Ágoston Reguly

Quick bio

I am an Assistant Professor at the Institute of Economics, Corvinus University of Budapest (Corvinus), an affiliated researcher at Financial Services Innovation Lab (FSIL), part of Scheller College of Business at Georgia Institute of Technology.

My main research interests are in empirical corporate economics and econometrics. My ongoing research covers topics including mergers and acquisitions, firms’ technology adoption, policy evaluation using machine learning and traditional econometric techniques.

Research

Ongoing:

• Central Bank Communication in a Small Open (Unorthodox) Economy - joint with Marton Kovacs
• Perverse Effects of Imperfectly Enforced Price Controls - joint with Boris Knapp
• The Long-run Performance of Mergers and Acquisitions - joint with Sudheer Chava
• Discovering Heterogeneous Treatment Effects in Regression Discontinuity Designs - R&R at Journal of Applied Econometrics
• Two book chapters: Machine Learning (with László Mátyás & Essie Maasoumi) and Data Privacy (with Zsigmond Pálvölgyi), prepared for A History of Econometrics: The Challenge of Remaining Relevant (Eds: Matyas & Pirotte)

Published:

• Modelling with Sensitive Variables - joint with Felix Chan and Laszlo Matyas, published in AStA Advances in Statistical Analysis
• When and How Much Do Fixed Effects Matter? - joint with Felix Chan and Laszlo Matyas. This is a book chapter from The Econometrics of Multi-dimensional Panels (2024) (Editor: Matyas, Publisher: Springer)
• The Use of Machine Learning in Treatment Effect Estimation (2022) - joint work with Robert Lieli and Yu-Chin Hsu. This is a book chapter from Econometrics with Machine Learning (Editors: Chan F., Matyas L., Publisher: Springer).

See more on my research here.

Codes

• Beta version of multisynthdid R-package, which runs Synthetic Diff-in-Diffs with multiple outcomes. See more here multisynthdid.
• RD-tree implements the machine learning algorithm from Discovering Heterogeneous Treatment Effects in Regression Discontinuity Designs. This is a MatLab implementation. There is a python implementation by Firat Yaman, however this only implements the parametric version.
• Split-sampling method which allows data-protection for sensitive variables discussed in Modelling with Sensitive Variables

Teaching

• Econometrics I for Applied Economists (BA) at Corvinus.

• Previously I have taught various Data Analysis courses at Central European University (CEU) and led Vertically Integrated Projects a research course for Georgia Tech students.

• Learn coding & data analysis with R

  • This course material guides you from basic statistics towards machine learning methods.
  • See my course material and problem description with tasks+solutions at the following github repo: Coding for Data Analysis in R.
  • This supplements the book Data Analysis for Business, Economics, and Policy by Gábor Békés (CEU) and Gábor Kézdi (U. Michigan) Published on 6 May 2021 by Cambridge University Press, and developed for Data Analytics courses at CEU. Check out gabors-data-analysis.com for more detial.
  • Thanks to Péter Duronelly and Ádám Víg you can access the same material in python as well here.

More information on the courses is available here.

Contact

email: agoston.reguly-at-uni-corvinus.hu or areguly6-at-gatech.edu