Jozef Barunik


About


Jozef Baruník is an Associate Professor at the Institute of Economic Studies, Charles University in Prague. He also serves as a head of the Econometrics department at the Czech Academy of Sciences. In his research, he develops mathematical models for understanding financial problems (such as measuring and managing financial risk), develops statistical methods and analyzes financial data. Especially, he is interested in asset pricing, high-frequency data, financial econometrics, machine learning, high-dimensional financial data sets (big data), and frequency domain econometrics (cyclical properties and behavior of economic variables).

Jozef’s work has appeared in the Review of Economics and Statistics, Econometrics Journal, Journal of Financial Econometrics, Journal of Financial Markets, Econometric Reviews, Journal of Economic Dynamics and Control, The Energy Journal, he is an Associate Editor of the Digital Finance, Journal of Economic Interaction and Coordination, and Kybernetika, and also referees frequently for several journals and grant agencies in the fields of econometrics, finance, and statistics.



Submitted Publications


The Dynamic Persistence of Economic Shocks (with L.Vacha) preprint draft (June 2023)
Revise and resubmit, Review of Economics and Statistics
replication code and package in Julia


Common Idiosyncratic Quantile Risk (with M.Nevrla) preprint draft (May 2023)
Revise and resubmit, Review of Finance


Deep Learning, Predictability, and Optimal Portfolio Returns (with M.Babiak)
preprint draft (July 2021), Revise and resubmit, Journal of Banking and Finance


Selected Publications


Dynamic industry uncertainty networks and the business cycle (with M.Bevilacqua and R.Faff)
Journal of Economic Dynamics and Control (2024), Vol. 1598, 104793


Persistence in Financial Connectedness and Systemic Risk (with M.Ellington)
European Journal of Operational Research (2024), forthcoming
code and package in Julia and Matlab


Quantile Spectral Beta: A Tale of Tail Risks, Investment Horizons, and Asset Prices (with M.Nevrla)
Journal of Financial Econometrics (2023), Vol. 21, No. 5, pp. 1590–1646
replication codes


Asymmetric Network Connectedness of Fears (with M.Bevilacqua, and R.Tunaru)
The Review of Economics and Statistics (2022), 104(6): 1-13
replication codes and data


Measurement of Common Risks in Tails: A Panel Quantile Regression Model for Financial Returns (with F.Cech)
Journal of Financial Markets (2021), 52, 100562,
replication codes and data


Forecasting dynamic return distributions based on ordered binary choice (with S.Anatolyev)
International Journal of Forecasting (2019), 35(3), pp.823-835,
code and package


Quantile Coherency: A General Measure for Dependence between Cyclical Economic Variables (with T.Kley)
The Econometrics Journal (2019), 22(2), pp. 131-152,
code and package
Supplementary material available


Total, asymmetric and frequency connectedness between oil and forex markets (with E. Kocenda)
The Energy Journal (2019),vol 40, pp. 157 - 174,
codes for introduced measures
Note the package frequencyConnectedness available here can be used to replicate the paper
Supplementary material available


Measuring the Frequency Dynamics of Financial Connectedness and Systemic Risk (with T.Krehlik)
Journal of Financial Econometrics (2018), 16 (2), pp. 271–296,
code and package
Supplementary material available


Do co-jumps impact correlations in currency markets? (with L.Vacha)
Journal of Financial Markets (2018), 37, pp.97-119,
codes for introduced measures


Modeling and Forecasting Persistent Financial Durations (with F.Zikes and N.Shenai)
Econometric Reviews (2017), 36:10, 1081-1110,
Supplementary material available


Estimation of financial agent-based models with simulated maximum likelihood (with J.Kukacka)
Journal of Economic Dynamics and Control (2017), 85, pp. 21-45,


Asymmetric volatility connectedness on forex markets (with E. Kocenda and L.Vacha)
Journal of International Money and Finance (2017), 77C, pp. 39-56,
code for replication


Cyclical properties of supply-side and demand-side shocks in oil-based commodity markets (with T.Krehlik)
Energy Economics (2017), 65, pp.208-218,
code and package codes for bootstrap


Good volatility, bad volatility: Which drives the asymmetric connectedness of Australian electricity markets? (with N. Apergis and M.Lau)
Energy Economics (2017), 66, pp.108- 115,
codes for introduced measures
Note the package frequencyConnectedness available here can be used to replicate the paper


Asymmetric connectedness on the U.S. stock market: Bad and good volatility spillovers (with E. Kocenda and L.Vacha)
Journal of Financial Markets (2016), 27, 55–78,
codes for introduced measures
Note the package frequencyConnectedness available here can be used to replicate the paper


Semiparametric Conditional Quantile Models for Financial Returns and Realized Volatility (with F.Zikes)
Journal of Financial Econometrics (2016), 14 (1), 185–226,
Supplementary material available


Forecasting the term structure of crude oil futures prices with neural networks (with B.Malinska)
Applied Energy (2016), 164, pp.366–379,


Modeling and forecasting exchange rate volatility in time-frequency domain (with T.Krehlik and L.Vacha)
European Journal of Operational Research (2016), 251 (1), pp. 329–340,
codes to paper


Volatility spillovers across petroleum markets (with E. Kocenda and L.Vacha)
The Energy Journal (2015), 36(3), 309-329
codes for introduced measures
Note the package frequencyConnectedness available here can be used to replicate the paper


Are benefits from oil - stocks diversification gone? A new evidence from a dynamic copulas and high frequency data (with K.Avdulaj)
Energy Economics (2015), 51, pp. 31-44,


Realizing stock market crashes: stochastic cusp catastrophe model of returns under time-varying volatility (with J.Kukacka)
Quantitative Finance (2015), 15 (8), pp. 1347-1364,


Realized wavelet-based estimation of integrated variance and jumps in the presence of noise (with L.Vacha)
Quantitative Finance (2015), 15 (6), pp. 959-973,
codes for introduced measures


Are Bayesian Fan Charts Useful? The Effect of Zero Lower Bound and Evaluation of Financial Stability Stress Tests (with M.Franta, R.Horvath,K.Smidkova)
International Journal of Central Banking (2014), 10(1), 159-187


Comovement of energy commodities revisited: Evidence from wavelet coherence analysis (with L.Vacha)
Energy Economics (2012), 34(1), pp. 241–247,


Can a stochastic cusp catastrophe model explain stock market crashes? (with M.Vosvrda)
Journal of Economic Dynamics and Control (2012), 33(10), 1824-1836


Full list of publications available here



Working Papers (Work in Progress)


Forecasting Volatility of Oil-based Commodities: The Model of Dynamic Persistence (with L.Vacha) preprint draft (February 2024)
replication code and package in Julia

Abstract: Time variation and persistence are crucial properties of volatility that are often studied separately in oil-based volatility forecasting models. Here, we propose a novel approach that allows shocks with heterogeneous persistence to vary smoothly over time, and thus model the two together. We argue that this is important because such dynamics arise naturally from the dynamic nature of shocks in oil-based commodities. We identify such dynamics from the data using localised regressions and build a model that significantly improves volatility forecasts. Such forecasting models, based on a rich persistence structure that varies smoothly over time, outperform state-of-the-art benchmark models and are particularly useful for forecasting over longer horizons.


Moderation or indulgence? Effects of bank distribution restrictions during stress (with J.A. Smith, E.Gerba and P.Katsoulis)
Bank of England Staff Working Paper No. 1 053 (Nov 2023)

Abstract: At the onset of the Covid‑19 crisis, several regulatory authorities issued a recommendation or request to banks to restrict their dividend and share buyback distributions. The purpose of this action was to increase banks’ resilience by not distributing retained earnings, and help them support the real economy given their unique role in doing so. These restrictions reflected the singular circumstances brought by Covid‑19. We evaluate the impact of these restrictions on banks’ resilience, lending and investors’ required rate of return. First, using a difference‑in‑differences analysis on an international sample of European banks, we find that restricted banks increased their available Common Equity Tier 1 (CET1) capital and resilience in every quarter while the restrictions were fully in place, before gradually reducing it once they were partly lifted. Second, using a data set on the universe of UK small and medium‑sized enterprise (SME) loans issued by nine UK banking groups, we find that restricted banks increased their lending volumes on smaller non‑government guaranteed loans throughout the implementation period. Third, using the international sample of European banks, we find that the restrictions increased shareholders’ required rate of return throughout the implementation period, with the impact on the required rate of return on capital partially offset by lower debtholders’ required rate of return. The results indicate that distribution restrictions can be an effective crisis tool to increase banks’ resilience and lending capacity.


Common Firm-level Investor Fears: Evidence from Equity Options (with M.Bevilacqua and M.Ellington)
preprint draft (Sept 2023)

Abstract: We identify a new type of risk, common firm-level investor fears, from commonalities within the cross-sectional distribution of individual stock options. We define firm-level fears that link with upward price movements as good fears, and those relating to downward price movements as bad fears. Such information is different to market fears that we extract from index options. Stocks with high sensitivities to common firm-level investor fears earn lower returns, with investors demanding a higher compensation for exposure to common bad fears relative to common good fears. Risk premium estimates for common bad fears range from -5.63% to -4.92% per annum.


Learning Probability Distributions of Day-Ahead Electricity Prices (with L.Hanus)
preprint draft (Oct 2023)

Abstract: We propose a novel machine learning approach to probabilistic forecasting of hourly day-ahead electricity prices. In contrast to recent advances in data-rich probabilistic forecasting that approximate the distributions with some features such as moments, our method is non-parametric and selects the best distribution from all possible empirical distributions learned from the data. The model we propose is a multiple output neural network with a monotonicity adjusting penalty. Such a distributional neural network can learn complex patterns in electricity prices from data-rich environments and it outperforms state-of-the-art benchmarks.

code and package in Julia


Currency Network Risk (with M.Babiak)
preprint draft (May 2023)

Presented at at 2023 WFA (San Francisco), 2023 SGF Conference (Zurich), 2023 EEA-ESEM (Barcelona)

Abstract: This paper identifies a new currency risk stemming from linkages between option-implied currency volatilities. A volatility network strategy that buys net recipients and sells net transmitters of transitory shocks to ex-ante currency volatilities generates significant excess returns. Net recipients are more exposed to volatility spillovers and compensate investors with higher average returns. In contrast, net transmitters are more resilient to volatility transmissions and offer a lower risk premium because they hedge against volatility interdependencies. When volatility linkages are controlled for contemporaneous correlations, the strategy is uncorrelated with popular benchmarks. The volatility network factor is also priced in a currency cross-section.

The video below illustrates the short-term currency network during the Global Financial Crisis. If the video does not play properly, feel free to watch it here.


Learning Probability Distributions in Macroeconomics and Finance (with L.Hanus)
preprint draft (Apr 2022)

Abstract: We propose a deep learning approach to probabilistic forecasting of macroeconomic and financial time series. Being able to learn complex patterns from a data rich environment, our approach is useful for a decision making that depends on uncertainty of large number of economic outcomes. Specifically, it is informative to agents facing asymmetric dependence of their loss on outcomes from possibly non-Gaussian and non-linear variables. We show the usefulness of the proposed approach on the two distinct datasets where a machine learns the pattern from data. First, we construct macroeconomic fan charts that reflect information from high-dimensional data set. Second, we illustrate gains in prediction of stock return distributions which are heavy tailed, asymmetric and suffer from low signal-to-noise ratio.
code and package in Julia coming soon


Risks of heterogeneously persistent higher moments (with J.Kurka)
preprint draft (March 2024)

Abstract: Using intraday data for the cross-section of individual stocks, we show that both transitory and persistent fluctuations in realized market and average idiosyncratic volatility, skewness and kurtosis are differentially priced in the cross-section of asset returns, implying a heterogeneous persistence structure of different sources of higher moment risks. Specifically, we find that idiosyncratic transitory shocks to volatility as well as idiosyncratic persistent shocks to skewness contain strong commonalities that are relevant to investors.



Students


Current (Full-Time) Doctoral Students

Attila Sarkany: Machine Learning in Finance

Josef Kurka: Horizon-specific risks, higher moments, and asset prices

Martin Hronec: Asset Pricing with Quantile Machine Learning

Matej Nevrla: Tail risks, asset prices, and investment horizons

Lukas Janasek: Quantile Deep Reinforcement Learning in Economics

Lenka Nechvatalova: Deep Reinforcement Learning and Portfolio Management

Lubos Hanus (joint supervision with L.Vacha) Dynamic density forecasting using machine learning


Past Doctoral Students

Krenar Avdulaj: Essays in Financial Econometrics

Frantisek Cech: Three Essays on Risk Modelling and Empirical Asset Pricing

Tomas Krehlik: Applications of Modern Spectral Tools in Financial Econometrics

Jiri Kukacka: Estimation of Financial Agent-Based Models

Barbora Gregor: Three Essays on Data-Driven Methods in Asset Pricing and Forecasting


Teaching

JEM217 - Advanced Econometrics (Masters)

The objective of the course is to help students understand several important modern techniques in econometrics and apply them in empirical research and practical applications. Emphasis of the course will be placed on understanding the essentials underlying the core techniques, and developing the ability to relate the methods to important issues faced by a practicioner.

By completing this course, students will be able to use a computer based statistical software to analyze the data, choose appropriate models and estimators for given economic application, understand and interpret the results in detail (diagnose problems, understand proper inference) and will be confident to carry out the analysis and conclusions with respect to appropriatness and limitation of the methodology used. Finally, students will have sufficient grounding in econometric theory to begin advanced work in the field.

JEM059,061 - Financial Econometrics I and II (Masters)

The objective of the course is to introduce advanced time series methods. Students will be able to use the modern financial econometric tools after passing this course and will be prepared to continue in the Quantitative Finance II course. Part of the course is also focused on the high frequency data econometrics.

JEM116 - Applied Econometrics (Masters)

The course concentrates on the practical use of econometric methods, reviewing the relevant methodology, its use, and the possible alternative modeling approaches. The lectures are supplemented by computer classes, where students can gain hands-on experience in applied econometric analysis. During the course we will especially focus on time series techniques applied to forecasting asset volatility, modeling inflation, exchange rate volatility and other topics that you may regularly encounter in economic and financial literature. The course focuses on following topics in econometrics: OLS, IV, ARIMA, GARCH, VAR, cointegration, non-linear model and limited dependent variable. The course is especially suited for students undertaking empirical exercise in writing their M.A. thesis.

JED412,413 - Advanced Financial Econometrics (PhD seminar)

The aim of this seminar is an analysis of macroeconomic systems primarily by methods of nonlinear dynamics, stochastics dynamics, and nonlinear time series analysis. Special application will be focused on a behavior of the financial markets.

web page of the course


Contact


doc. PhDr. Jozef Baruník, Ph.D.

Associate Professor
Department of Macroeconomics and Econometrics
Institute of Economic Studies, Faculty of Social Sciences
Charles University
Opletalova 26
Prague 1, 110 00,Czech Republic

barunik [AT] fsv [DOT] cuni [DOT] cz
00 (420) 776 259273
https://ies.fsv.cuni.cz/en/contacts/people/29200599


Head of the Department
Department of Econometrics
Institute of Information Theory and Automation
Academy of Sciences of the Czech Republic
Pod Vodarenskou Vezi 4
Prague 8, 182 08,Czech Republic

barunik [AT] utia [DOT] cas [DOT] cz
00 (420) 776 259273
https://www.utia.cas.cz/people/barunik