, a leading peer-to-peer (p2p) online lending platform, switched its selling mechanism from multiunit uniform-price descending auctions to open fixed rates at the end of 2010. We exploit the policy change to study the relative transaction efficiency of the two popular selling mechanisms in general for p2p lending and other related financial products. Our analysis of Prosper's detailed listing and bid data shows that there is 1) a significant increase in its transaction efficiency associated with the policy switch; 2) strong evidence of lenders herding and strategically delaying bidding in the auctions, and, to a significantly lesser extent, under fixed rates; and 3) a U-shaped bimodal distribution of the amount of money raised per listing in Prosper's auctions but not with its fixed rates. Analysis of game-theoretic models of the two selling mechanisms shows that the interplay between lenders' herding and strategic timing of entry in the two dynamic selling mechanisms can explain the relatively higher transaction efficiency under open fixed rates, as well as the other documented patterns in bidding and funding outcomes. Implications for designing selling mechanisms for p2p lending and other related financial products are discussed.

Resubmitted to the Management Science. (with Hong Luo and Jing Xia). 
Pricing idiosyncratic products is often challenging because the seller, ex ante, lacks information about the demand for individual items. This paper develops a model of dynamic pricing for idiosyncratic products that features the optimal stopping structure and a seller that learns about item-specific demand through the selling process. The model is estimated using novel panel data of a leading used-car dealership. Policy experiments are conducted to quantify the value of the demand information that the dealer obtains through the initial assessment and subsequent learning in the selling process. With the dealer’s average net profit per car in the estimation sample being around $1150, the initial assessment is worth around $101, and the subsequent learning in the selling process helps improve the dealer’s profit by at least $269. These estimates suggest a potentially high return to taking the “information-based” approach to pricing idiosyncratic products. 

Estimating Production Functions with Robustness Against Errors in the Proxy Variables, new version 2017. (with Yingyao Hu and Yuya Sasaki). under second-round review at the Review of Economic Studies
This paper proposes a new approach to the identification and estimation of production functions. It extends the literature on structural estimation of production functions, started by the seminal work of Olley and Pakes (1996), by relaxing the scalar-unobservable assumption about the proxy variables. The key additional assumption needed in the identification argument is the existence of two conditionally independent proxy variables (e.g. the investment and the material input). The assumption seems reasonable in some important cases. The proposed estimation method is straightforward to apply, and can be conveniently augmented to deal with measurement errors in the input variables.

Expectation-Based Reference-Dependent Preferences: Evidence from the Used Car Retail Market, revise and resubmit, at Marketing Science, 2016. (with Haiyan Liu).
This paper empirically tests the theory of expectation-based reference-dependent preferences in the setting of consumers shopping for used cars. A unique feature of the paper’s empirical strategy is that it uses a panel dataset on the non-negotiable daily listing prices for cars advertised online by a large national dealership to pin down the prices that car buyers expected and the actual prices. The paper finds that an unanticipated price in- crease (drop) significantly lowers (increases) the car’s daily sale probability; and that the asymmetry in the reference-price effects is driven partly by price decreases being bad news for cars’ unobserved quality. 

Short-Run Needs and Long-Term Goals: A Dynamic Model of Thirst Management (with Ahmed Khwaja and K. Sudhir), Marketing Science, 2015.
Beverage consumption occurs many times a day in response to a variety of needs that change throughout the day. In making their choices, consumers self-regulate their consumption by managing short run needs (e.g., hydration and mood pickup) with long-term goals (e.g., health). Using unique intra-day beverage consumption, activity and psychological needs data, we develop and estimate a model of high frequency consumption choices that accounts for both intra-day changes in short run needs and individual level unobserved heterogeneity in the degree of self-regulation. A novel feature of the model is that it allows for dynamics of consumption and stockpiling at the level of product attributes. The model is used to evaluate introduction of new products in the beverage category and gain insight into the linkage between self-regulation and excess consumption. Broadly, the modeling framework of balancing short run needs with long-term goals has wide ranging applications in choices where long term effects are gradual (e.g., nutrition, exercise, smoking and preventive health care).

Is Advertising Informative? Evidence from Contraindicated Drug Prescriptions, new version, 2017. (with Matt Shum and Wei Tan). revise and resubmit at Quantitative Marketing and Economics.
Crestor, an important but controversial cholesterol-lowering drug, is contraindicated for use by senior and Asian patients. In this paper, we exploit this fact and a unique doctor-level prescription and advertising (detailing, to be exact) exposure data for statin drugs to examine the hypothesis of informative advertising. Our tests are based on a simple empirical learning model in which a doctor learns about the match quality of Crestor for different types of patients through advertising signals. We find strong evidence for the informative-advertising hypothesis: The match quality signaled by doctor-level advertising for contraindicated patients is significantly inferior to that signaled for the other patients. Our results are robust to doctor-level advertising being correlated with doctor-specific unobserved factors and/or differential trend in individual doctors' attitude towards Crestor.

When to Haggle, When to Hold Firm? Lessons from the Used Car Retail Market. new version 2016. Under review at Marketing Science.
Though haggling has been the conventional way for auto retailers to sell cars, the last two decades have witnessed the systematic adoption of no-haggle prices by many large dealerships, including the largest new car and used car dealership chains. This paper develops a structural empirical model to estimate sellers' profits under posted price and haggling, and investigates how market conditions affect sellers' optimal pricing formats. The model incorporates a simple class of bargaining mechanisms into a standard random-coefficient discrete choice model. With the extension, the product-level demand system is estimated using data with only list prices, and the unobserved price discounts are also recovered in the estimation. The counterfactual experiments yield a few interesting findings. First, dealers' adopted pricing formats seem superior to the alternative ones. Second, dealers enjoying larger market power through vertical differentiation and carrying a large number of models are more likely to have posted price as their optimal pricing format.

What Makes Insurance Companies Voluntarily Share Their Proprietary Customer Information. 2011. 
This paper provides a novel game-theoretic explanation for why voluntary information-sharing arrangements in the insurance and banking industry can be self-enforcing---namely, why the information exchanges, organized as independent for-profit corporations, can count on insurance companies and banks to continue to report their private customer information to them.

Work in Progress 
The Voice of Customers and Productivity Dynamics of Service Employees, (with K. Sudhir)

Dynamic Cross-selling by Multitasking Service Agents, (with K. Sudhir).

Semi-parametric Estimation of the Stochastic Utility Model of Choice: the Case with Dynamic Adverse Selection, Preliminary draft, (with Jeremy Fox and Haiyan Liu)

Information Disclosure and Observation Learning: Evidence from CarFax Reports, Preliminary draft, (with Haiyan Liu and Hong Luo)