yield very low rents. The most important bidder is a public company. Therefore it has less incentive to collude, and its bids would limit the profits from colluding. The other fact that supports competitive bidding is the bidders’ very low accounting profits in this market (Valkama and Finkkilä 2003).
Another important consideration is that the pseudo-cost estimation is based on as- suming Bayesian Nash equilibrium behavior. As HHS admit, this is not an innocuous assumption, although they claim that FPSB auctions seem particularly well suited to this approach. Experimental results by Bajari and Hortacsu (2005) encourage the use of these structural econometric tools. In my data, the players are experienced compa- nies bidding for large contracts. Thus the assumption of rational behavior is plausible. Kagel and Levin (1986) find in their experiments that bidders learn equilibrium behav- ior only through experience. In my data, the standardized variances of bids were much larger in the first tranche than in later ones. This could be due to learning. Because of the small size of the data set, I do not remove the first tranche from the data.
yield very low rents. The most important bidder is a public company. Therefore it has less incentive to collude, and its bids would limit the profits from colluding. The other fact that supports competitive bidding is the bidders’ very low accounting profits in this market (Valkama and Finkkilä 2003).Another important consideration is that the pseudo-cost estimation is based on as- suming Bayesian Nash equilibrium behavior. As HHS admit, this is not an innocuous assumption, although they claim that FPSB auctions seem particularly well suited to this approach. Experimental results by Bajari and Hortacsu (2005) encourage the use of these structural econometric tools. In my data, the players are experienced compa- nies bidding for large contracts. Thus the assumption of rational behavior is plausible. Kagel and Levin (1986) find in their experiments that bidders learn equilibrium behav- ior only through experience. In my data, the standardized variances of bids were much larger in the first tranche than in later ones. This could be due to learning. Because of the small size of the data set, I do not remove the first tranche from the data.
การแปล กรุณารอสักครู่..
