The Cost of Delay: Evidence from the Ethereum Transaction Fee Market
by Yinhong “William” Zhao
Abstract
Delaying a financial transaction can be costly, but the cost of delay is difficult to estimate in traditional
finance. I exploit the unique data offering and market design of the Ethereum blockchain to estimate the
cost of delaying financial transactions in decentralized finance (DeFi). I construct a dynamic auction
model for the Ethereum transaction fee market that relates users’ optimal transaction fee bids to their delay
cost functions and network conditions, and I structurally estimate the delay cost functions for different
users and transaction types. The average cost of delaying a transaction by one minute is 8.78 US dollars,
but the distribution of delay costs is highly skewed to the right. Delay costs are higher for complex
transactions and users who trade more frequently. I estimate that welfare loss due to network delay on
Ethereum was 14.03 million US dollars per day in July 2021, and I apply the delay cost estimates to
evaluate the welfare losses under alternative transaction fee mechanisms.
Campbell Harvey, Faculty Advisor
Michelle Connolly, Faculty Advisor
JEL Codes: D44; G10; L17;
Incentive Programs for Neglected Diseases
By Pranav Ganapathy
We propose and evaluate an auction mechanism for the priority review voucher program. The 2007 voucher program rewards drug developers for regulatory approval of novel treatments for neglected tropical diseases. Previous papers have proposed auctioning vouchers for the priority review voucher program but have offered neither a mathematical model nor a framework. We present a mechanism design problem with one pharmaceutical company producing one drug for a neglected tropical disease. The mechanism that maximizes the regulator’s expected surplus is a take-it-or-leave-it offer, with three different offers based on low, intermediate, and high neglected disease burdens. We demonstrate how mechanism design can be applied to settings in which the buyer pays for public access to a product with regulatory speed. Finally, this paper may be useful to policymakers seeking to improve access to voucher drugs through modifications of the program.
Advisors: Professor David Ridley, Professor Giuseppe Lopomo, Professor Michelle Connolly| JEL Codes: I1, D44, D82
Pricing and Pack Size Brand, Quantity, and Cost Considerations in Pricing Multipacks of Toothpaste
By Stephanie Wiehe
The US market for toothpaste, like many other goods, is shifting towards selling
in bulk. Multipacks of toothpaste require quantity discounts to incentivize consumers, making buying in bulk a great deal for the savings-minded toothpaste-shopper. It is more difficult to understand, however, producers’ willingness to sell multipacks of toothpaste, when margins are necessarily slimmer than single tubes due to quantity discounts. This paper explores the consumer’s decision in purchasing toothpaste as an interaction between savings on price and inventory considerations, like shopping and carrying costs. My model combines aspects of prior works on second degree price discrimination and quantity discounts with alterations to fit the intricacies of the market for toothpaste. The model’s predictions support the possibility of pack size as a tool for second degree price discrimination as shopping and carrying costs constitute two markets with different price elasticities of demand for single and multipacks of toothpaste. This work adds to the existing literature on storable goods and non-linear pricing and brings a new economics-based approach to a question faced by toothpaste producers.
Advisors: Professor Allan Collard-Wexler, | JEL Codes: L11; L42; D4
Investigating a Case of Alleged Collusion in Michigan Public Oil and Gas Lease Auctions
By Lucas Do
The state of Michigan administers oil and gas lease auctions semiannually through the Department of Natural Resources. In June 2012, the international news outlet Reuters published allegations of bid-rigging following the Department’s May 2010 auction. This paper empirically investigates the validity of Reuters’ allegations by analyzing auction bid sheets from 2008 to 2018 as well as other data reflecting market conditions over time. To this end, I first formulate a benchmark structural model of bidders’ valuations and estimate it with auction data from a period during which I assume no collusion occurred. Then, I extend the benchmark model by endogenizing bidders’ decision to collude. Using the extended model and the estimated benchmark parameters, I apply the simulated method of moments to solve for the collusive probability that “best” explains the observed bids during the alleged period of collusion. After discovering strong evidence for bid-rigging, I run counterfactual simulations to estimate the revenue damage caused to the state of Michigan by this non-competitive bidding behavior. I find that the hypothetical revenue damage, summed over the entire alleged collusive period, totals over $450 million. However, although these findings lend support to Reuters’ allegations and are contrary to the Department of Justice’s conclusion in 2014 after they had probed the case, they should be approached only with caution, given the limitations of the available data on the potential bidders.
Advisors: Professor James Roberts, Professor Michelle Connolly | JEL Codes: L4, D44, L71
Evaluation of the Impact of New Rules in FCC’s Spectrum Incentive Auction
By Elizabeth Lim, Akshaya Trivedi and Frances Mitchell
On March 29, 2016, the FCC initiated its first ever two-sided spectrum auction. The auction closed approximately one year later, having repurposed a total of 84 megahertz (MHz) of spectrum. The “Incentive Auction” included three primary components: (1) a reverse auction where broadcasters bid on the price at which they would voluntarily relinquish their current spectrum usage rights, (2) a forward ascending clock auction for flexible use wireless licenses which determined the winning bids for licenses within a given geographic region, and (3) an assignment phase, where winning bidders from the forward auction participated in single-bid, second price sealed auctions to determine the exact frequencies individual licenses would be assigned within that geographic region. The reverse auction and the forward auction together constituted a “stage.” To guarantee that sufficient MHz were cleared, the auction included a “final stage rule” which, if not met, triggered a clearing of the previous stage and the start of a new stage. This rule led to a total of four stages taking place in the Incentive Auction before the final assignment phase took place. Even at first glance, the Incentive Auction is unique among FCC spectrum auctions. Here we consider the estimated true valuation for these licenses based on market conditions. We further compare these results to more recent outcomes in previous FCC spectrum auctions for wireless services to determine if this novel auction mechanism
impacted auction outcomes.
Advisor: Michelle Connolly | JEL Codes: L5, O3, K2, D44, L96
Regulatory Uncertainty: The Impact of the 2015 Open Internet Order on Broadband Infrastructure Investment
By Dane Bourcy Burkholder and Chin Jie Lim
This paper analyzes the impact of the United States Federal Communication’s (FCC) March 2015 Open Internet Order (OIO) on broadband infrastructure investment outcomes such as changes in speed of services, market entry. We find that higher broadband investment levels deter potential entrants and may weed out competition amongst incumbent ISPs from December 2014 to December 2016. The 2015 OIO appears to have negatively impacted the probability of an ISP entering a census block for the first time by 7.17% during any six-month time periods from June 2015 to December 2016 compared to the time period from June 2010 to December 2014.
Advisor: Dr. Michelle Connolly | JEL Codes: D21, D25, D42, L20, L50, L96
A Brief Review and Analysis of Spectrum Auctions in Canada
by Martínez-Cid, Wenfei Jiao, and Zeren Zhang
Abstract
We begin by explaining the importance of efficient spectrum allocation and reviewing Canada’s recent spectrum allocation history. We then use a dataset covering more than 1,200 licenses auctioned from 2001 to 2015 that seeks to account for each auction’s particular rules. Our results confirm that measures of demand such as population covered, income levels, frequency levels, bandwidth, etc. indeed drive license valuation. We also quantify the negative impact on price of setting aside particular license auctions for new entrants, suggesting that the set-aside provision constitutes an implicit subsidy for those firms.
Michelle Connolly, Faculty Advisor
JEL Codes: D44, D45, D47, L51, O33
Structural Estimation of FCC Bidder Valuation
By Renhao Tan, Zachary Lim, and Jackie Xiao
We modify a method introduced in Fox and Bajari (2013) which structurally estimates the deterministic component of bidder valuations in FCC spectrum auctions based on a pairwise stability condition: two bidders cannot exchange two licenses in a way that increases the sum of their valuations, and we apply it to C block auctions 5, 22, 35 and 58. Our modifications improve the fit of the Fox and Bajari (2013)’s estimator especially in similar auctions involving big bidders. We find that there is evidence of significant “cross-auction” complementaries between licenses sold in a particular auction and those already owned by these endowed bidders.
Advisor: Michelle Connolly | JEL Codes: D44, D45, H82, L82
BIDDING FOR PARKING: The Impact of University Affiliation on Predicting Bid Values in Dutch Auctions of On-Campus Parking Permits
By Grant Kelly
Parking is often underpriced and expanding its capacity is expensive; universities need a better way of reducing congestion outside of building costly parking garages. Demand based pricing mechanisms, such as auctions, offer a possible solution to the problem by promising to reduce parking at peak times. However, faculty, students, and staff at universities have systematically different parking needs, leading to different parking valuations. In this study, I determine the impact university affiliation has on predicting bid values cast in three Dutch Auctions of on-campus parking permits sold at Chapman University in Fall 2010. Using clustering techniques crosschecked with university demographic information to detect affiliation groups, I ran a log-linear regression, finding that university affiliation had a larger effect on bid amount than on lot location and fraction of auction duration. Generally, faculty were predicted to have higher bids whereas students were predicted to have lower bids.
Advisor: Alison Hagy, Allan Collard-Wexler, Kent Kimbrough | JEL Codes: C38, C57, D44, R4, R49 | Tagged: Auctions, Parking, University Parking, Bidder Affiliation, Dutch Auction, Clustering
Optimal Ordering in Sequential English Auctions: A Revenue-Comparison Model for 18th Century Art Auctions in London and Paris
By Amaan Mitha
We develop a model based on several auction parameters to test the widely held notion that in a sequential English auction, it is optimal for the seller to arrange the lots in order of decreasing value. We test this model against two datasets of 18th century auctions, one of various auctions from Paris and the other from Christie’s sales in London. We find that the Paris data support the claim, while the Christie’s data seem to refute the optimal strategy. We also find a rationale for bidders in the Christie’s auctions to alter their strategies, accounting for the discrepancy.
Advisor: Neil De Marchi | JEL Codes: D4, Z11 | Tagged: Auctions, English Auction, Lot Ordering, Optimal Auction Strategy, Sequential Auctions
