Econometrics and Data Science Methods for Business, Economics and Finance

August 21-25, 2023 in Amsterdam

Faculty

Francisco Blasques is Professor of Econometrics and Data Science at Vrije Universiteit Amsterdam. His research focuses mostly on the theory and practice of dynamic modeling and time-series econometrics.

Siem Jan Koopman is Professor of Econometrics at Vrije Universiteit Amsterdam. His research focuses mostly on time-series modeling using state space and related methods including score-driven models and dynamic factor models.

Meet the lecturers

Course

This summer school will cover fundamental topics in econometrics and data science. The content ranges from predictive and causal methods for time-series analysis, to state-space methods and filtering techniques for high-dimensional datasets. Participants will learn how to design, test and evaluate quantitative models and methods in Business, Economics and Finance.

Given the interdisciplinary nature of the summer school, we will begin with a review of basic methods in econometrics, data science, structural modeling and time series. Practical cases are developed for different purposes in the fields of business, economics, and finance. For each topic, we cover both the theory and methodology, as well as hands-on applications with real data. In particular, we illustrate their use and their importance for all practical purposes, we implement the basic methods in a computer lab, and we assess their performance in a real data setting.

Participants will work in small groups to develop (a) structural models for the support of marketing and pricing decisions in business, (b) designing time series models for macroeconomic forecast, (c) a case on extracting and forecasting signals from noisy business data using the Kalman filter, and (d) a case on incorporating vast data resources for measuring and nowcasting current economic activity.

Covered topics

• Day 1    Time Series Analysis and Forecasting: This workshop covers the following aspects of time-series analysis and econometrics: Basic properties of time series; Estimation and specification of ARMA-type and dynamic ML models; Prediction and impulse response functions; Distributed lags and error correction; Unit roots, integration and co-integration.
• Day 2    Predictive and Causal Data Science: This workshop covers the following aspects of causal data science: Predictive versus prescriptive analytics; Pitfalls of predictive analytics; Methods of causal inference; Data-driven policy evaluation and decision-making.
  
• Day 3    Group work on a data project: All summer school participants work on a real data project within small groups. In the morning, the assignment is introduced. During the day, the groups are working on the project with the support of academic researchers in econometrics and data science.
  
• Day 4    Signal Extraction, Filtering and Scenario Analysis: This workshop covers the following aspects of causal data science: Local level and trend models, with and without regression effects; Filtering, smoothing and forecasting using state space methods; Messy features: missing data, noise and outliers, intervention analysis; Scenario analysis and decision-making. 
  
• Day 5    Dynamic Big Data, Factor Analysis and Nowcasting: This workshop covers the following aspects of causal data science: Principal components, factor analysis, and dynamic factor models; Filtering, smoothing and forecasting using multivariate state space methods; Messy features in high-dimensional settings: missing data, noise and outliers, unbalanced panels; Nowcasting, scenario analysis and decision-making.
  

For more specific information see Course Outline.

Level

This summer course welcomes (research) master students, PhD students, post-docs and professionals from all disciplines and industries (finance, economic policy, business studies) with a quantitative background and who are interested in learning state-of-the art econometrics, data science, and time series methods.

Admission requirements

Basic knowledge of statistical inference and regression analysis. Basic knowledge of programing (R, Python or MATLAB). Formal background in quantitative studies (mathematics, statistics, engineering, business analytics, finance, etc.) is required from students (at the level of a first-year course in a Master study). No formal background in Econometrics or Statistics will be assumed.