A Sequence of Essays on Sequences of Auctions

Essay I (with Gagan Ghosh and Heng Liu). The existence of declining prices in sequential auctions is a well-documented empirical pattern. Three explanations that can explain the puzzle are bidders being risk averse, loss averse, or ambiguity averse. We use a data set on bids and prices from sequential auctions of train tickets to confirm the existence of declining prices. We further document bidder behavior that is inconsistent with bidders being risk averse or bidders being loss averse.Essay II (with Gagan Ghosh and Heng Liu). We build on the findings in the first essay by studying a sequential second price auction model with ambiguity averse bidders. The studied mechanism closely resembles the one used to sell the train tickets. We provide an identification result that enables us to estimate the primitives of the model. Using the train ticket bidding data we then recover both the distribution of valuations and bidders' "worst-case'' beliefs. We perform policy experiments using the recovered objects showing that the seller would have lost 18% of the revenues if all ambiguity was removed. On the other hand, the seller would have earned 11% more if the tickets were sold using first price auctions instead of second price auctions. This is in line with our theoretical result.Essay III (with Ola Andersson and Jim Ingebretsen Carlson). We report findings from a controlled lab experiment on sequential first price auctions that are consistent with the declining price anomaly in sequential auctions. We find that prices are higher in the first than in the second auction both in a Risk treatment, where bidders know the distribution of other bidders' valuations, and in an Ambiguity treatment, where bidders do not know the distribution of other bidders' valuations. Essay IV. I derive a structural test that can be implemented using bidding data from sequential auctions to evaluate if bidders behave as if they are price risk averse. The test is fully non-parametric. I show in a Monte-Carlo simulation that the test can be used to differentiate between risk neutral and price risk averse bidders