DTA

Corso Ordinario Secondo Livello

GERACI, NICOLÒ

etd-10142024-230142

Predicting yesterday: an evaluation of Dynamic Factor Models for nowcasting

Classe Scienze Sociali

SCIENZE ECONOMICHE E MANAGERIALI - SCIENZE ECONOMICHE E MANAGERIALI

relatore Prof. CORSI, FULVIO
Tutor Prof. MINA, ANDREA
Presidente Prof. IRALDO, FABIO
Membro Dott.ssa CANTARELLI, PAOLA
Membro Dott. GIACHINI, DANIELE
Membro Prof. BARONTINI, ROBERTO
Membro Prof. BOTTAZZI, GIULIO
Membro Prof. CINQUINI, LINO
Membro Prof.ssa VAINIERI, MILENA
Membro Prof. TENUCCI, ANDREA
Membro Prof. ROVENTINI, ANDREA
Membro Prof. MONETA, ALESSIO
Membro Prof. DI MININ, ALBERTO
Membro Prof. TURCHETTI, GIUSEPPE

• forecasting
• nowcasting
• time series

25/11/2024;

parziale

Nowcasting involves predicting the recent past, present, or near future of key economic indicators like GDP or inflation using real-time or high-frequency data. It has gained im- portance due to delays in official data releases, such as GDP figures, which are typically available six weeks after the reference quarter. By using more frequent data like monthly industrial output or weekly labor statistics, nowcasting fills this gap, offering early insights into economic activity. Dynamic Factor Models (DFMs) are a popular tool in this context, condensing large datasets into a few common factors that capture key dynamics. Origi- nating from the work of Geweke and Sargent in the 1970s, DFMs have become central to economic forecasting. Their state-space representation allows for Kalman filtering, which handles mixed-frequency and missing data, making them ideal for nowcasting. However, interpreting the extracted factors can be challenging. This thesis explores DFMs applied to nowcasting and tests popular estimation methods through Monte Carlo simulations and an empirical application using data from the Federal Reserve Bank of Saint Louis.

Nome file	Dimensione
Ci sono 1 file riservati su richiesta dell'autore. Contatta l'autore