ascending auction escalates, people mistakenly weigh the total utility of the item against the marginal bid price, rather than the total price.

The presumed advantage of the Dutch auction is that it plays on people's risk aversion - once the bid-price drops below the subjective valuation of the highest potential bidder, risk aversion presumably kicks in to force the subjective valuation information out, thus maximizing sale price. (An interesting side-question here is whether the risks associated with letting the bid-price descend below subjective value are appropriately risks taken in maximizing gains from trade, or in minimizing losses since the existence of gains from trade have already been revealed to the highest potential bidder and thus the option to accept the bid and purchase the item represents quasi-possession of the gains that exist at that bidprice.) However, the descending system keeps bidders directly focused on total bid-price, rather than the margin. Thus, it might get the highest valuer to reveal their true preference, but it won't get them to change it or disregard it.

The English system is hypothesized to get people to alter their valuation in the course of bidding. I've floated the idea that they lose sight of it by becoming focused on the margins rather than the total price. It may be a combination. Both would result in similar behavior, though my suggestion is perhaps less likely to explain behavior of institutional bidders. I'm thinking more along the lines of the people sitting in the chairs at Sotheby's. It would be fascinating to try to isolate the causes empirically.


Post a Comment

<< Home