One day, possibly soon, someone will likely receive a Nobel prize for auction design. Amend that: another Nobel Prize. William Vickrey received one in 1996 for work in this area, and in 2007 Leonid Hurwicz, Eric Maskin and Roger Myerson were honored for related work. But this field is smokin’; it’s time for another.
One problem that has been solved is how to auction related commodities, like swaths of electromagnetic spectrum. Different swaths frequently have different values and the values are interrelated. A series of separate auctions would be inefficient because bidders can typically exercise more market power in separate auctions. Furthermore, the amount they demand of one auctioned good will typically be affected by the price of others. The trick, it would appear, is to auction the goods together in a way that allows the bidders to get the bundles of goods they most want.
Paul Milgrom and Robert Wilson of Stanford, along with Preston McAfee of Yahoo! came up with a “simultaneous ascending auction” that appears to work quite well for selling such bundles. Buyers are allowed to bid on parcels of their own design until the markets clear.
But there are defects to this approach. First, the resulting auctions are vulnerable to collusion by the bidders. Second, the auctioneer has to decide exactly how much she wants to sell before she knows the prices the objects will sell for. Third, and crucially, these auctions may take numerous rounds, and in some markets – for example, liquid financial assets – values can change dramatically in the space of minutes.
Paul Klemperer of Oxford, for his part, has come up with remedies for these defects of the multi-period auction while preserving its most attractive features. The Bank of England asked him for a design that would allow it to auction loans linked to varying qualities of collateral – and to manage the process in a very short time-frame. The goal: to be able to inject liquidity into the banking system very rapidly, but in a manner that channeled resources to those who valued them most.
Klemperer’s approach, which has already been successfully used by the Bank, could have important applications elsewhere; electricity markets come to mind.
Auction theory, like a lot of economic theory, has often amounted to a dazzling display of intellectual creativity to no practical end. Klemperer and company are showing that the payoff is just around the corner.
