Research

Experimental finance

Experimental asset markets: A survey of recent developments. [with Owen Powell] Journal of Behavioral and Experimental Finance, 12 (2016): 14-22. link to the paper

We review the latest research on experimental asset markets, where the values of the traded assets are homogeneous across all agents. Such markets have been shown to be prone to substantial mispricing, frequently in the form of a bubble-and-crash pattern. This calls into question the efficiency of such markets. The studies reviewed in this survey consider how market efficiency is affected by the structure of the market, the properties of the traded assets and the characteristics of traders.

The robustness of mispricing results in experimental asset markets. [with Owen Powell] link to WP under revision

Many experiments have been conducted on market mispricing, however there is a distinct lack of guidance over how mispricing should be measured. This raises concerns about the sensitivity of mispricing results to variations in the measurement procedure. In this paper, we investigate the robustness of previous results with respect to four variations: the choice of interval length, the use of the bid-ask spread as a price proxy, the choice of aggregation function, and controlling for observable market characteristics. While a majority of previous results are unaffected, roughly 30% do change significance.

Bubbles, experience and success. [with Dmitry Gladyrev and Owen Powell] link to most recent version submitted

Older version: The Effect of Financial Selection in Experimental Asset Markets. (2014). Vienna Economics Papers, No. 1404.

One of the most robust findings in the experimental asset market literature is the experience effect: markets are highly efficient, i.e. they generate prices close to fundamentals, as long as at least some traders are familiar with the environment. In this paper, we show that the experience effect is sensitive to the previous success of experienced traders. Specifically, markets populated by traders who were previously most and least successful are less efficient than those populated by moderately successful traders.

Group size and bubbles. [with Owen Powell] link to most recent version under revision

In this study we compare asset market outcomes produced by a small sample of large markets and a large sample of small markets, where market size is determined by the number of traders. The market outcomes that we consider include overpricing and absolute mispricing in markets with inexperienced and experienced traders, as well as the effect of experience on these measures of market efficiency. We find that even though a large sample of small markets produce noisier measures, they neither stochastically dominate the measures produced by a small sample of large markets, nor are they dominated by the latter.

Role of information in (housing) markets. [with Tatiana Filatova] work-in-progress

Tournaments

Peer effects in (sport) tournaments. [with Frantisek Kopriva] work-in-progress

In this project we measure peer effects in settings where contestants compete within a large group but directly observe only one of the other contestants - the “peer” - using data from speed-skating competitions. We compare such peer effects across tournament designs and genders.

Bounded rationality

Consumer choice of a three-part tariff as a search. link to most recent version under revision

Older version: Overcoming Consumer Biases in the Choice of Pricing Schemes: A Lab Experiment. (2010). CERGE-EI Working Paper Series No. 418.

This paper presents a lab experiment on consumer choice of a three-part tariff designed with the mouse tracking tool, Mouselab. The experiment reveals a bias towards the tariff that provides the number of ``free'' consumption units exactly equal to the expected consumption. One explanation for the observed bias comes from treating consumer choice of a three-part tariff as a search.

A lab experiment on consumer choice of a three-part tariff: Insights from process data. link to most recent version under revision

Older version: Pricing-Scheme Choice: How Process Affects Outcome. (2010). CERGE-EI Working Paper Series, No.411.

Standard IO models are based on the assumption that consumers use their future demand estimates to evaluate net utility from each offer made to them and choose the offer associated with the highest net utility. This, in turn, is based on the implicit assumption that consumers with stationary demand build correct beliefs about their future demand based on its realization in the past. This paper challenges both assumptions. Process and choice data from a lab experiment designed with the Mouselab tool are analyzed to understand how consumers with imperfect ability to infer their future demand choose a three-part tariff. A large proportion of suboptimal choices (33.3%) is observed and is partially attributed to the failure of consumers to evaluate their net utility associated with each offered tariff.

Price discrimination with boundedly rational consumers: When to make dominated offers? link to most recent version under revision

I suggest a bounded rationality explanation for the presence of dominated choice alternatives in markets. By dominated alternatives I mean those that are never chosen by fully rational consumers. These are alternatives that violate consumers' participation or incentive compatibility constraints. The model is built upon the assumption that, due to time or cognitive limitations, consumers have only a sub-set of available alternatives in their consideration sets, from which they choose the best. The necessary condition for the monopolist to benefit from offering dominated alternatives is that consumers with a higher willingness to pay are more likely to have them in their consideration sets.

Explaining the choice of telephone tariffs and health club contracts in one model. (2008). CERGE-EI Discussion Paper Series, No.2008-192.

The recent empirical evidence (Miravete (2003) and Della Vigna & Malmendier (2006)) seems to indicate certain inconsistencies in consumer behavior: people choose optimal calling plans but display a bias towards flat-rate health club contracts even though pay-per-visit would be optimal. These inconsistencies cannot be explained within a fully rational benchmark, and this stimulates a search for more suitable models. In this paper, I, first, discuss whether such concepts as hyperbolic discounting and immediate temptation can be applied for the contract choice problem. The main result is that, within both concepts, the bias towards flat-rate health club contracts is rationalized by the fact that consumers perceive a marginal utility from a visit to a gym differently at the moment of choosing a contract and at the moment of potential consumption. Another finding is that seeming mistakes in the contract choice are driven by different factors for sophisticated and naive consumers. I justify the presence of naivety in the environment with a repeated choice using the concept of limited memory and other cognitive constraints. Based on these findings, I introduce an alternative model of contract choice that is consistent with the choice of both, calling plans and health club contracts.