Homepage of Guillaume Chevillon

Guillaume Chevillon

Research: Economics, Analytics, Forecasting
CoDirector ESSEC | CentraleSupélec Master in Data Sciences & Business Analytics

Research

Welcome to my homepage

News

January-February 2020:

I am visiting the Department of Economics at the Business School of the University of New South Wales, Sydney. I will be giving seminars at
Feb 5, Economics Department, UNSW Business School
Feb 14, Business Analytics, Sydney University Business School
Feb 26, Macquarie University
March 4, Economics Department, Sydney University

Our paper with Anurag Banerjee and Marie Kratz "Probabilistic Forecasting of Bubbles and Flash Crashes" is now online of the website of The Econometrics Journal. https://doi.org/10.1093/ectj/utaa004
Our paper withSophocles Mavroeidis and Zhaoguo Zhan, "Robust inference in structural VARs with long-run restrictions" is now published in Econometric Theory, 36(1), pp. 86-121, 2020 (open source access).

October 2019:

Pleased to be invited to the Joint Research Centre of the European Commission in the program on Big Data & Forecasting of Economic Developments run by Sebastiano Manzan.
2020 QS ranking of masters in Business Analytics:
Our Master in Data Sciences & Business Analytics is now
#1 in Europe, #3 worldwide (after the MIT and UCLA)

May 2019:

I have a new article published in Le Monde « Il semble illusoire de contrôler a priori les outils d’intelligence artificielle car leurs conséquences sont quasi imprévisibles » where I discuss the issues surrounding scientific evaluations of the long run effects of social network algorithms and whether/how recent proposals of Public Agencies for Algorithmic Security might be feasible.

********************

About me

I am an Econometrician who works on Economics and Forecasting. I am Professor at ESSEC Business School, in the greater Paris area, and until recently, I also was a member of the Macroeconomics Center at CREST.

I am the Academic Director (for ESSEC) of the ESSEC-CentraleSupélec Master in Data Sciences & Business Analytics. Together with Nikos Paragios from CentraleSupelec, we've been running this program since we welcomed the first cohort in 2015. I'm very proud of this Master which has been very successful so far, with great students and professors involved.

-> It is the first joint master between a business school and an engineering school that was accredited by the French Ministry of Higher Education at the master level.
-> ranked #1 in Europe, #3 worldwide in 2020 QS ranking of Masters in Business Analytics

On the research side, as Economics cannot be a proper experimental science (contrary to Physics and Natural Sciences, economists cannot and will not conduct large scale experiments on economies), if we have any hope for it ever to become a proper "science" rather than a set of opinions, we need to be able to refute and reject wrong theories.

This is the purpose of econometricians: we develop tools to judge economic theories by their empirical relevance. The lack of experimentation implies that we have to resort to historical data and see what laws and principles are permanent and hidden. This is in fact a form of data sciences developed specifically with social sciences in mind.

In this context my interests have also been recently about the interplay between Data Sciences and Social Sciences since both are ultimately about understanding human behavior and how we can have a progressive modeling strategy, i.e. truly moving towards science.

More specifically my research interests lie in time series econometrics and forecasting, with a special interest in Macroeconomics (esp. Dynamics of deviations from Rational Expectations) and Finance (Forecasting of Asset Prices). I also have other work on risk premia in oil prices and the human origin of global warming.

I hold a degree of Ingénieur des Mines de Paris, an M.Phil. & a D.Phil. in Economics from the University of Oxford and have been a visiting researcher at Oxford, NYU, Brown and the NY Fed. I have also worked for a few years as an applied forecaster at OFCE (SciencesPo), in Paris. I have taught in the past at Oxford, SciencesPo, Dauphine, ENA, HEC and Orléans.

At ESSEC, I belong the the group of Statisticians & Econometricians.

contacts: chevillon@essec.edu

twitter: @gchevil

Official short bio:

Guillaume Chevillon is Professor of Econometrics at ESSEC Business School, and an associate researcher in the Macroeconomics lab at CREST, both in Paris. His field of research is econometric theory with applications in macroeconomics and finance. His first line of research consists in the study of persistence and inference in models of expectations under bounded rationality or learning. A second research interest concerns "atypical" sources of dynamic dependence and their use in modeling and estimation (near stationarity, spurious long memory, collapsing bubbles...). A third line focuses on forecasting under non-stationarity. He has published his work in, inter alia, the Journal of Econometrics, the Journal of Monetary Economics, Econometric Reviews, Econometric Theory and the Journal of Economic Dynamics and Control. Guillaume holds a D.Phil. in Economics from the University of Oxford and an MSc in Engineering from Ecole des Mines de Paris. At ESSEC, he is director of the ESSEC-Centrale Master in Data Sciences & Business Analytics. He has been a visiting scholar or professor at, inter alia, Brown, Oxford, NYU and the NY Federal Reserve. He has also taught at several French universities and Grandes Ecoles and is regularly featured in the French press and on the radio.

Older information:

December 2018:

Our paper with Sophocles Mavroeidis and Zhaoguo Zhan "Robust inference in structural VARs with long-run restrictions" was accepted for publication at Econometric Theory (open source on the Editor's website)
My opinion contribution in Libération was also published in La Libre Belgique, "Et si on réinitialisait les réseaux sociaux en 2019 ?"
I'm pleased to be serving on the program committees of the Workshop in Financial Econometrics (Nantes, 21-22 March 2019, call still open) and of the 27th Symposium of the SNDE at the Dallas Fed (March 28-29, 2019)

November 2018: I was interviewed by Le Monde about the recent good news on the Master in Data Sciences & Business Analytics, and more generally about the interplay between Engineering & Management, see the interview here.

September 2018:

I have a new article published by Libération "Et si on réinitialisait les réseaux sociaux" (English version: Should we reset social networks?)

July 2018: I'm pleased to participate in the Scientific Committee of the 10th French Econometrics Conference that will take place at Paris School of Economics on Nov 29-30, 2018. Deadline for submissions is Sept 15.

April 2018: I have two papers about different origins of long memory that are just making it to the Journal of Econometrics.

Generating Univariate Fractional Integration within a Large VAR(1) with Alain Hecq (Maastricht) and Sébastien Laurent (Aix-Marseille) will be published in May 2018.
Learning can generate long memory with Sophocles Mavroeidis (Oxford) was published in May 2017. We have a companion paper: Perpetual Learning and Apparent Long Memory (Journal of Economic Dynamics & Control, 2018).

March 2018: I'm honored to have been elected to the Executive Committee of the Society for Nonlinear Dynamics and Econometrics. The 26th annual symposium organized by Ippei Fujiwara & Yasuo Hirose at Keio University was a great success.

- I published an op-ed in Libération on 17/11/16 "Des algorithmes dangereux pour le débat démocratique" and you can see me at a roundtable with (inter alia) Axelle Lemaire, Secrétaire d'Etat pour le numérique et l'innovation at Théâtre de la Gaîté Lyrique, on 19/11. (video here)

- France Info: I no longer participate in Emmanuel Davidenkoff's series "Un jour, une question" but the topics I've covered can be found here.