Jean-François Richard

Contact

Posvar 4917
Pittsburgh, PA 15260
412-648-1732

About Jean-François Richard

Jean-Francois Richard is a Distinguished University Professor of Economics. He holds a secondary appointment in the department of Statistics. His research interests include econometric modeling, Bayesian methods, time series, empirical game theoretic models (auctions, collusion) and, over the last ten years, computational methods relying upon Monte Carlo simulations : Efficient Importance Sampling and applications to high-dimensional latent variable models such as stochastic volatility models and panels with spatial and temporal correlations.

Detailed research interests, courses taught and selected publications are also available, as is his contact information.

Research

Numerical Integration, with an early emphasis on poly-t densities and, more recently, on Efficient Importance Sampling for Likelihood based inference in high-dimensional latent processes with non-Gaussian observable variables such as: stochastic volatility, probit and count models, spatial models (including spatial and temporal correlation).

Econometric Modeling of Time series, where Richard contributed to the development of important modeling concepts such as exogeneity and encompassing. Together with David Hendry, Richard also developed a set of internally consistent criteria for the validation of econometric time series models.

Empirical Auction Models, where Richard contributed to the analysis of collusive behavior in auctions and procurements, the development of estimation techniques for empirical game theoretic models and of numerical techniques for approximating equilibrium solutions to analytically intractable games such as asymmetric first price auctions.

Empirical Applications using some of the above methods in a variety of areas such as money demand, Allais paradox, procurements (DOD, US Forest Service, school milk programs), stochastic volatility, winner’s curse, legal systems and economic development, local taxation, GDP forecasting and urban crimes.

Research Interest

Bayesian Methods, with emphasis on Simultaneous Equation Models, Errors-in-Variables models, models with autoregressive errors and inference on exogeneity.

Degrees

  • "License" in Physics, University of Louvain, 1965
  • "License et maitrise" in Economics, University of Louvain, 1968
  • PhD in Economics, University of Louvain, 1973

Awards

  • W. Hallam Tuck Fellow, University of Chicago, 1968-1969
  • “Prix des Alumni de la Fondation Universitaire”, 1979 (attributed every five years to a Belgian economist under 36)
  • Fellow of the Econometric Society since 1980
  • Fellow of the Journal of Econometrics since 2007

Courses Taught

ECON 2150, General Econometrics: This is an intermediate level graduate course in Econometrics. Covered topics include: structural change, specification and model selection, systems of regression equations, nonlinear regressions, instrumental variables, simultaneous equation models, semi-parametric and non- parametric estimation, minimum distance estimation, generalized method of moments, maximum likelihood and testing principles.

ECON 2260, Advanced Econometrics I: This graduate course consists of an introduction to the econometric analysis of time series. Following a review of basic concepts (estimation, testing and numerical optimization), we first analyze reduced form models: stationary autoregressive moving average processes. Next, we discuss structural models with emphasis on unit roots, cointegration and Error Correction Mechanisms. Finally, we discuss dynamic state-space models: linear Gaussian models and the Kalman Filter, non-linear and/or non-Gaussian models and particle filters, together with extensions.

ECON 0170, Introduction to Mathematical Modeling in Social Sciences: This in an undergraduate course that is not regularly offered. It aims at awakening students’ interest in the use of a few simple and yet powerful mathematical techniques in the social sciences. It is largely non-calculus based and has limited overlap with more traditional mathematics course offerings. It is mostly self-contained and has no prerequisites. It can be taken by students at any time during their undergraduate years.

Publications