My research focuses on rational inattention and applied time series econometrics.
- Published papers
- Working papers
- Other research
Choosing what to pay attention to (2021)
Theoretical Economics, forthcoming
This paper studies static rational inattention problems with multiple actions and multiple shocks. We solve for the optimal signals chosen by agents and provide tools to interpret information processing. By relaxing restrictive assumptions previously used to gain tractability, we allow agents more latitude to choose what to pay attention to. Our applications examine the pricing problem of a monopolist who sells in multiple markets and the portfolio problem of an investor who can invest in multiple assets. The more general models that our methods allow us to solve yield new results. We show conditions under which the multimarket monopolist would optimally choose a uniform pricing strategy, and we show how optimal information processing by rationally inattentive investors can be interpreted as learning about the Sharpe ratio of a diversified portfolio.
SciPy 1.0: fundamental algorithms for scientific computing in Python (2020)
Nature methods, 2021, with Pauli Virtanen, Ralf Gommers, and 108 others
SciPy is an open-source scientific computing library for the Python programming language. Since its initial release in 2001, SciPy has become a de facto standard for leveraging scientific algorithms in Python, with over 600 unique code contributors, thousands of dependent packages, over 100,000 dependent repositories and millions of downloads per year. In this work, we provide an overview of the capabilities and development practices of SciPy 1.0 and highlight some recent technical developments.
Forecasting US inflation in real time (2021)
with Kirstin Hubrich
We perform a real-time forecasting exercise for US inflation, investigating whether and how additional information–additional macroeconomic variables, expert judgment, or forecast combination–can improve forecast accuracy and robustness. In our analysis we consider the pre-pandemic period including the Global Financial Crisis and the following expansion–the longest on record–featuring unemployment that fell to a rate not seen for nearly sixty years. Distinguishing features of our study include the use of published Federal Reserve Board staff forecasts contained in Tealbooks and a focus on forecasting performance before, during, and after the Global Financial Crisis, with relevance also for the current crisis and beyond. We find that while simple models remain hard to beat, the additional information that we consider can improve forecasts, especially in the post-crisis period. Our results show that (1) forecast combination approaches improve forecast accuracy over simpler models and robustify against bad forecasts, a particularly relevant feature in the current environment; (2) aggregating forecasts of inflation components can improve performance compared to forecasting the aggregate directly; (3) judgmental forecasts, which likely incorporate larger and more timely datasets, provide improved forecasts at short horizons.
Mechanics of static quadratic Gaussian rational inattention tracking problems (2018)
This paper presents a general framework for constructing and solving the multivariate static linear quadratic Gaussian (LQG) rational inattention tracking problem. We interpret the nature of the solution and the implied action of the agent, and we construct representations that formalize how the agent processes data. We apply our approach to a price-setting problem and a portfolio choice problem - two popular rational inattention models found in the literature for which simplifying assumptions have thus far been required to produce a tractable model. In contrast to prior results, which have been limited to cases that restrict the number of underlying shocks or their correlation structure, we present general solutions. In each case, we show that imposing such restrictions impacts the form and interpretation of solutions and implies suboptimal decision-making by agents.
Mechanics of linear quadratic Gaussian rational inattention tracking problems (2017)
Note: This is an previous version of the working paper Mechanics of static quadratic Gaussian rational inattention tracking problems, although it contains some sections not included there. In particular, it expands on the dynamic case and provides more detail on the equilibrium solution to the rational inattetion price-setting problem.
This paper presents a general framework for constructing and solving the multivariate static linear quadratic Gaussian (LQG) rational inattention tracking problem. We interpret the nature of the solution and the implied action of the agent, and we construct representations that formalize how the agent processes data. We apply this infrastructure to the rational inattention price-setting problem, confirming the result that a conditional response to economics shocks is possible, but casting doubt on a common assumption made in the literature. We show that multiple equilibria and a social cost of increased attention can arise in these models. We consider the extension to the dynamic problem and provide an approximate solution method that achieves low approximation error for many applications found in the LQG rational inattention literature.
Estimating time series models by state space methods in Python: Statsmodels (2015)
This paper describes an object oriented approach to the estimation of time series models using state space methods and presents an implementation in the Python programming language. This approach at once allows for fast computation, a variety of out-of-the-box features, and easy extensibility. We show how to construct a custom state space model, retrieve filtered and smoothed estimates of the unobserved state, and perform parameter estimation using classical and Bayesian methods. The mapping from theory to implementation is presented explicitly and is illustrated at each step by the development of three example models: an ARMA(1,1) model, the local level model, and a simple real business cycle macroeconomic model. Finally, four fully implemented time series models are presented: SARIMAX, VARMAX, unobserved components, and dynamic factor models. These models can immediately be applied by users.
Index of Common Inflation Expectations (2020)
with Hie Joo Ahn
This note develops a new index of common inflation expectations that summarizes the comovement of various inflation expectation indicators based on a dynamic factor model. This index suggests that inflation expectations were relatively stable between 1999 and 2012, and then experienced a downward shift that persisted, despite some fluctuations, at least through the beginning of the COVID-19 pandemic in early 2020. Since then it has successfully captured pandemic-driven concerns, first falling on fears of a prolonged recession, and then rising as the US economy has recovered and anxiety about inflation has grown.