Job market candidate
Tel. +34 93 542 2687
Available for Interviews at
Simposio de la Asociación Española de Economía (SAEe), December 14-16, Barcelona, Spain
Allied Social Science Associations (ASSA), January 5-7, Philadelphia, US
| || |
Applied Economic Theory, Behavioral Economics, Experimental Economics
"Thompson Sampling: Endogenously Random Behavior in Games and Markets" (Job Market Paper)
Experimental studies have shown different degrees of randomness both over time and across different setups. This paper proposes Thompson sampling as a way to explain this randomness endogenously. Thompson sampling means that individuals update their subjective probability distribution of unknown parameters in a Bayesian way, make a random draw from the posterior, and then act optimally, conditional on that draw. For different experimental datasets (2x2 games and forecasting in markets), the empirical fit of Thompson sampling is compared to two other hypotheses of noisy decision-making: Bayesian learning with exogenous shocks and quantal response equilibrium (QRE). In datasets where the amount of randomness does not vary, the difference between Thompson sampling and other models is insignificant. Conversely, in datasets where the amount of randomness varies substantially, Thompson sampling provides a better fit than the other two benchmarks.
“Monetary Policy Rules in a Non-Rational World: A Macroeconomic Experiment”,(Finalist for the CEF Graduate Student Contest: prize $950)
This paper introduces a new learning-to-forecast experimental design, where subjects in a virtual New-Keynesian macroeconomy based on Woodford (2013) need to forecast individual instead of aggregate outcomes. Using this design, I analyze the impact of different interest rate rules on expectation formation and expectation-driven fluctuations. Even if the Taylor principle is fulfilled, instead of quickly converging to the REE, the experimental economy exhibits persistent purely expectation-driven fluctuations not necessarily around the REE. Only a particularly aggressive monetary authority achieves the elimination of these fluctuations and quick convergence to the REE. To shed light on how monetary policy endogenously influences behavior, this paper uses Thompson sampling: individuals use the Kalman filter to learn the underlying state but initiate their beliefs by random draws from a subjective posterior distribution.
"The Curse of a Simplified Macroeconomic Model – An Experimental Study" A key question in economic research and policymaking is to what extent one can use simplified models. This paper addresses this question by comparing the results of two learning-to-forecast laboratory experiments: one design based on the reduced form of a New-Keynesian model; and one design closer to the model's microfoundations based on the structural form of a New-Keynesian model. While the experimental economy based on the reduced form converges quickly to the rational expectations equilibrium (REE), the experimental economy based on the structural form exhibits slow convergence and fluctuations not necessarily around the REE. This paper explains the differences using Thompson sampling, a learning approach of endogenously noisy decision-making.
"Noisy Bayesian Learning in an Oligopoly Newsvendor Market with Demand Inertia"(with Rosemarie Nagel and Nicolaas J. Vriend)
This paper shows that Thompson sampling can be applied to explain the learning dynamics in an oligopoly market game. While in the standard adaptive learning model by Marcet and Sargent (1989) and Evans and Honkapohja (2001) fluctuations are frequently created by exogenous noise, we follow the proposal by Mauersberger (2017) to introduce endogenous noise in the cognitive process by using Thompson sampling. We use the experimental data of Nagel and Vriend (1999), in which subjects are firms in an oligopoly market with imperfect information about their environment. Subjects need to make two decisions: how much to produce of a perishable good and the number of signals (advertisement) sent to their consumers. Our simulations show that Thompson sampling can explain several features of aggregate and individual data such as the convergence speed to the equilibrium and the level of fluctuations.
"Heterogeneity in Microeconomic Experiments" (with Rosemarie Nagel)
Prepared for the Handbook of Computational Economics, Volume 4, Heterogeneous Agents. Editors: Cars Hommes and Blake LeBaron (soon to be posted)
In this chapter on experiments in microeconomics, we discuss heterogeneity through the interplay of economic models, theoretical solutions, experimental and behavioral economics, psychology, and neuroeconomics in laboratory and field experiments. First, we present the different areas within experimental economics guided by several surveys. We then suggest a classification and structure of archetypal games and markets through the Keynesian Beauty contest formulation and discuss the main experimental findings. We provide an overview of experimental methods including for example the strategy method, cognitive and population measures to extract and to understand the heterogeneity of human behavior contrasted by theoretic solutions. The gap between rational choice theory and random behavior is bridged through different parsimonious boundedly rational models.
Research in Progress
"Testing Rational Inattention" (with Michael Woodford)
"An Experimental Test of Forward Guidance" (with Rosemarie Nagel, Stephanie Schmitt-Grohe and Martin Uribe)
"Playing against the Field and the “Visibility” of Mixed Strategies" (with Rosemarie Nagel, Ingrid M.T. Rohde and Shmuel Zamir)