Embedded Files
Dr Christis Katsouris
BS.c. (Hons, Bath), MS.c. (Warwick), MBA (UCY), MS.c. (UCY), Ph.D. (Southampton); Postdoc at the University of Helsinki.
© Christis G. Katsouris Institute of Econometrics and Data Science
“Our Research Laboratory at the Christis G. Katsouris Institute of Econometrics and Data Science, is designed with no a priori formal Academic Departments in mind. We conduct academic research and collaborate across scientific disciplines driven by shared values such as team curiosity, passion for problem-solving and commitment to making a positive impact in order to pioneer novel approaches across research themes that tackle socio-economic and environmental challenges.”
Dr. Christis Katsouris Ph.D., University of Southampton
# open mindedness, # enhancing knowledge, # expanding boundaries
Photo Credit: © Christis Katsouris (2012)
The Laboratory for Research in Statistical Methods of Econometrics and Machine Learning Applications at The Christis G. Katsouris Institute of Econometrics and Data Science was founded by Dr. Christis Katsouris in July 2024. For a full list of our research collaborators, please visit our Collaborations page.
MISSION STATEMENT
The Research Lab's mission is the development and evaluation of novel methodological approaches and to further encourage intellectual discussions around cutting-edge advances in theoretical and applied econometrics, statistics and economic modelling. We focus on enhancing the capabilities of statistical and data oriented research with the aim to promote interdisciplinary studies across economic, finance, actuarial and social sciences. The scope of our Research Laboratory is to bridge the gap in theoretical, applied and empirical studies across areas such as statistical inference for linear and nonlinear stochastic processes, empirical and applied macroeconometrics, time series and panel data econometrics, computational and high-dimensional econometrics with focus on developing novel machine learning methods and causal inference techniques. Further topics include robust estimation and inference, forecasting methodologies as well as economic analysis that supports an evidence-based approach to policy making. In summary, we aim to foster academic collaborations and accelerate the advancement of the econometric and statistical fields via theoretical and data-driven applications that address important questions of relevance to policy-making.
RESEARCH AREAS
Econometric and Computational Methods
Statistical Learning Theory and Methods
Macroeconometrics and Empirical Macroeconomics
Time Series and Financial Econometrics with Machine Learning
# Sequential Decision Problems
These are statistical decision problems which consist of test functions such as the Cusum or the Mosum statistics along with suitable stopping criteria (sequential decision rules) and are used in real-time monitoring schemes or in OOS sequential monitoring schemes for break-point detection purposes. Although the literature which addresses specific aspects related to break-point detection in these settings has exploded the last decades, no significant progress has been made with respect to the sequential decision problem at hand.
# Econometrics and Data Science Problems
© Christis G. Katsouris Institute of Econometrics and Data Science
News:
02 August 2024: During summer 2023 when Dr Christis Katsouris was working as a Visiting Lecturer in Economics at University of Exeter Business School, he supervised a postgraduate taught dissertation titled: “Forecast Evaluation under the presence of Instabilities”. Within this steam of literature an active research project is titled: “Inferring Predictive Accuracy in Nested Predictive Regressions Robust against Parameter Instability”.
02 September 2024: Joint working papers with research collaborators.
02 October 2024: I joined the Cyprus Statistical Society as an admitted member.
Working Papers
[1] Research Article (July 2024)
Title: "Estimating Dynamic Panel Data Models with Bank-Specific and Macro Determinants of Non-Performing Loans".
Co-author: Jiayang Li, MSc in Financial Economics Graduate, University of Exeter Business School.
[2] Research Article (August 2024)
Title: "Improved Uniform Confidence Interval Estimation for Autoregressive Parameter under Nonstationary Volatility".
Co-author: Dr. Xuewen Yu, Assistant Professor, Fudan University.
[3] Research Article (September 2024)
Title: "Testing for Unobserved Heterogeneity in Dynamic Panel Data Regressions under Diagonal Bilinear Dependence".
Co-author: Dr. Aziz Lmakri, Assistant Professor, Hassan II University of Casablanca.
[4] Research Article (October 2024)
Title: "Shrinkage Estimation in High-Dimensional Panel Data Models with Multiple Structural Breaks under Regressors Sparisty".
Co-author: Dr. Benjamin Poignard, Associate Professor, Keio University.
Disclaimers:
Funding: No funding or any research grant received from any funding agencies in the public, commercial, or non-for-profit sectors.
Conflicts of interest: No known competing financial interests or personal relationships that could have appeared to influence the work reported in these research projects.
Research Topics
I. Statistical Learning Theory and Methods
© Christis G. Katsouris Institute of Econometrics and Data Science
A fundamental research domain which facilitates the study of the asymptotic and nonasymptotic properties of machine learning estimators is the so-called statistical learning theory. Consider for example our aim is to estimate an unknown univariate function based on a nonparametric regression model. Then an important research object of interest for the statistician is the convergence rate of the mean integrated square error of such an estimator and the minimax optimality property of the MISE which tell us how fast the population mean squared distance between the estimator and the unknown function f shrinks to zero uniformly when this function ranges over a smoothness class, as the sample size increases.
Statistical learning theory is a growing field and these concepts are influencing recent developments in the structural econometrics and macro-finance literature; especially when the main objective is to bridge the gap between deep learning methods, high-dimensional nonlinear optimisation techniques and heterogeneous agent models in continuous-time settings. The statistical informativeness and econometric properties of these methods worth further study.
Photo Credit: © Christis Katsouris (2023)
References:
> Statistical Theory and Applications
Amorino., C., Brownlees, C., and Ghosh, A. (2025). "Concentration Inequalities for Suprema of Empirical Processes with Dependent Data via Generic Chaining with Applications to Statistical Learning". Preprint arXiv:2511.00597.
Derumigny, A., Girard, L., and Guyonvarch, Y. (2025). "Can We Have It All? Non-asymptotically Valid and Asymptotically Exact Confidence Intervals for Expectations and Linear Regressions". Preprint arXiv:2507.16776.
Cattaneo, M. D., Cox, G.F., Jansson, M., and Nagasawa, K. (2025). "Continuity of the Distribution Function of the Argmax of a Gaussian Process". Preprint arxiv:2501.13265.
Tan, F., Guo, X., and Zhu, L. (2025). "Weighted Residual Empirical Processes, Martingale Transformations, and Model Specification Tests for Regressions with Diverging Number of Parameters". Journal of Econometrics, 252, 106113.
Xu, Y., and Zeevi, A. (2025). "Towards Optimal Problem Dependent Generalization Error Bounds in Statistical Learning Theory". Mathematics of Operations Research, 50(1), 40-67.
Yuan, Z., and Spindler, M. (2025). "Bernstein-type Inequalities and Nonparametric Estimation under Near-Epoch Dependence". Journal of Econometrics, 251, 106054.
Cordoni, F., and Sancetta, A. (2024). "Consistent Causal Inference for High-Dimensional Time Series". Journal of Econometrics, 246(1-2), 105902.
Zhu, Y. (2024). "Phase Transitions in Nonparametric Regressions". Journal of Econometrics, 105640.
Katsouris, C. (2023). "High dimensional Time Series Regression Models: Applications to Statistical Learning Methods". Preprint arXiv:2308.16192.
> Statistical Theory and Methods
Lan, H., et al. (2025). "The Bias-Variance Trade-off in Data-Driven Optimization: A Local Misspecification Perspective". Preprint arXiv:2510.18215.
Shen, Z., Wang, C., Wang, S., and Yan, Y. (2025). "High-Dimensional Differentially Private Quantile Regression: Distributed Estimation and Statistical Inference". Preprint arXiv:2508.05212.
Li, S., and Zhang, L. (2024). "Fairm: Learning Invariant Representations for Algorithmic Fairness and Domain Generalization with Minimax Optimality". Preprint arXiv:2404.01608.
Bilodeau, B., Negrea, J., and Roy, D. M. (2023). "Relaxing the iid Assumption: Adaptively Minimax Optimal Regret via Root-Entropic Regularization". Annals of Statistics, 51(4), 1850-1876.
Chzhen, E., and Schreuder, N. (2022). "A Minimax Framework for Quantifying Risk-Fairness Trade-Off in Regression". Annals of Statistics, 50(4), 2416-2442.
Chen, S. X., and Peng, L. (2021). "Distributed Statistical Inference for Massive Data". Annals of Statistics, 49(5), 2851-2869.
Jordan, M. I., Lee, J. D., and Yang, Y. (2019). "Communication-Efficient Distributed Statistical Inference". Journal of the American Statistical Association, 114(526), 1-39.
II. Estimation and Inference Methods in Structural Econometrics
© Christis G. Katsouris Institute of Econometrics and Data Science
A second fundamental research domain in econometrics and statistics which facilitates the development of econometric methodologies, structural econometric models and associated econometric theory is the so-called information-theoretic approach. Related objects of interest include the optimality properties of GMM estimators, sequentially iterated GMM, empirical GMM, their convergence rates and the efficiency of optimal procedures such as simulation-based estimators for density functions (e.g., see the novel simulation-based estimator of the quantile density function proposed by Liao, Li & Fan (2024, arXiv:2410.03557)). Overall, applications span frameworks which include both linear regressions as well as time series regressions. Recently, there is a growing attention to the use of these approaches in high-dimensional data environments. A related suggestion to extend GMM techniques with gaussian kernel functionals in the case of infinite-dimensional regressors was discussed during a presentation by Katsouris (2024).
The first stream of literature considers nonnested testing for competitive statistical models to evaluate the specification of a statistical model against a specific alternative model. Within this stream of literature, a commonly used approach is to test for nonnested conditional moment restrictions. The second stream of literature considers nested testing for competitive statistical models to evaluate the specification of a statistical model against a specific alternative model. Further applications of interest, include estimation methods for parameters of structural models where empirical moments are matched (such as the minimum distance estimation approach).
Photo Credit: © Christis Katsouris (2009)
References:
Argañaraz, F., and Escanciano, J. C. (2024). "Machine Learning Debiasing with Conditional Moment Restrictions: An Application to LATE". Preprint arXiv:2410.23785.
Cho, J. S., and Phillips, P.C.B. (2024). "GMM Estimation with Brownian Kernels Applied to Income Inequality Measurement". Working paper 232, School of Economics, Yonsei University.
Katsouris C. (2024). "Teaching Presentation in Econometrics with an Application to Estimating Heterogeneous Treatment Effects at the Quantiles". Interview for the position of Lecturer in Econometrics, Department of Economics, University of Manchester, (28th May 2024).
Liao, X., Li, X., and Fan, Q. (2024). "Robust Bond Risk Premia Predictability Test in the Quantiles". Preprint arXiv:2410.03557.
Sørensen, J.R.V. (2024). "Testing a Class of Semi-or Nonparametric Conditional Moment Restriction Models using Series Methods". Econometric Theory, 40(4), 827-858.
Chernozhukov, V., Newey, W.K., and Santos, A. (2023). "Constrained Conditional Moment Restriction Models". Econometrica, 91(2), 709-736.
Sueishi, N. (2016). "A Simple Derivation of the Efficiency Bound for Conditional Moment Restriction Models". Economics Letters, 138, 57-59.
III. Identification and Estimation Methods in Macroeconometrics
© Christis G. Katsouris Institute of Econometrics and Data Science
A third fundamental related research domain is the statistical identification and estimation of econometric models which can facilitate pragmatic empirical research studies especially within the area of macroeconometrics. This approach has two important benefits: first it allow us to obtain suggestive explanations of economic phenomena rather than to validate or falsify existing theories, and second such a statistical decision theory approach, provides the means for developing econometric methodologies which are robust to data-driven stylized facts. For example, the classical Gaussian assumption has been found an inadequate mechanism for identifying Structural Vector Autoregressive models, especially when using time series data with heavy tails.
Developing identification and estimation methods that incorporate the properties of non-Gaussian time series provides a way to conduct pragmatic empirical research such as modelling the impact of climate shocks to the economy. Moreover, the identification of macroeconomic belief shocks, such as inflation expectations, using the non-Gaussianity property can provide a mechanism for robust estimation of counterfactual opinion forecasts, contributing this way to the literature related to the measurement of macroeconomic uncertainty.
Photo Credit: © Christis Katsouris (2009)
References:
Caggese, A., and Mesters, G. (2025). "Identifying Firm-Level Financial Frictions using Theory-Informed Restrictions". Working paper, Pompeu Fabra University.
Kolesár, M., and Plagborg-Møller, M. (2024). "Dynamic Causal Effects in a Nonlinear World: the Good, the Bad, and the Ugly". Preprint arXiv:2411.10415.
Hoesch, L., Lee, A., and Mesters, G. (2024). "Locally Robust Inference for Non‐Gaussian SVAR Models". Quantitative Economics, 15(2), 523-570.
Collard, F., Feve, P., and Guay, A. (2024). "Believe it or not, it's All About the Beliefs!". CEPR Discussion Paper (No. 19103). Available at https://hdl.handle.net/10419/301938.
Katsouris, C. (2023). "Structural Analysis of Vector Autoregressive Models". Preprint arXiv:2312.06402.
Lanne, M., and Luoto, J. (2021). "GMM Estimation of Non-Gaussian Structural Vector Autoregression". Journal of Business & Economic Statistics, 39(1), 69-81.
Lanne, M., Meitz, M., and Saikkonen, P. (2017). "Identification and Estimation of Non-Gaussian Structural Vector Autoregressions". Journal of Econometrics, 196(2), 288-304.
IV. Simulation-based Econometrics and Machine Learning Methods
© Christis G. Katsouris Institute of Econometrics and Data Science
Recent statistical literature documents a vigorous debate about the role of statistical methodologies in ensuring reproducibility and replicability of empirical findings in the disciplines of social sciences. The research theme focuses on statistical and econometric estimation and testing procedures with machine learning techniques and statistical guarantees. Moreover, addressing the issue of robust estimation and inference in the presence of missing observations, which is commonly discussed in empirical research, requires novel methodologies. Relevant examples include large datasets with firm fundamentals which could have complex missing patterns. In this case, conventional imputation methods, that typically work well in i.i.d settings are no longer valid, rendering classical inference for panel data with cross-sectional dependence as well as inference in efficient-frontier panel models unreliable, when the data missingness is not properly addressed using econometric techniques. Lastly, making code available and thoroughly referencing prior works has been found to increase citations; which is one of the main incentive and reward systems for academic work. Encouraging diversity of thought by critically evaluating the contributions of a wide spectrum of perspectives in the literature, is a good academic practice.
Photo Credit: © Christis Katsouris (2011)
References:
A. Missing Data Problem in Time Series Models
Duan, J., and Pelger, M. (2025). "Imputation-Powered Inference for Missing Covariates". NBER Working Paper (No. w34535). Available at nber/w34535.
Bryzgalova, S., Lerner, S., Lettau, M., and Pelger, M. (2024). "Missing Financial Data". Available at SSRN 4106794.
Freyberger, J., Höppner, B., Neuhierl, A., and Weber, M. (2024). "Missing Data in Asset Pricing Panels". Review of Financial Studies, hhae003.
Bai, J., and Ng, S. (2021). "Matrix Completion, Counterfactuals, and Factor Analysis of Missing Data". Journal of the American Statistical Association, 116(536), 1746-1763.
Rho, S. H., and Vogelsang, T. J. (2019). "Heteroskedasticity Autocorrelation Robust Inference in Time Series Regressions with Missing Data". Econometric Theory, 35(3), 601-629.
B. Statistical Methods and Inference with Machine Learning
> Generalized Linear Models
Caner, M. (2023). "Generalized Linear Models with Structured Sparsity Estimators". Journal of Econometrics, 236(2), 105478.
Chen, X., Li, H., and Zhang, J. (2023). "Complete Subset Averaging Approach for High-Dimensional Generalized Linear Models". Economics Letters, 226, 111084.
Vansteelandt, S., and Dukes, O. (2022). "Assumption-Lean Inference for Generalised Linear Model Parameters". Journal of the Royal Statistical Society Series B, 84(3), 657-685.
> Double Machine Learning Estimation and Inference
Liu, L., Mukherjee, R., and Robins, J. M. (2024). "Assumption-Lean Falsification Tests of Rate Double-Robustness of Double-Machine-Learning Estimators". Journal of Econometrics, 240(2).
Rothenhäusler, D., and Bühlmann, P. (2023). "Distributionally Robust and Generalizable Inference". Statistical Science, 38(4), 527-542.
Wang, M., et al. (2023). "Jackknife Model Averaging for High‐Dimensional Quantile Regression". Biometrics, 79(1), 178-189.
> Causal Inference with Non-Response/Attrition/Censoring
Ober-Reynolds, D. (2026). "Robustness to Missing Data: Breakdown Point Analysis". Journal of Econometrics, 253, 106151.
Linton, O., Seo, M.H., and Whang, Y.J. (2023). "Testing Stochastic Dominance with Many Conditioning Variables". Journal of Econometrics, 235(2), 507-527.
Fricke, H., Frölich, M., Huber, M., and Lechner, M. (2020). "Endogeneity and Non‐Response Bias in Treatment Evaluation–Nonparametric Identification of Causal Effects by Instruments". Journal of Applied Econometrics, 35(5), 481-504.
Nguimkeu, P., Denteh, A., and Tchernis, R. (2019). "On the Estimation of Treatment Effects with Endogenous Misreporting". Journal of Econometrics, 208(2), 487-506.
C. On Measuring Research Innovation and Impact
Arts, S., Melluso, N., and Veugelers, R. (2025). "Beyond Citations: Measuring Novel Scientific Ideas and their Impact in Publication Text". Review of Economics and Statistics, 1-33.
Campos-Mercade, P., and Mengel, F. (2024). "Non-Bayesian Statistical Discrimination". Management Science, 70(4), 2549-2567.
Wrobel, J., et al. (2024). "Partnering with Authors to Enhance Reproducibility at JASA". Journal of the American Statistical Association, 1-3.
Babina, T., et al. (2023). "Cutting the Innovation Engine: How Federal Funding Shocks Affect University Patenting, Entrepreneurship, and Publications". Quarterly Journal of Economics, 138(2), 895-954.
Brantner, C. L. et al. (2023). "Methods for Integrating Trials and Non-Experimental Data to Examine Treatment Effect Heterogeneity". Statistical Science, 38(4), 640-654.
Gustafson, P. (2023). "Parameter Restrictions for the Sake of Identification: Is There Utility in Asserting That Perhaps a Restriction Holds?". Statistical Science, 38(3), 477-489.
Parmigiani, G. (2023). "Defining Replicability of Prediction Rules". Statistical Science, 38(4), 543-556.
Connelly, R., et al. (2016). "The Role of Administrative Data in the Big Data Revolution in Social Science Research". Social Science Research, 59, 1-12.
Ewens, M., Tomlin, B., and Wang, L. C. (2014). "Statistical Discrimination or Prejudice? A Large Sample Field Experiment". Review of Economics and Statistics, 96(1), 119-134.
Coding Examples
Time Series Econometrics
Applied Econometrics
Bibliography: Econometric and Statistical Theory
Time Series Analysis
James D. Hamilton: Time Series Analysis.
Wayne A. Fuller: Introduction to Statistical Time Series.
Shumway and Stoffer: Time Series Analysis and its Applications.
Douc, Moulines and Stoffer: Nonlinear Time Series: Theory, Methods and Applications.
Econometric Theory
Hayashi, F.: Econometrics.
James Davidson: Econometric Theory.
Davidson R. and MacKinnon J.G: Econometric Theory and Methods.
Phoebus J. Dhrymes: Mathematics for Econometrics.
Halbert White: Asymptotic Theory for Econometricians.
Badi H. Baltagi (Edited by): Theoretical Econometrics.
George G. Judge et al: The Theory and Practice of Econometrics.
Michio Hatanaka: Time-Series Based Econometrics.
Lars P. Hansen and Thomas J. Sargent: Rational Expectations Econometrics.
Phillips P.C.B: Lecture Notes on Nonstationary Time Series (1988, 1995).
Statistical Theory
Hastie, Tibshirani and Friedman: The Elements of Statistical Learning.
James, Witten, Hastie, Tibshirani: An Introduction to Statistical Learning.
Lucien Le Cam: Asymptotic Methods in Statistical Decision Theory.
Bickel and Doksum: Mathematical Statistics.
van der Vaart A.W.: Asymptotic Statistics.
van de Geer S. and Bühlmann: Statistics for High-Dimensional Data.
Johnson and Wichern: Applied Multivariate Statistical Analysis.
Reinsel and Velu: Multivariate Reduced-Rank Regression.
Robert J. Serfling: Approximation Theorems of Mathematical Statistics.
Horvth Lajos and Piotr Kokoszka: Inference for Functional Data with Applications.
Aneiros, Horová, Hušková, and Vieu: Functional and High-Dimensional Statistics and Related Fields.
Probability Theory
Chow and Teicher: Probability Theory.
Billingsley P.: Probability and Measure, Convergence of Probability Measures.
Pollard D.: Convergence of Stochastic Processes.
Stout W.F.: Almost sure convergence.
Ward Whitt: Stochastic Process Limits.
Davidson J.: Stochastic Limit Theory.
Hall and Heyde: Martingale Limit Theory and its Application.
Van der Vaart A.W. and Wellner: Weak Convergence and Empirical Processes.
Häusler and Luschgy: Stable Convergence and Stable Limit Theorems.
Csörgö and Révész: Strong Approximations in Probability and Statistics.
Csörgö and Horváth. Limit Theorems in Change-Point Analysis.
Jacod and Shiryaev: Limit Theorems for Stochastic Processes.
Hans Crauel: Random Probability Measures on Polish Spaces.
Andreas E. Kyprianou: Fluctuations of Lévy Processes with Applications.
Gennady Samorodnitsky: Stochastic Processes and Long Range Dependence.
Applied Probability Theory
Haan and Ferreira: Extreme Value Theory.
Embrechts, Klüppelberg and Mikosch: Modelling Extremal Events.
Stuart Coles: An Introduction to Statistical Modeling of Extreme Values.
Snyder and Miller: Random Point Processes in Time and Space.
Sidney I. Resnick: Heavy-Tail Phenomena. Probabilistic and Statistical Modeling.
John P.Nolan: Univariate Stable Distributions: Models for Heavy Tailed Data.
Bibliography: Stochastic Calculus
Stochastic Finance and Financial Engineering Theory
Dana and Jeanblanc: Financial Markets in Continuous time.
Karatzas and Shreve: Brownian Motion and Stochastic Calculus.
Cont and Tankov: Financial Modelling with Jump Processes.
Philip E. Protter: Stochastic Integration and Differential Equations.
Ikeda and Watanabe: Stochastic Differential Equations and Diffusion Processes.
Ruppert and Matterson: Statistics and Data Analysis for Financial Engineering.
Resampling Methods
Chernick: Bootstrap Methods.
Good: Permutation, Parametric, and Bootstrap Tests of Hypotheses.
A.D.Davison and D.V.Hinkley: Bootstrap Methods and their Applications.
Peter Hall: The Bootstrap and Edgeworth Expansion.
Economic, Financial and Actuarial Indicators
The VIX indicator tracks expected volatility of the US stock market and is a measure of investor sentiment and asset pricing dynamics. The index is constructed using option pricing and market micro-structure theories and modelling techniques.
The presence of volatility fluctuations has motivated researchers to investigate the macroeconomic factors which explain these fluctuations as well as to propose novel ways to predict and test the predictive accuracy of forecasts obtained from econometric models conditional on the volatility uncertainty paradigm.
The Economic Policy Uncertainty Index has been developed by Baker et al. (2009, QJE). The index is constructed based on machine learning techniques and textual analysis which captures sentiment regarding the uncertainty in economic policy making.
The Economic Policy Uncertainty Index has been developed by Baker et al. (2009, QJE). The index is constructed based on machine learning techniques and textual analysis which captures sentiment regarding the uncertainty in economic policy making.
The EUI has motivated researchers to investigate the transmission mechanism (i.e., spillover effects and causal linkage) of uncertainty shocks to macroeconomic variables. Recent studies have documented the presence of nonlinear dynamics and asymmetric effects over the business cycle.
BTC Price Index
BTC Price Index
The Bitcoin price index monitors the price of cryptocurrencies, a worldwide payment system which enables two willing online parties to transact directly with each other without relying on a financial intermediary. Transactions are recorded in a publicly distributed ledger called a blockchain.
The Bitcoin price index monitors the price of cryptocurrencies, a worldwide payment system which enables two willing online parties to transact directly with each other without relying on a financial intermediary. Transactions are recorded in a publicly distributed ledger called a blockchain.
The strong volatility experienced by cryptocurrency exchanges has motivated researchers to investigate the existence of speculative bubbles, defined as a “persistent, systematic divergence of market price of a security from its fundamental value”.
The strong volatility experienced by cryptocurrency exchanges has motivated researchers to investigate the existence of speculative bubbles, defined as a “persistent, systematic divergence of market price of a security from its fundamental value”.
References:
Paye, B. S. (2012). "‘Déjà vol’: Predictive Regressions for Aggregate Stock Market Volatility using Macroeconomic Variables". Journal of Financial Economics, 106(3), 527-546.
Bekaert, G., and Hoerova, M. (2014). "The VIX, the Variance Premium and Stock Market Volatility". Journal of Econometrics, 183(2), 181-192.
Andreou, E., and Ghysels, E. (2021). "Predicting the VIX and the Volatility Risk Premium: The role of short-run funding spreads Volatility Factors". Journal of Econometrics, 220(2), 366-398.
Katsouris, C. (2021). "Forecast evaluation in Large Cross-Sections of Realized Volatility". Preprint arXiv:2112.04887.
Mancini, T., Calvo-Pardo, H., and Olmo, J. (2021). "Granger Causality Detection in High-Dimensional Systems using Feedforward Neural Networks". International Journal of Forecasting, 37(2), 920-940.
Pitarakis, J. Y. (2023). "A novel approach to Predictive Accuracy testing in Nested Environments". Econometric Theory, 1-44.
References:
Baker, S. R., Bloom, N., and Davis, S. J. (2016). "Measuring Economic Policy Uncertainty". Quarterly Journal of Economics, 131(4), 1593-1636.
Davis, S. J. (2016). "An index of Global Economic Policy Uncertainty". National Bureau of Economic Research Working paper (No. w22740).
Ahir, H., Bloom, N., and Furceri, D. (2022). "The World Uncertainty Index". National Bureau of Economic Research Working paper (No. w29763).
Virolainen, S. (2024). "A Statistically Identified Structural Vector Autoregression with Endogenously Switching Volatility Regime". Journal of Business and Economic Statistics, 1-20.
Virolainen, S. (2024). "Identification by non-Gaussianity in Structural Threshold and Smooth Transition Vector Autoregressive Models". Preprint arXiv:2404.19707.
References:
Chu, C. S. J., Stinchcombe, M., and White, H. (1996). "Monitoring Structural Change". Econometrica: Journal of the Econometric Society, 1045-1065.
Katsouris, C. (2017). "Sequential break-point detection in stationary time series: An application to monitoring economic indicators". Preprint arXiv:2112.06889.
Bouri, E. et al. (2019). "Co-explosivity in the Cryptocurrency Market". Finance Research Letters, 29, 178-183.
Phillips, P.C.B., and Shi, S. (2020). "Real time Monitoring of Asset Markets: Bubbles and crises". In Handbook of Statistics (Vol. 42, pp. 61-80). Elsevier.
Horváth, L., Liu, Z., Rice, G., and Wang, S. (2020). "Sequential Monitoring for Changes from Stationarity to Mild non-stationarity". Journal of Econometrics, 215(1), 209-238.
Horvath, L., Trapani, L., and Wang, S. (2024). "Sequential Monitoring for Explosive Volatility Regimes". Preprint arXiv:2404.17885.
The GAT time series correspond to temperature anomalies since the main objective is to track changes in temperature. These temperature anomalies are calculated as the difference between the observed temperature and the long-term average temperature for each location and date.
The index is an objective measure of observed changes in extreme weather and sea levels which allows to monitor climate trends. It is updated quarterly using meteorological season data. Such observations can be employed as a proxy of weather shocks into macroeconometric models.
Diffusion of technological developments, ongoing digital transformation efforts as well as the existing "productivity puzzle" motivates new economic thinking in relation to novel ways for modelling such complex and interconnected systems.
References:
Hansen, J., et al. (2006). "Global Temperature Change". Proceedings of the National Academy of Sciences, 103(39), 14288-14293.
Phillips, P.C.B., Leirvik, T., and Storelvmo, T. (2020). "Econometric Estimates of Earth’s Transient Climate Sensitivity". Journal of Econometrics, 214(1), 6-32.
Pretis, F. (2020). "Econometric Modelling of Climate Systems: The equivalence of energy balance models and cointegrated vector autoregressions". Journal of Econometrics, 214(1), 256-273.
Wagner, M., Grabarczyk, P., and Hong, S. H. (2020). "Fully Modified OLS Estimation and Inference for Seemingly Unrelated Cointegrating Polynomial Regressions and the Environmental Kuznets Curve for Carbon Dioxide Emissions". Journal of Econometrics, 214(1), 216-255.
Fouquet, R., & O'Garra, T. (2022). "In Pursuit of Progressive and Effective Climate Policies: Comparing an Air Travel Carbon Tax and a frequent flyer levy". Energy Policy, 171, 113278
Mancini, T., Calvo-Pardo, H., and Olmo, J. (2022). "Environmental Engel Curves: A Neural Network Approach". Journal of the Royal Statistical Society Series C: Applied Statistics, 71(5), 1543-1568.
References:
Acevedo et al. (2020). "The effects of Weather Shocks on Economic Activity: What are the channels of impact?". Journal of Macroeconomics, 65, 103207.
Gallic, E., and Vermandel, G. (2020). "Weather Shocks". European Economic Review, 124, 103409.
Kim, H. S., Matthes, C., and Phan, T. (2022). "Severe Weather and the Macroeconomy". Federal Reserve Bank of Richmond Working Papers, 21-14R.
Ciccarelli, M., and Marotta, F. (2024). "Demand or Supply? An empirical exploration of the effects of Climate Change on the Macroeconomy". Energy Economics, 129, 107163.
Lanne, M., and Virolainen, S. (2024). "A Gaussian Smooth Transition Vector Autoregressive Model: An application to the Macroeconomic Effects of Severe Weather Shocks". Preprint arXiv:2403.14216.
References:
Mayer, E., Rüth, S., and Scharler, J. (2016). "Total factor productivity and the Propagation of Shocks: Empirical evidence and implications for the Business Cycle". Journal of Macroeconomics, 50, 335-346.
Osotimehin, S. (2019). "Aggregate Productivity and the Allocation of Resources over the Business Cycle". Review of Economic Dynamics, 32, 180-205.
Barrage, L. (2020). "Optimal Dynamic Carbon Taxes in a Climate–Economy Model with Distortionary Fiscal Policy". Review of Economic Studies, 87(1), 1-39.
Czarnitzki, D., Fernández, G. P., and Rammer, C. (2023). "Artificial Intelligence and Firm-Level Productivity". Journal of Economic Behavior & Organization, 211, 188-205.
Duernecker, G., and Sanchez-Martinez, M. (2023). "Structural Change and Productivity Growth in Europe—Past, Present and Future". European Economic Review, 151, 104329.
Colaneri, K., Frey, R., and Köck, V. (2024). "Random Carbon Tax Policy and Investment into Emission Abatement Technologies". Preprint arXiv:2406.01088.
For the interested reader see also, "Theory of Viscosity Solutions of Hamilton-Jacobi Equations at the Boundary".
“Understanding productivity trends with respect to business cycle theories is important for maximising welfare and reducing income inequalities. For instance, if productivity is counter-cyclical and labor is reallocated with substantial time lags, an interpretation of a short-term increase in productivity could be that the most productive worker is attracted and retained during recessions. Perhaps we can then entertain the idea that new economic growth models which take into account sustainable development, environmental and welfare concerns in conjecture with the optimal rate of technological adaptation are needed. Using policy tools, as carbon tax on channels such as investment income rather than as a tax to be absorbed by economic agents, can accelerate the effectiveness of measures towards the Green transition.”
Dr Christis Katsouris, Ph.D. University of Southampton
© Christis G. Katsouris Institute of Econometrics and Data Science
What is Complex Systems Science?
The many challenges and crises of our times such as the climate change, efforts to ensure sustainable development and to deal with inequalities of various forms involves solving problems with increased complexity. Providing effective solutions to such complex problems requires unifying knowledge spanning several different sciences and disciplines. Therefore, the current transition from a mode of fragmentation and "compartmentalisation" of knowledge to an interdisciplinary approach where methodological reasoning and human values co-exist harmonously, is a driving-force for new thinking.
Evolutionary Dynamics of Economic and Financial Systems: such as the modelling of financial networks using econometric models which incorporate network topology dynamics (via the adjacency matrix). Applications include monitoring systemic risk, financial contagion and network connectivity. Moreover, the modelling of the interaction between climate and economic systems based on structural shock analysis is of interest. Further examples include the change-point detection in possibly nonstationary time series data (e.g., break recovery or estimation and testing for structural breaks), as well as model parameter estimation for nearly unstable or mildly explosive processes. For instance, financial time series are classified as 'mildly explosive' in periods where the economy experiences a faster-than-expected growth.
Biological, Ecological and Ocean Systems: includes the time series analysis and modelling of oscillation sequences which have applications when considering the stochastic behaviour of nearly, mildly and purely unstable processes in time series econometrics as well as when modelling diseases and chronic conditions such as the real-time monitoring of blood sugar fluctuations for diabetic patients in biostatistics. Overall the main characteristic of these systems is the presence of uniform persistence which, from the ecological point of view, persistence in a given community results when each species in the community is saved from extinction in the long-run.
© Christis G. Katsouris Institute of Econometrics and Data Science
Photo Credit: © Christis Katsouris (2009)
“A transition from a fragmentation and compartmentalisation mode of knowledge to an interdisciplinary approach where methodological reasoning and human values co-exist harmonously, is a driving-force for new thinking.”
Dr Christis Katsouris, Ph.D. University of Southampton
References:
From the econometric/statistical theory & methods perspective, relevant frameworks are grouped across interdependent research themes.
a. On Modelling Financial Networks and Macrofinance Risk Spillovers
Katsouris, C. (2024). "A Spatio-Temporal Autoregressive Distributed Lag Model with an Application to Monitoring Spatial Heterogeneity in Ecological Dynamics of Marine Habitats". Research Proposal 28 January 2024 (available upon request).
Katsouris, C. (2021). "Optimal Portfolio Choice and Stock Centrality for Tail Risk Events". Preprint arXiv:2112.12031.
Cited by: Hatami, Z. (2022). "A New Approach for Analyzing Financial Markets using Correlation Networks and Population Analysis". PhD thesis. University of Nebraska at Omaha.
Zhang, X., Opschoor, A., and Lucas, A. (2021). "The Importance of Heterogeneity in Dynamic Network Models Applied to European Systemic Risk". Tinbergen Institute, Discussion Paper (No. TI 2021-085/III). Available at https://hdl.handle.net/10419/248769.
Hale, G., and Lopez, J.A. (2019). "Monitoring Banking System Connectedness with Big Data". Journal of Econometrics, 212(1), 203-220.
Baruník, J., and Křehlík, T. (2018). "Measuring the Frequency Dynamics of Financial Connectedness and Systemic Risk". Journal of Financial Econometrics, 16(2), 271-296.
Demirer, M., Diebold, F.X., Liu, L., and Yilmaz, K. (2018). "Estimating Global Bank Network Connectedness". Journal of Applied Econometrics, 33(1), 1-15.
Blasques, F., Koopman, S.J., Lucas, A., and Schaumburg, J. (2016). "Spillover Dynamics for Systemic Risk Measurement using Spatial Financial Time Series Models". Journal of Econometrics, 195(2), 211-223.
Diebold, F. X., and Yılmaz, K. (2014). "On the Network Topology of Variance Decompositions: Measuring the Connectedness of Financial Firms". Journal of Econometrics, 182(1), 119-134.
Billio, M., Getmansky, M., Lo, A. W., and Pelizzon, L. (2012). "Econometric Measures of Connectedness and Systemic Risk in the Finance and Insurance Sectors". Journal of Financial Economics, 104(3), 535-559.
Koopman, S.J., Lucas, A., and Schwaab, B. (2011). "Modeling Frailty-Correlated Defaults using Many Macroeconomic Covariates". Journal of Econometrics, 162(2), 312-325.
Duffie, D., Eckner, A., Horel, G., and Saita, L. (2009). "Frailty Correlated Default". Journal of Finance, 64(5), 2089-2123.
Hong, Y., Liu, Y., and Wang, S. (2009). "Granger Causality in Risk and Detection of Extreme Risk Spillover between Financial Markets". Journal of Econometrics, 150(2), 271-287.
b. On Change-Point Detection
Lin, Y., Poignard, B., Pong, T. K., and Takeda, A. (2024). "Break Recovery in Graphical Networks with D-trace Loss". Preprint arXiv:2410.04057.
Xu, H., Wang, D., Zhao, Z., and Yu, Y. (2024). "Change-Point Inference in High-Dimensional Regression Models under Temporal Dependence". Annals of Statistics, 52(3), 999-1026.
Stival, M., Bernardi, M., and Dellaportas, P. (2023). "Doubly-Online Change-Point Detection for Monitoring Health Status during Sports Activities". Annals of Applied Statistics, 17(3), 2387-2409.
Engle, R. F., Ledoit, O., and Wolf, M. (2019). "Large Dynamic Covariance Matrices". Journal of Business & Economic Statistics, 37(2), 363-375.
Yuan, H., Xi, R., Chen, C., and Deng, M. (2017). "Differential Network Analysis via Lasso Penalized D-trace Loss". Biometrika, 104(4), 755-770.
Chan, N. H., Yau, C. Y., and Zhang, R. M. (2014). "Group LASSO for Structural Break Time Series". Journal of the American Statistical Association, 109(506), 590-599.
Kolar, M., and Xing, E. P. (2012). "Estimating Networks with Jumps". Electronic Journal of Statistics, 6, 2069.
c. On Probability Forecasts
Price, I., et al. (2024). "Probabilistic Weather Forecasting with Machine Learning". Nature, 1-7.
Martin, G.M., et al. (2022). "Optimal Probabilistic Forecasts: When do they Work?". International Journal of Forecasting, 38(1), 384-406.
Lai, T.L., Gross, S.T., and Shen, D.B. (2011). "Evaluating Probability Forecasts". Annals of Applied Statistics, 39(5): 2356-2382.
d. On Modelling Biological Processes
Xie, Y., Fu, H., Huang, Y., Pozdnyakov, V., and Yan, J. (2025). "Recurrent Events Modeling based on a Reflected Brownian Motion with Application to Hypoglycemia". Biostatistics, 26(1), kxae053.
Blayneh, K.W. (2023). "Uniform Persistence and Backward Bifurcation of Vertically Transmitted Vector-Borne Diseases". Research in Mathematics, 10(1), 2264581.
Zeldow, B., et al. (2019). "A Semiparametric Modeling Approach using Bayesian Additive Regression Trees with an Application to Evaluate Heterogeneous Treatment Effects". Annals of Applied Statistics, 13(3), 1989-2010.
Discussion Topic A.
Main Challenges of Neural Network Systems
We briefly discuss some of the main challenges that neural network systems face:
Lack of neuroplasticity and continual learning
Lack of ability of switching between tasks
The significance of such neural network systems in enhancing the capabilities of machine learning and deep learning applications for causal and statistical inference purposes as well as in other industrial applications are, undoubtedly enormous. However, the ability of deep learning methods in environments where continual learning is essential might exhibit gradual lose of plasticity - a fundamental property of the human brain; and in fact this could be one of the main limitations of these systems. Recently, some studies propose solutions to tackle the fact that deep learning methods lose their ability to learn (plasticity loss) over time, such as via an algorithmic procedure that continuously updates the 'diversity of neurons'.
Relevant aspects for further research include the domain of unsupervised learning without feedback. From the statistical perspective, this implies a learning environment under model uncertainty. Take for example, the case of life insurance and credit risk modelling, then a related issue is the problem of superhedging for payment streams in continuous time under model uncertainty. In other words, this covers the case when we are interested to model an event which occurs as a surprise and is itself not observable under the reference information flow.
Another example, suppose that a neuroscientist is interested in the biological mechanisms that facilitate associative learning between neurons. In particular, suppose that the neuroscientist is interested to identify 'environments' within which such mechanism might break down. These settings require modelling approaches which capture interactions between neurons using machine learning techniques. From the econometrics point of view, relevant questions include: (i) suitable methodologies for identification and estimation purposes, (ii) statistical testing formulations and (iii) statistical evaluation techniques which permit to assess the performance of these methodologies with real and simulated datasets.
𝕊tat∧Prob_ML{metrics}
References:
Probability Theory
Do, Q.H., Nguyen, B.T., and Ho, L.S.T. (2024). "A Generalization Bound of Deep Neural Networks for Dependent Data". Statistics & Probability Letters, 208, 110060.
Biagini, F., and Zhang, Y. (2019). "Reduced-Form Framework under Model Uncertainty". Annals of Probability, 29 (4), 2481-2522.
During a Hybrid Seminar at Statistical Sciences Research Institute, University of Southampton (9th December 2021).
[Seminar Speaker: Dr. Sebastian Engelke, University of Geneve].
During a Hybrid Seminar at Statistical Sciences Research Institute, University of Southampton (24th March 2022).
[Seminar Speaker: Dr. Haeren Cho, University of Bristol].
During a Hybrid Seminar at Statistical Sciences Research Institute, University of Southampton (November 2022).
[Seminar Speaker: Marvin Pförtner, PhD student, University of Tübingen].
During a Hybrid Seminar at Statistical Sciences Research Institute, University of Southampton (11th November 2022).
[Seminar Speaker: Dr. Henry Reeve, University of Bristol].
During the Young Statisticians Europe Online Workshop: "Recent Challenges in Model Specification Testing based on Different Data Structures" (9th November 2022).
During the Young Statisticians Europe Online Workshop: "Recent Challenges in Model Specification Testing based on Different Data Structures" (9th November 2022).
During an Econometrics Seminar at the University of Exeter Business School (27th March 2023).
[Seminar Speaker: Prof. Robert Taylor, University of Essex].
During an Econometrics Seminar at the University of Exeter Business School (28th March 2023)
[Seminar Speaker: Prof. Anna Simoni, Institut Polytechnique de Paris].
During a Time Series and Machine Learning Reading Group Session at the University of Southampton (April 2023).
During a Hybrid Seminar at the Department of Economics, University of Southampton (17th October 2023).
[Seminar Speaker: Dr. Chaowen Zheng, University of Southampton].
# Spatial Panel AR Model, # Interactive Fixed Effects Structure # Two-Step Estimation Procedure
During a TS and ML Reading Group Session (3rd November 2023).
[Guest Speaker: Dr. Shen Guohao, Department of Applied Mathematics, Hong Kong Polytechnic University].
# Statistical Machine Learning, # Nonparametric Methods, # Deep Learning
During a Hybrid Seminar at Statistical Sciences Research Institute, University of Southampton (16th May 2024).
# Change-Point Detection, # Rate of Convergence, # Error Bounds
Participation to Workshops, Seminars & Reading Groups in Econometrics & Statistics
# Building a Positive Research Culture, # Team Curiosity, # Collaborative Learning, # Cross-Departmental Research
Discussion Topic B.
Reinforcement Learning and Bandit Algorithms
Reinforcement learning describes how an agent can learn an optimal action policy in a sequential decision process, through repeated experience. Specifically, from the computational perspective, RL’s main objective is to find the most suitable action model to maximise total cumulative reward for the RL agent. In economics and finance related applications include the optimal portfolio selection problem (such as matrix completion bandit), online statistical problems (such as minimax problems in treatment choice and online policy inference), among others.
Deep learning estimation techniques can be used for modelling sustainable human behaviour, in simulated environments of human interactions, which is especially useful for obtaining insights on how human players can benefit from acting as social planner's who care about maximising the welfare of all participants in a game. Moreover, deep learning approaches have also seen growing attention in the treatment effects estimation literature. However, one of the main challenges of the particular framework, from the computational perspective, is the presence of increasing number of model parameters. Therefore, optimising computational resources during inference is crucial especially in environments with limited resources, by allowing more efficient and cost-effective solutions without sacrificing performance.
Source: medium.com
Reading Group:
Reinforcement Learning and Bandit Algorithms Joint Reading Group (2024).
Book:
Lattimore, T., & Szepesvári, C. (2020). Bandit Algorithms. Cambridge University Press.
References:
Reinforcement Learning
Lowet, A.S., Zheng, Q., Meng, M. et al. (2025). "An Opponent Striatal Circuit for Distributional Reinforcement Learning". Nature.
Duan, C., Li, J., and Xia, D. (2024). "Online Policy Learning and Inference by Matrix Completion". Preprint arXiv:2404.17398.
Koster, R., et al. (2024). "Using Deep Reinforcement Learning to Promote Sustainable Human Behaviour on a Common Pool Resource Problem". Preprint arXiv:2404.15059.
Jiang, Y., Olmo, J., and Atwi, M. (2024). "Deep Reinforcement Learning for Portfolio Selection". Global Finance Journal, 62, 101016.
Charpentier, A., Elie, R., and Remlinger, C. (2021). "Reinforcement Learning in Economics and Finance". Computational Economics, 1-38.
Mnih, V., ..., and Hassabis, D. (2015). "Human-level Control through Deep Reinforcement Learning". Nature, 518(7540), 529-533.
Discussion Topic C.
Topological Data Analysis for Time Series
Statistical Properties of Persistence homologies
Statistical Inference on Manifolds
Generative Models: Mathematical Foundations
Topological data analysis is build on fundamental concepts and results commonly used in algebraic topology, viewing data shapes - such as data clouds, as the objects of statistical interest while the underline mathematical mechanism is free of coordinates and invariant to deformations. These properties allows us to conduct explanatory analysis and implement feature extraction in supervise learning environments as well as to develop suitable frameworks for statistical inference using labeled or unlabeled data. A relevant example is the identification and estimation of covariate shifts in distributions with unlabeled data. More recently, these applications started to gain attention for econometric inference purposes.
“Topological Data Analysis provides an alternative way for the statistical analysis of high-dimensional data. For example, Uncovering causal relations based on the shape of data can provide new insights to many other disciplines.”
Dr Christis Katsouris
References:
Statistical Theory
Meitz, M. (2024). "Statistical Inference for Generative Adversarial Networks and Other Minimax Problems". Scandinavian Journal of Statistics.
Qiu, Y., Gao, Q., and Wang, X. (2024). "Adaptive Learning of the Latent Space of Wasserstein Generative Adversarial Networks". Preprint arXiv:2409.18374.
Zhou, X., et. al. (2023). "A Deep Generative Approach to Conditional Sampling". Journal of the American Statistical Association, 118(543), 1837-1848.
Jumper, J., ..., and Hassabis, D. (2021). "Highly Accurate Protein Structure Prediction with AlphaFold". Nature, 596(7873), 583-589.
Kpotufe, S., and Martinet, G. (2021). "Marginal Singularity and the Benefits of Labels in Covariate-Shift". Annals of Statistics, 49(6), 3299-3323.
Petersen, A., and Müller, H. G. (2019). "Fréchet Regression for Random Objects with Euclidean Predictors". Annals of Statistics, 47(2), 691-719.
Probability Theory
Onaran, E., Bobrowski, O., and Adler, R. J. (2023). "Functional Limit Theorems for Local Functionals of Dynamic Point Processes". Preprint arXiv:2310.17775.
Pardon, J. (2011). "Central Limit Theorems for Random Polygons in an Arbitrary Convex Set". Annals of Probability, 39(3), 881-903.
Drees, H., and Rootzén, H. (2010). "Limit Theorems for Empirical Processes of Cluster Functionals". Annals of Statistics, 38(4): 2145-2186.
Time Series and Topological Data Analysis
Gao, H., Kaltenbach, S., and Koumoutsakos, P. (2024). "Generative Learning for Forecasting the Dynamics of High-Dimensional Complex Systems". Nature Communications, 15(1), 8904.
Martínez-Sánchez, Á., Arranz, G. and Lozano-Durán, A. (2024). "Decomposing Causality into its Synergistic, Unique, and Redundant Components". Nature Communications 15, 9296.
Suzuki, K., Matsuzaki, S. I. S., and Masuya, H. (2022). "Decomposing Predictability to Identify Dominant Causal Drivers in Complex Ecosystems". Proceedings of the National Academy of Sciences, 119(42), e2204405119.
Garside, K., Henderson, R., Makarenko, I., and Masoller, C. (2019). "Topological Data Analysis of High Resolution Diabetic Retinopathy Images". PloS One, 14(5), e0217413.
Gidea, M., and Katz, Y. (2018). "Topological Data Analysis of Financial Time Series: Landscapes of Crashes". Physica A: Statistical Mechanics and its Applications, 491, 820-834.
Google Sites
Report abuse