**INFORMATION ABOUT PROJECT,
SUPPORTED BY RUSSIAN SCIENCE FOUNDATION**

*The information is prepared on the basis of data from the information-analytical system RSF, informative part is represented in the author's edition. All rights belong to the authors, the use or reprinting of materials is permitted only with the prior consent of the authors. *

COMMON PART

Project Number22-18-00588

Project titleRobust methods in econometrics, economics and finance: From the analysis of crises, structural shocks and financial contagion to inequality measurement and the analysis of the efficiency of economic policy measures

Project LeadIbragimov Rustam

AffiliationFederal State Budgetary Educational Institution of Higher Education "Saint-Petersburg State University",

Implementation period | 2022 - 2024 |

Research area 08 - HUMANITIES AND SOCIAL SCIENCES, 08-155 - Prognosis of social and economic development, governmental regulation of economy and regulation of socio-economic processes

KeywordsRobust methods, robust econometric inference, econometric models, estimation, forecasting, crises, structural shocks, financial contagion, income inequality, income distribution, efficiency, economic policy measures, market regulation, social and economic development, cross-country and cross-regional comparisons

PROJECT CONTENT

Annotation

Econometric and statistical analysis, modelling and forecasting in economics and finance are significantly complicated by the presence of problems of dependence, heterogeneity and outliers in the studied data, as well as problems of endogeneity in the estimated models. The above problems in economic and financial observations and models arise, in particular, due to the interdependence of key indicators in economics and finance; clustering of volatility, the impact of crises and other structural shocks such as Black Monday of October 19, 1987, Flash Crash of May 6, 2010 and the onset of the COVID-19 pandemic; and also due to financial contagion and the spread of crises, as in the case of the start of the global financial crisis in 2008. The presence of dependence, heterogeneity, and outliers in the analyzed data naturally leads to changes in the asymptotics of the considered statistical estimates, in particular, their limiting variances and standard errors. Moreover, the presence of endogeneity in econometric, economic and financial models makes the estimates of their parameters inconsistent, i.e. biased - even in large samples.
The standard and widely used approaches to econometric and statistical analysis and modeling for problems of endogeneity, dependence, heterogeneity, emissions and extreme values in economic and financial indicators and time series are given by the instrumental variable methods; asymptotic methods based on consistent standard errors and approximations for distributions of test statistics in large samples; and models of the dynamics of financial and economic time series and their volatility. Widely used methods of asymptotic statistical and econometric analysis include the approaches based on Heteroskedasticity and autocorrelation consistent (HAC) standard errors for time series and cluster standard errors for panel data. Asymptotic methods often use the convergence of t-statistics of the estimated parameters of economic and financial models to the standard normal distribution under the null hypothesis. Key models of economic and financial dynamics include predictive regressions, factor models, and GARCH models for financial returns and foreign exchange rates. Important contributions to the development and applications of the above approaches and methods were made, in particular, by laureates of the Nobel Prize in economics J. D. Angrist, G. W. Imbens and D. Card (2021 Nobel Prize), A. Banerjee, E. Duflo and M. Kremer (2019 Nobel Prize), E. F. Fama, L. P. Hansen and R. J. Shiller (2013 Nobel Prize) and R. F. Engle and C. W. J. Granger (2003 Nobel Prize).
The use of instrumental variable regressions allows one to consistently estimate the economic and financial models under the analysis, and the use of consistent standard errors makes it possible to use standard asymptotic distributions, as in the case of asymptotic normality of HAC t-statistics in the analysis of time series regressions, for example, predictive regressions and models for financial returns and asset prices. The asymptotic theory of instrumental variable regression estimates, such as those in two-step least squares (TSLS) regression, and asymptotic theory of statistical and econometric analysis based on consistent standard errors are now well-developed and widely used. However, unfortunately, applications of the asymptotic theory and methods of instrumental variables and consistent - under dependence and heterogeneity - standard errors in the case of samples of economic and financial data observed in practice are often complicated by the fact that the above methods of asymptotic econometric analysis have poor statistical properties in finite samples (see, among others, Andrews, 1991, Andrews and Monahan, 1992, den Haan and Levin, 1997, and Ibragimov and Muller, 2010; all references are provided in the supplementary file). These problems are observed even in the case of fairly simple econometric models and estimates, especially under pronounced dependence and heterogeneity.
The search for solutions to problems in applications of asymptotic methods based on consistent errors in the analysis of economic and financial data with structures of dependence, heterogeneity and outliers observed in practice has led to the development of a number of relatively recently proposed robust statistical and econometric methods. Ibragimov and Muller (2010) (see also section 3.3 in the monograph Ibragimov, Ibragimov and Walden, 2015) provide new approaches to robust testing of hypotheses on a single parameter of economic and financial models (for example, a parameter in predictive regressions for financial returns) under autocorrelation and heterogeneity in the data that do not require consistent estimation of the limiting variances of parameter estimates and their standard errors. These approaches are based on the use of t-statistics calculated using groups of observations in the sample under consideration and the critical values of the standard Student-t distributions. The extensive numerical analysis in Ibragimov and Muller (2010) and subsequent works in the literature shows that the statistical properties of robust t-statistic inference approaches often outperform the properties of widely used traditional inference methods based on consistent standard errors, especially when there is pronounced dependence and heterogeneity. in the data.
Naturally, many important problems in economics, finance, econometrics and related fields require the analysis of statistical hypotheses for two or more parameters. As leading examples, one may mention fundamental problems of analyzing treatment effects and the effectiveness of economic policy measures; the analysis of structural changes and the impact of crises and other structural shocks on economic and financial markets; cross-country and cross-regional comparisons of socio-economic development and its dynamics; tests of joint hypotheses in regression models; the analysis of Granger causality in predictive models and models of the relationship between key economic and financial indicators; and many others.
In addition, research on many problems of economics, finance and econometrics is based on two-step procedures for statistical and econometric analysis and estimation. Standard examples are given by two-stage least squares (TSLS) methods used in instrumental variable analysis in these areas under the problems of endogeneity in the models and data considered and the implied inconsistency of ordinary least squares (OLS) estimates. Two-stage inference methods are also naturally used and, in fact, are the only possible methods of statistical analysis, in the study of factors influencing inequality of income and wealth, including inequality in the tails of income/wealth distributions. Two-stage procedures also form the basis for the methods for estimation and inference on the parameters of copula dependence structures, e.g., in models of the interdependence of economic and financial markets and their financial contagion properties. Since the second step of two-stage inference methods uses estimates or predicted values of the dependent variable obtained at the first stage, this introduces additional uncertainty in the estimates of the studied parameters obtained by these methods. This leads to important problems on calculation of the correct limiting variances and estimation of standard errors of two-stage estimates and, more generally, on asymptotic validity of statistical tests and inference based on the estimates obtained.
The main objectives and contributions of the project consist in research in the following two key directions and related areas:
1. Development of econometric toolbox for robust statistical and econometric analysis of hypotheses on two or more parameters in economics and finance with applications in important areas in economics, finance and econometrics, including robust tests for structural breaks in economic and financial markets and in their dynamics; robust econometric analysis of the effectiveness of economic policy measures and treatment effects; robust tests of joint hypotheses on several regression parameters; cross-country and cross-regional comparisons of the dynamics of socio-economic development and market models; and the analysis of Granger causality in predictive models, among others.
2. Development of new and improved approaches to robust statistical and econometric analysis in two-stage inference methods under endogeneity, dependence, heterogeneity and outliers in models and data. Research in this direction will allow one to unify the analysis based on many methods widely used in econometrics, finance and economics, including the above approaches to instrumental statistical and econometric analysis; identification and assessment of factors and determinants influencing income and wealth inequality both over the whole distributions of income and wealth as well as in their tails; and the analysis and estimation of copula dependence structures for economic and financial markets and financial contagion.
The methods and approaches used in the research on the project will include, among others, the probabilistic results on conservativeness properties of one- and two-sample t-tests for the means of small samples of independent heterogeneous Gaussian observations in the literature (see Bakirov, 1989, 1998, Bakirov and Szekely, 2005, Ibragimov and Muller, 2010, 2016, and references therein), as well as new results on their refinements and generalizations, including generalizations to the case of heavy-tailed observations, inequalities for sums of symmetric and unimodal random variables, and generalizations to the important case of F -tests for joint hypotheses and probability inequalities for quadratic forms.

Expected results

Econometric and statistical analysis, modelling and forecasting in economics and finance are significantly complicated by the presence of problems of dependence, heterogeneity and outliers in the studied data, as well as problems of endogeneity in the estimated models. The above problems in economic and financial observations and models arise, in particular, due to the interdependence of key indicators in economics and finance; clustering of volatility, the impact of crises and other structural shocks such as Black Monday of October 19, 1987, Flash Crash of May 6, 2010 and the onset of the COVID-19 pandemic; and also due to financial contagion and the spread of crises, as in the case of the start of the global financial crisis in 2008. The presence of dependence, heterogeneity, and outliers in the analyzed data naturally leads to changes in the asymptotics of the considered statistical estimates, in particular, their limiting variances and standard errors. Moreover, the presence of endogeneity in econometric, economic and financial models makes the estimates of their parameters inconsistent, i.e. biased - even in large samples.
The standard and widely used approaches to econometric and statistical analysis and modeling for problems of endogeneity, dependence, heterogeneity, emissions and extreme values in economic and financial indicators and time series are given by the instrumental variable methods; asymptotic methods based on consistent standard errors and approximations for distributions of test statistics in large samples; and models of the dynamics of financial and economic time series and their volatility. Widely used methods of asymptotic statistical and econometric analysis include the approaches based on Heteroskedasticity and autocorrelation consistent (HAC) standard errors for time series and cluster standard errors for panel data. Asymptotic methods often use the convergence of t-statistics of the estimated parameters of economic and financial models to the standard normal distribution under the null hypothesis. Key models of economic and financial dynamics include predictive regressions, factor models, and GARCH models for financial returns and foreign exchange rates. Important contributions to the development and applications of the above approaches and methods were made, in particular, by laureates of the Nobel Prize in economics J. D. Angrist, G. W. Imbens and D. Card (2021 Nobel Prize), A. Banerjee, E. Duflo and M. Kremer (2019 Nobel Prize), E. F. Fama, L. P. Hansen and R. J. Shiller (2013 Nobel Prize) and R. F. Engle and C. W. J. Granger (2003 Nobel Prize).
The use of instrumental variable regressions allows one to consistently estimate the economic and financial models under the analysis, and the use of consistent standard errors makes it possible to use standard asymptotic distributions, as in the case of asymptotic normality of HAC t-statistics in the analysis of time series regressions, for example, predictive regressions and models for financial returns and asset prices. The asymptotic theory of instrumental variable regression estimates, such as those in two-step least squares (TSLS) regression, and asymptotic theory of statistical and econometric analysis based on consistent standard errors are now well-developed and widely used. However, unfortunately, applications of the asymptotic theory and methods of instrumental variables and consistent - under dependence and heterogeneity - standard errors in the case of samples of economic and financial data observed in practice are often complicated by the fact that the above methods of asymptotic econometric analysis have poor statistical properties in finite samples (see, among others, Andrews, 1991, Andrews and Monahan, 1992, den Haan and Levin, 1997, and Ibragimov and Muller, 2010; all references are provided in the supplementary file). These problems are observed even in the case of fairly simple econometric models and estimates, especially under pronounced dependence and heterogeneity.
The search for solutions to problems in applications of asymptotic methods based on consistent errors in the analysis of economic and financial data with structures of dependence, heterogeneity and outliers observed in practice has led to the development of a number of relatively recently proposed robust statistical and econometric methods. Ibragimov and Muller (2010) (see also section 3.3 in the monograph Ibragimov, Ibragimov and Walden, 2015) provide new approaches to robust testing of hypotheses on a single parameter of economic and financial models (for example, a parameter in predictive regressions for financial returns) under autocorrelation and heterogeneity in the data that do not require consistent estimation of the limiting variances of parameter estimates and their standard errors. The proposed robust tests of statistical and econometric hypotheses for the parameter under consideration are very easy to use: Following the approaches to robust statistical analysis in Ibragimov and Muller (2010), the data sample is split into a certain fixed number q of groups (say, into q = 2, 4, 8 groups of consecutive observations in time series for which regression models are estimated); then the analyzed parameter is estimated using the data in each group of observations, and statistical conclusions in tests of hypotheses about the value of the parameter (for example, the hypothesis about the equality of the regression coefficient to zero) are made on the basis of t-statistics in group estimates of the parameter and the critical values of the Student's standard distribution with q-1 degrees of freedom. For example, in the case of the number of groups equal to q = 4, the quantiles of the Student-t distribution with three numbers of freedom are used. The above approaches to robust testing of hypotheses allow constructing, in a standard way, robust confidence intervals for unknown values of the parameter of the analyzed economic and financial models using the estimates of the parameter and quantiles of Student-t distributions. According to the theoretical results in Ibragimov and Muller (2010), the proposed approaches to robust statistical and econometric analysis are asymptotically valid under the general assumptions that the group estimates of the parameter under study are asymptotically normal and asymptotically independent. These conditions are satisfied for many widely used models and dependence/heterogeneity structures in economics and finance, including dependence models in time series, cluster and spatially dependent samples, and panel data. In particular, it is important to note that the asymptotic normality of parameter estimators calculated over groups of observations generally holds under the same assumptions as the asymptotic normality of full-sample estimators. The extensive numerical analyses in Ibragimov and Muller (2010) and subsequent works in the literature show that the statistical properties of t-statistic approaches to robust inference often outperform the properties of widely used traditional robust inference methods based on consistent standard errors, especially when there is pronounced dependence and heterogeneity in the data (see also paragraph 3.3 in the monograph Ibragimov, Ibragimov and Walden, 2015, as well as the recent study in Esarey and Menger, 2019). The methods and approaches to robust statistical and econometric inference proposed in Ibragimov and Muller (2010) have been used in applied and theoretical analysis in a number of works in econometrics, economics and finance, including, among others, Bloom et al. (2013), Krueger et al. (2017), Blinder and Watson (2016), Verner and Gyongyosi (2018), Anatolyev (2019), Chen and Ibragimov (2019) and Gargano et al. (2019).
Naturally, many important problems in economics, finance, econometrics and related fields require the analysis of statistical hypotheses for two or more parameters. As leading examples, one may mention fundamental problems of analyzing treatment effects and the effectiveness of economic policy measures; the analysis of structural changes and the impact of crises and other structural shocks on economic and financial markets; cross-country and cross-regional comparisons of socio-economic development and its dynamics; tests of joint hypotheses in regression models; the analysis of Granger causality in predictive models and models of the relationship between key economic and financial indicators; and many others.
In addition, research on many problems of economics, finance and econometrics is based on two-step procedures for statistical and econometric analysis and estimation. Standard leading examples are given by two-stage least squares (TSLS) methods used in instrumental variable analysis in these areas under the problems of endogeneity in the models and data considered and the implied inconsistency of ordinary least squares (OLS) estimates: At the first stage, the regressions of endogenous regressors on instruments are estimated, and the second stage of the procedures is based on regressions of the dependent variable on the values of these regressors predicted at the first stage. Two-stage inference methods are also naturally used and, in fact, are the only possible methods of statistical analysis, in the study of factors influencing inequality of income and wealth, including inequality in the tails of income/wealth distributions: Here, the analysis is based on regressions of estimates of inequality measures (for the entire income/wealth distributions or in their tails). Important related problems of two-stage estimation arise in tail index regressions for key economic and financial indicators, e.g., financial returns, in the analysis of determinants of heavy tails and financial and economic crises, with regressions run for tail index estimates. Two-stage inference procedures also form the basis of methods for estimation of copula dependence structures, e.g., models of the interdependence of economic and financial markets and their financial contagion. At the first stage of the methods, the parameters of the one-dimensional distributions of the economic and financial indicators considered are estimated, and the second stage is based on maximum likelihood estimation of copula dependence parameters.
The above discussion emphasizes the theoretical and applied significance and relevance of research in the following two directions and related areas on which the work on the project will be concentrated:
1. Development of econometric toolbox for robust statistical and econometric analysis of hypotheses on two or more parameters in economics and finance with applications in important areas in economics, finance and econometrics, including robust tests for structural breaks in economic and financial markets and in their dynamics; robust econometric analysis of the effectiveness of economic policy measures and treatment effects; robust tests of joint hypotheses on several regression parameters; cross-country and cross-regional comparisons of the dynamics of socio-economic development and market models; and the analysis of Granger causality in predictive models, among others.
2. Development of new and improved approaches to robust statistical and econometric analysis in two-stage inference methods under endogeneity, dependence, heterogeneity and outliers in models and data. Research in this direction will allow one to unify the analysis based on many methods widely used in econometrics, finance and economics, including the above approaches to instrumental statistical and econometric analysis; identification and assessment of factors and determinants influencing income and wealth inequality both over the whole distributions of income and wealth as well as in their tails; and the analysis and estimation of copula dependence structures for economic and financial markets and financial contagion.
Research on the project will result in the development of a toolbox of new methods and approaches to robust statistical and econometric analysis and hypothesis testing on several parameters of economic and financial models under the analysis, as well as new and improved methods for statistically justified analysis using two-stage inference procedures. A unifying characteristic of the methods and approaches the project concentrates on consists in the use in them of conservativeness properties of test statistics, such as two-sample t-statistics in tests of hypotheses about the equality of two parameters in the analysis of structural changes, treatment effects, and the analysis of the effectiveness of economic policy and regulation measures, and t- and F-statistics in tests of hypotheses on one, two or more parameters of OLS and TSLS regressions and inference using instrumental variables. A further key distinguishing feature of the methods and approaches consists in the provided possibility of the analysis of many important problems in various areas of research in economics and finance from a unified point of view. As discussed above, these areas include instrumental methods of inference in economics and finance under problems of endogeneity in the models and variables considered; methods for estimation of copula dependence structures that characterize the interdependence and financial contagion properties of economies and markets under the analysis; cross-country and cross-regional comparisons of socio-economic development and its components, such as income distribution and inequality; analysis of factors affecting the dynamics of key economic and financial indicators, including financial returns, foreign exchange rates, inflation and income inequality; as well as the related problems in the analysis of Granger causality in predictive models, among others.
The project will provide a wide range of applications of the developed new and improved methods of robust statistical and econometric inference under the problems of endogeneity, dependence, heterogeneity and outliers in the data and models under study. In particular, we plan to conduct a wide-scale critical analysis of the results in the literature on key models in economics and finance analyzed on the basis of instrumental regression methods, such as the standard models and estimates demand functions in economic and financial markets; Phillips curve and Okun's law models relating the dynamics of inflation, employment and economic growth; models of the effects of education level on wages and incomes (see the work of Nobel laureates in economics this year J. D. Angrist, G. W. Imbens and their co-authors), and others. The work on the project will further provide a wide range of empirical applications of the developed robust statistical and econometric methods in the analysis and comparisons of the dynamics of socio-economic development of modern developed economies and emerging markets, including the markets in Russia, its regions and the countries of the former USSR. These applications will include robust analysis of structural breaks in the dynamics of the markets in consideration due to crises and other structural shocks such as the onset of the COVID-19 pandemic; statistical analysis and estimation of copula structures of interdependence and financial contagion in the economies under study, with conclusions about the spread of crises, as in the case of the global financial crisis of 2008; as well as the development and applications of predictive models for the dynamics of key economic and financial indicators of developed and emerging markets and economies, including economic growth, financial returns, foreign exchange rates, inflation rates and wages. An important component of the empirical results and applications will be provided by large-scale cross-country and cross-regional comparisons of key economic, financial and social indicators of the dynamics of economies under consideration, including GDP growth, income distributions and inequality and the factors affecting them.
The methods and approaches used in the research on the project will include, among others, the probabilistic results on conservativeness properties of one- and two-sample t-tests for the means of small samples of independent heterogeneous Gaussian observations in the literature (see Bakirov, 1989, 1998, Bakirov and Szekely, 2005, Ibragimov and Muller, 2010, 2016, and references therein), as well as new results on their refinements and generalizations, including generalizations to the case of heavy-tailed observations, inequalities for sums of symmetric and unimodal random variables, and generalizations to the important case of F -tests for joint hypotheses and probability inequalities for quadratic forms. The methods and approaches will also use new results on justification of the applicability of robust methods based on the conservative properties of test statistics, including t-statistics, in a wide range of important problems. These problems include the analysis of nonlinear dependence, clusters of volatility and (non-)efficiency of economic and financial markets modelled using GARCH time series and copulr dependence structures; cross-country and cross-regional analysis of socio-economic development, including income distributions and inequality; problems of estimation and forecasting of economic and financial dynamics based on predictive regressions; as well as instrumental variable methods of statistical and econometric analysis under the problems of endogeneity, with applications to the analysis of demand functions in economics and finance, the study of factors affecting financial returns, foreign exchange rates, income distributions and inequality and their heavy-tailedness properties.
The work on the project will further result in the development of research and educational contributions and collaborations of an active group of Russian and foreign scientists in the areas of statistical, econometric and computer analysis and mathematical and computer modelling of financial and economic processes and markets. The project's contributions to research and education will include, inter alia, research and work in the following important areas and the achievement of the following expected results.
- Development of a research and educational center in econometrics, business analytics and related fields (Center for Econometrics and Business Analytics - CEBA, https://ceba-lab.org) on the basis of St. Petersburg State University with its further funding from the internal resources of the university within its organizational structure.
- Development of research and educational cooperation between members of the research group, the Center for Econometrics and Business Analytics - CEBA, St. Petersburg State University, Imperial College Business School, the University of Sydney Business School, leading researchers in econometrics, statistics, computer science and related fields in Russia and abroad, including members of the Project's Scientific Council (S. Anatolyev, Russian School of Economics; U. K. Müller, Department. of Economics, Princeton University; P. C. B. Phillips, Department of Economics, Yale University) and other world-class experts in the fields of the project and related areas.
- Development and teaching of educational courses for students and young researchers at St. Petersburg State University and other higher educational institutions of Russia; development and teaching of courses in econometrics, business analytics, computer and mathematical modelling of financial and economic markets and phenomena, computer and statistical analysis of large databases of financial and economic data and analysis of financial and economic time series at St. Petersburg State University and other universities of the Russian Federation. Organization of author's short-term, semester and annual courses taught by leading Russian and foreign scientists in the Center in the above areas (https://ceba-lab.org/courses).
- Organization of international research conferences, seminars and summer schools on the basis of the Center for Econometrics and Business Analytics - CEBA, St. Petersburg State University and other higher educational institutions of the Russian Federation with participation of leading scientists and researchers working in financial markets, government agencies and IT, and leading researchers in the field of the project in Russia and abroad. Organization of continuous work of weekly online seminars in econometrics, business analytics and related areas at CEBA (https://ceba-lab.org/seminars); organization of annual iCEBA conferences at the Center (https://ceba-lab.org/conference22, https://ceba-lab.org/conference).
- Development and statistical and computer analysis of a large-scale database on the key economic and financial indicators of the socio-economic development of various economies, including the markets in Russia, its regions and the countries of the former USSR analyzed in the course of the project.
- Development of new and improved methods for robust statistical and econometric analysis of large databases of dependent and heterogeneous financial and economic data, with a focus on the development of robust statistical and econometric methods for analyzing hypotheses on two or more parameters, hypotheses of equality of parameters of economic and financial models and joint hypotheses, as well as statistically valid two-step econometric inference procedures. Applications of new and improved methods developed during the project work in econometric models of instrumental variable analysis, cross-country and cross-regional comparisons of models of socio-economic development, tests of structural shifts due to crises and their spread, and other important problems.
- Implementation of the developed new robust statistical and econometric methods in widely used software packages for computer, mathematical and statistical modelлing and analysis such as R, Python and Stata; development and application of new software based on them.
- Development and applications of the new robust approaches in statistical, mathematical and computer modelling and forecasting of the dynamics of key financial and economic variables and indicators in developed and emerging economies and markets, including markets in Russia, its regions and countries of the former USSR, as well as in modelling and forecasting the impact of crises and their spread and financial contagion phenomena on the economies and markets considered.
- Adaptation of modern and newly developed robust methods of statistical, mathematical and computer modelлing and robust statistical analysis of financial and economic markets for efficient implementation on high-performance (parallel and distributed) computing systems and their software.
- Application of modern computer and information technologies, high-performance (parallel and distributed) computing systems and their software in modelлing and robust statistical analysis of large databases on financial and economic markets affected by crises, their propagation and financial contagion.
- Application of modern and newly developed methods of statistical, mathematical and computer modelling and robust statistical analysis in the study of the dynamics of economic development in Russia, post-Soviet economies and emerging markets affected by the on-going COVID-19 pandemic, financial and economic crises, financial contagion and interdependence. Application of research results in the development of economic policy recommendations.
- Publication of the results obtained over the course of research and work on the project in papers in leading peer-reviewed foreign and international scientific journals and their presentation at leading international conferences and seminars in economics, finance, econometrics and statistics, computer and mathematical modelling, computer and information technologies and at meetings of international professional societies in these areas.
- It is expected that the project will provide a large-scale contribution to the development of international cooperation in scientific research and education in the fields of statistical, mathematical and computer modelling of financial and economic markets and processes, including crises and financial contamination, and statistical and econometric analysis of large databases of financial and economic data. It will include the development and applications of modern methods of robust econometric, statistical, mathematical and computer modelling and analysis of large databases on financial and economic markets; including databases on Russia, post-Soviet economies and emerging markets; as well as interdisciplinary and broad international scientific and educational cooperation in the project areas. The project will contribute to the development of key priority areas of development in Russia, including the development of information and computer systems and technologies, the use of high-performance computing systems in the analysis of financial and economic markets, the development of software for robust statistical and computer analysis and modelling, education of students and young scientists and the development of recommendations on statistically justified economic and financial analysis, economic policy and anti-crisis measures in Russia and abroad.

REPORTS