First-Price Sealed-Bid Tender Versus English Auction: Evidence from Land Sales
Posted: 29 Nov 2010
Date Written: November 29, 2010
We compare the impact on seller’s revenue when prepared vacant land is sold through two similar auction formats: the first-price sealed-bid tender and the ascending English auction. The Revenue Equivalence Theorem (Vickrey, 1961) states that a seller will obtain the same payoff regardless of the auction format applied. The four common auction formats are: first-price auction, second-price auction, ascending English auction, and descending Dutch auction. However, the Revenue Equivalence Theorem holds under the following conditions: (1) bidders are risk-neutral; (2) the independent private values assumption applies; (3) bidders are symmetric; and (4) lack of collusion amongst bidders (Maskin and Riley, 2000). If any of these assumptions is violated, the Revenue Equivalence Theorem no longer holds, leading to differences in seller’s revenue conditional on the type of auction format applied. With reference to the real estate market, any of the four assumptions can be easily violated. Hence, the question of which auction format to use in selling real estate is no longer a trivial question, especially since real estate is an extremely expensive and scarce commodity. In this paper, we analyze whether a seller will have a higher payoff under the first-price sealed-bid tender or the ascending English auction when bidders are asymmetric. The assumption that bidders have symmetric value distribution functions is rather strong, especially in land auctions. In this case, the bidders participating in the auction may range from small construction firms (‘weak’ bidders) to large development firms (‘strong’ bidders). The bidders’ value distribution function will no longer be symmetric as each type of bidders have different opportunity costs of completing the property development – bidders’ value distribution functions will be dependent on factors such as financial strength, technical capability, economies of scale, etc., which are specific to each firm. Maskin and Riley (2000) predict that when bidders are asymmetric, seller’s expected revenue in the sealed-bid tender may be higher or lower than in the English auction. In addition, we will also analyze which auction format is preferred by the different type of bidders when there are entry costs to participating in the auction. From timber auctions in the United States, Athey, Levin, and Seira (2008) find that sealed-bid auctions systematically attract more small bidders. Using a unique data for land sales that is close to a set of data generated by an economic experiment in a real-world business setting, our preliminary results show that even though the English auction generates higher land prices for the seller, statistically, there is no significant difference in land prices between the first-price sealed-bid tender and the English auction. The land sales are organized by one monopoly seller who sold about 7,000 square feet of prepared vacant land designated for residential development in three phases during the period 1993:Q4 to 1994:Q2. In the first two phases, the seller auctioned land parcels through the sealed-bid tender. In Phase Three, the seller sold 30 percent of the remaining land parcels through the ascending English auction while at the same time simultaneously selling the rest of the land parcels through the sealed-bid tender. Our empirical analysis shows that the seller’s choice of land parcels to be sold through the two auction formats is random, thus ensuring that there is no evidence of selectivity bias, at least from the seller’s choice of auction format. We also find that with entry costs, small firms prefer to participate in the English auction rather than the sealed-bid tender auction. Lastly, we are also able to observe sellers’ revenue when only one auction format was introduced and when two auction formats were introduced. Our preliminary results indicate that land prices increase in Phase Three, when the seller simultaneously uses two auction formats to sell prepared vacant land.
Keywords: First-Price Sealed-Bid Tender, Ascending English Auction, Asymmetric Bidders, Prepared Vacant Land, Entry Costs
JEL Classification: F2
Suggested Citation: Suggested Citation