Abstract

High dimensionality in economic modeling is tackled by using econometric techniques that overwhelm classical assumptions about observed variables. We refer to moment structure models, such as factor models. In order to capture more flexible dependence structures we exploit copula functions instead of gaussianity. The proposed methodology is applied to high dimensional dataset concerning several economic sectors (such as tourism, housing market, industrial production, etc.). New timelier indices based on the extracted latent factors are provided. Our goal is to anticipate official reports on the health of the sectors or of the whole economic system.

Projektleiter

Enrico Foscolo

Forschungs­schwerpunkte

Quantitative methods and economic modeling

Schlagwörter

Stochastic models, Empirical economics, Time series, Decision theory, Statistical inference, Rank-based statistics, Computational statistics, Copula function, Data analysis, Dependence models, Financial markets, Quantitative risk management

Start- und Enddatum

01.09.2012
31.08.2015