Research Projects

  • PRIN 2022: Realized Random Graphs: A New Econometric Methodology for the Inference of Dynamic Networks (Principal Investigator: Giuseppe Buccheri)

  • This project aims to develop a novel econometric methodology to simplify the inference of dynamic network models while maintaining flexibility in parameterization. Inspired by the concept of "realized volatility" from financial econometrics, the methodology involves modeling a time-series of cross-sectional estimates of dynamic network parameters. This approach, termed Realized Random Graphs (RRG), allows for mapping nonlinear state-space graph models into simple linear time-series models, enabling easier estimation and faster convergence rates. The project involves both theoretical and empirical work, including applications in community detection and the assessment of financial policies' impact on systemic risk​.

    Funded by the Ministry of University and Research (n. 2022MRSYB7). Total budget: 217.235 EUR. Budget of the local unit: 125.067 EUR

  • PRIN 2022: Probabilistic Methods for Energy Transition - (Verona Local Unit Supervisor: Athena Picarelli)

    The aim of this project is to couple recent results on optimal control theory with machine learning numerical methods to optimise and evaluate the performances of renewable energy production plans. We consider models where the state dynamics (e.g. the price or the availability of resources) and the payoff (e.g., the production or distribution cost, the realized cash flow, the environmental impact) of a single agent (e.g., energy producer) depends on many different random quantities and on the behavior of other agents. We emphasize that when dealing with renewable energy sources, the production capacity depends on the availability of specific resources, and this introduces both an agent-specific randomness and a common noise that affects several agents (for instance particular atmospheric conditions affecting all the producers of some specific area). Mathematically, this type of problem corresponds to a high (or even infinite) dimensional stochastic optimal control problem whose state variable is of McKean Vlasov type.

  • PRIN 2022 PNRR: Climate Change, Uncertainty and Financial Risk: Robust Approaches based on Time-Varying Parameters (Verona Local Unit Supervisor: Giuseppe Buccheri)

  • This project focuses on the impact of climate change on financial portfolios. It aims to develop methodologies to analyze and manage climate risks within these portfolios, providing tools for regulators, policymakers, and the private sector to transition towards a sustainable economy while maintaining financial stability. The project is structured around three main targets:

    • Propagation of Uncertainty: Investigating how uncertainty affects econometric models used to project portfolio losses.
    • Identification of Climate Risk Drivers: Developing methodologies to distinguish climate risks from traditional financial factors and assessing their impact on asset prices.
    • Propagation Mechanisms of Climate Risks: Studying how climate risks affect portfolios and the financial system using high-dimensional network models.
    The project will use advanced econometric and network modeling techniques, leveraging the expertise of the research team to provide empirical and methodological advances with significant societal societal impact.

    Funded by by the Ministry of University and Research (n. P20229CJRS). Total budget: 269.315 EUR. Budget of the local unit: 63.163EUR.

  • FIS 2021: (Principal Investigator - Roberto Renó)

  • PRIN 2020: Dynamic models for a fast-changing world: An observation-driven approach to time-varying parameters (Verona Local Unit Supervisor: Giuseppe Buccheri)

  • This project aims to develop new econometric models with time-varying parameters. These models are based on the Score-Driven (SD) approach, which allows parameters to evolve over time based on past observations. This approach is advantageous as it enables the parameters' evolution to be driven by actual past shocks, allowing for better understanding and prediction of socio-economic dynamics. The project involves both theoretical and applied research, focusing on the statistical properties of SD models, their connections with Bayesian and Machine Learning methods, and their applications in areas such as macroeconometric models, high-dimensional covariance matrix forecasting, and the econometric analysis of financial markets.

    Funded by the Ministry of University and Research (n. 20205J2WZ4). Total budget: 582.558 EUR. Budget of the local unit: 137.132 EUR

  • PRIN 2017: HiDEA (Principal Investigator: Roberto Renó) Visit the HIDEA Project Page

    The research project HiDEA (Advanced Econometric methods for High-frequency Data) aims at developing new theoretical ideas in the econometrics of high-frequency data, and to apply them to solve open economic and financial problems. In particular, the main theoretical advancements of the project will be: developing the econometrics of stale prices, the econometrics of flash crashes, the econometrics of large cross-sections. The main applications will be: measuring market liquidity from transaction prices, study policy implications for the prevention of mini flash crashes unrelated to market fundamentals, exploiting the informational content in the cross sections of high-frequency data for analyzing market behavior and for systemic risk monitoring, and understanding the behavior and role played by high-frequency traders in the price formation process.