مشوق های کیفیت خدمات فروشندگان در مزایده انتشار تجارت سازمان حفاظت محیط زیست

Cason (1993, J. Environ. Econom. Management25, 177–195, doi:10.1006/jeem 1993.1041) argued that the auction which the EPA used in order to start the market for sulfur allowances may reduce the efficiency of the market since it gives sellers an incentive to understate their valuation. In this paper we show that the sellers' incentives are even more perverse than Cason suggested when we take into account that sellers can also submit a bid. We show that sellers have an incentive to set their asking price equal to 0 while simultaneously hedging their bets by submitting a positive bid.

Since Dales 4 , economists have recommended tradeable permits as an efficient instrument of environmental policy. However, the development of this instrument in policy practice has been slow. In the 1970s emissions trading rules emerged in the U.S. Because there was no conscious design, the rules were unclear and Ž . controversial and required a lot of government intervention Liroff 11 . In 1990, the first large-scale, consciously designed emission trading scheme was introduced in the U.S. It was applied to the SO emissions of electric utilities. 1 In 2 Ž. Phase I 1995 2000 a limited number of electric utilities participated in the Ž. scheme. In Phase II from 2000 on all electric utilities participate. Sulfur allowances can be traded in two different ways. One way is to trade privately between utilities, possibly with the intermediation of a broker. By now, the lion’s share of allowances is traded in this way. The other trading option is the annual auction in March, organized by the EPA. This auction was first held in Ž. 1993. At the auction the EPA sells the small part 2.8% of the total amount ofallowances that is not grandfathered directly to the utilities. The revenues of the auction are distributed among the allowance holders. Electric utilities and other interested parties can not only bid at the auction but also offer allowances for sale. The way in which the auction is conducted is unique. 2 The bids are ranked from high to low. The allowances from the EPA are sold to the highest bidders. The other suppliers are ranked according to their asking price, from low to high. The lowest asker is matched to the highest remaining bidder, etc., as long as the asking price is below the bid price. A successful bidder pays his bid price to the seller to whom he is matched. This was the first time that this auction design had been implemented. Also, it had not been analyzed before. Cason 1 is the only paper which provides an analysis of the auction. 3 The author claims that the market-clearing prices are too low and not all gains from trade are exhausted. The EPA’s rules can therefore generate significantly biased price signals and reduce the efficiency of the al- lowance trading market. 4 In a standard uniform price auction, all trades take place at the market-clearing price, given all bidding and asking prices submitted. In such an auction buyers have an incentive to bid slightly lower than their valuation, whereas sellers have an incentive to ask a price that is slightly higher than their Ž . valuation see for example Vickrey 15 or McAfee and McMillan 12 . In the EPA auction, however, Cason shows that a seller has an incentive to offer units at prices below her costs. Given that she will sell, the lower her asking price the higher the bid to which she will be matched and therefore the higher her expected revenue. Given that bidders still have an incentive to bid below their valuations, this results in downward-biased price signals which may reduce the efficiency of the market. Naturally, for the purposes of his paper Cason uses a model that is a highly simplified representation of the actual EPA auction. He assumes that buyers and sellers have demand, resp. supply, of only unit of the good. Recent research suggests that in multiunit auctions it is very hard to achieve efficient outcomes, Ž . even if the auction is properly designed see Klemperer 10 . Also, Cason assumes that the distribution of bids is given, and he does not take the interaction between the auction and the secondary market into account. Joskow et al . 8, pp. 12 13 argue that auction prices can only depart systematically from competitive market prices if the nonauction part of the market is seriously imperfect. Also, they argue that the presence of a private market leads to overstatement, not understatement, of reservation prices. Joskow et al . 8 criticize Cason’s assumption of private values. This paper analyzes a less obvious extension of Cason’s analysis. We allow for the possibility that sellers can also submit bids at the auction. We show that if that is the case then the equilibrium derived in Cason is no longer a Bayesian Nash equilibrium. Specifically, sellers can improve upon that equilibrium by submitting an asking price of 0 while simultaneously hedging their bets by submitting a bid equal to their valuation. This possibility is unique to this particular auction design. Therefore, we believe that it is a useful extension of Cason’s analysis and a further addition to his point. We show that the sellers’ incentives in this particular auctiondesign are even more perverse than Cason suggested. We believe our analysis does not modify any of Cason’s assumptions. Cason does not explicitly assume that sellers cannot submit a bid. It seems that he merely overlooked this option, as has everyone else analyzing this market or being active in it. 5 We now know that the auction did not live up to its expectations. Why should we then want to analyze this auction, now that it has proven to be a less relevant trading option, at least for private supply? First, there is the intellectual challenge of trying to find an equilibrium to this auction. Second, with a better understanding of how the auction works and what the incentives of the participants are, we may be able to explain why it failed to be relevant. An auction with a different design might have been a more attractive trading option and might have given a more reliable price signal to the private market. This is also relevant for future auctions of tradeable emission permits or other items. Third, we show that market partici- pants may be able to make money in the EPA auctions by using our approach rather than the equilibrium strategy derived by Cason. The remainder of this paper is organized as follows. Section 2 reviews the experiences with the EPA auction. In Section 3 we introduce the formal descrip- tion of the players and the rules of the auction. We also discuss the equilibrium derived by Cason. In this equilibrium sellers only set an asking price. In Section 4 we show that when all other sellers follow Cason’s strategy any one seller can achieve a higher payoff by submitting an asking price and a bid price, rather than only an asking price. In Section 5 we discuss the implications for a possible Bayesian Nash equilibrium in our setup. It can be shown that if an equilibrium exists then sellers either submit an asking price equal to 0 plus a positive bid or they only submit a positive asking price and no bid. Section 6 concludes the paper.

Cason 1 argued that the auction which the EPA used to start the market for sulfur allowances can lead to inefficient market outcomes. The setup of the auction gives both buyers and sellers an incentive to understate their valuation of an allowance. Therefore, equilibrium prices are too low, which leads to biased signals and reduces the efficiency of the allowance trading market. In this paper we showed that the incentives in this auction may be even more perverse than Cason suggested. In particular, we showed that sellers have an incentive to set their asking price equal to 0 while simultaneously hedging their bets by submitting a positive bid. We were not able to find the Bayesian Nash equilibrium in this auction when this possibility is taken into account. What we do know is that if an equilibrium exists sellers either set only a positive asking price, or set an asking price equal to 0 and submit a positive bid as well. As far as we know, no one has ever simultaneously submitted an asking price and a bid at the EPA auction yet: The EPA does not release the names of those who offer allowances for sale at the auction. However, even if we had this information we would face the problem that many participants use trade names or act through brokers to hide their identity. We also could not find evidence of sellers actually submitting an asking price of 0 in these auctions. Yet we do not believe there is anything in the institutional framework of these auctions that restricts sellers from following our strategy. The strategy we suggest has not been discussed in the literature about the EPA auction. Also, other literature on auctions does not study the possibility. The latter is probably due to the fact that such a strategy is only profitable in a peculiar setup like that of the EPA auction. We now know that the EPA auction was a failure, in the sense that hardly any private parties sold allowances at the auctions. As a result, it took several years before an efficient market in sulfur allowances developed. Cason 1 already predicted difficulties, arguing that equilibrium prices at the auction would be too low, making the market inefficient. Our analysis shows that when sellers can also submit bid prices their incentives in the auction are even more perverse and their optimal strategies even harder to determine, thus shedding even more doubt on the efficiency of such an auction. Therefore, it does not come as a surprise thatpotential sellers preferred to bypass the auctions altogether and waited for an efficient market to develop.