Anthony Lee Zhang

PhD Program, Economics
PhD Program Office
Graduate School of Business
Stanford University
655 Knight Way
Stanford, CA 94305

Anthony Lee Zhang

I'm a fifth-year PhD candidate in Economics at the Stanford Graduate School of Business. I'm on the job market and will be available for interviews at the ASSA 2019 Annual Meeting.
Research Interests
Market Design
Industrial Organization
Financial Markets
Microeconomic Theory

Job Market Paper

Competition and Manipulation in Derivative Contract Markets
This paper develops metrics and methods for quantifying the manipulability of cash-settled derivative contract markets. Many derivative contracts, such as futures, options, and swaps, are settled based on price benchmarks, which are calculated based on trade prices of underlying assets. Derivative contract volume is often much larger than the volume of underlying trade used to construct price benchmarks, so these markets may be vulnerable to manipulation: contract holders may trade the underlying asset in order to move benchmarks and influence contract payoffs. I show that derivative contract markets can be much larger than underlying markets without creating large incentives for manipulation, as long as underlying markets are sufficiently competitive. I show how to estimate manipulation-induced benchmark distortions using commonly observed market data. I propose a simple manipulation index which can be used as a diagnostic metric to detect potentially manipulable contract markets, similar to the Herfindahl-Hirschman index (HHI) in antitrust. I apply my results to study contract market competitiveness using the CFTC Commitments of Traders reports, to measure the manipulability of the LBMA gold price benchmark, and to propose a less manipulable design for the CBOE Volatility Index (VIX).
Publications
Redesigning Spectrum Licenses PDF
with Paul Milgrom and Glen Weyl; Regulation, 2017, 40(3): 22–26
Radio spectrum in the US is traditionally assigned using licenses with renewable 10- or 15-year terms, to encourage costly infrastructure investments. However, such long-term licenses are difficult to reassign as more valuable uses for spectrum arise. Short-term licenses expedite reassignment of spectrum to innovate entrants, but lower incentives for long-term investment. Existing license designs navigate this tradeoff poorly; depreciating licenses, introduced in Weyl and Zhang (2018), are a better alternative.
Implementability, Walrasian Equilibria, and Efficient Matchings PDF
with Piotr Dworczak; Economics Letters, 2017, 153 pp. 57–60.
In mechanism design, implementable allocation rules are allocations of goods to an agent's types which are incentive-compatible under some transfer payments. We demonstrate an analogy between implementable alloction rules, Walrasian equilibria, and efficient matchings. Implementable allocation rules correspond to Walrasian equilibria, in an economy in which the agent's types are represented by consumers with quasilinear utility and unit demand. Due to the welfare theorems, Walrasian equilibria are equivalent to efficient matchings between consumers and goods.
Working Papers
Depreciating Licenses (with Glen Weyl, January 2018) PDF
A large body of work in economics studies optimally allocating assets using auctions; comparatively little work analyzes how to design licenses governing the assets that are auctioned. In this paper, we argue that license design faces a fundamental tradeoff. Long-term or perpetual licenses improve incentives for owners to invest to maintain and improve assets, but short-term licenses are better for allocative efficiency. We propose a new license, called the depreciating license, which improves on this tradeoff. Depreciating license owners periodically announce valuations at which they are willing to sell their licenses, and pay a percent of these valuations as license fees. This encourages reallocation while creating high and time-stationary investment incentives.
A Mechanism Design Approach to Identification and Estimation (with Brad Larsen, July 2018) PDF
In many trading games, such as auctions and bargaining, agents take actions which affect the probability that they receive a good and monetary transfer payments they make or receive. In this paper, we show that agents' choices on a menu of probabilities and transfers available in equilibrium can be used to identify agents' values in many such trading games. This "empirical menu" approach can accomodate various extensions, such as certain kinds of unobserved heterogeneity and partially observed actions. We apply these results to study bargaining efficiency, competition and surplus division in used car bargaining.
Auctions with Liquidity Subsidies (November 2018) PDF
This paper proposes liquidity subsidies for improving allocative efficiency and price discovery in multi-unit auctions. In the proposed subsidy scheme, the market administrator divides some amount of subsidy revenue between agents proportional to their marginal contribution to the slope of auction aggregate demand at the equilibrium price. These subsidies cause agents to bid more aggressively, increasing the slopes of their submitted bid curves. This decreases bid shading, increases allocative efficiency, and lowers the variance of auction prices.
Work in Progress
Thickness in the US Housing Market
with Nadia Kotova
A Machine Learning Approach to Bargaining Game Estimation
With Brad Larsen
Last Updated 1 Jan 2019