Gustavo Fruet Dias

PhD Student

Room number: W317
Tel: +44 20 7882 8844
Fax: +44 20 8983 3580
Email: g.fruetdias@qmul.ac.uk

Research keywords: Forecast macroeconomic variables using large datasets, Factor models, Iterative Estimation and Heteroscedastic models
Research supervisors: George Kapetanios, Andrea Carriero

My research is focused on applying Iterative estimation methods on different econometric models, as a way to improve forecast performance. We believe those methods are quite competitive when compared to the ones usually applied in the literature, since they produce consistent and under some scenarios, efficient estimators. My PhD dissertation will be split in two parts:

In the first part, we aim to improve forecast of macro economic variables using large datasets. It is widely reported by the literature, that when the number of variables in a system increases, the forecast performance decreases dramatically. Following that, we propose the use of iterative estimations methods to overcome the “curse of dimensionality”, and thus improve forecast accuracy.

In the second part, we aim to improve forecast accuracy of financial assets risk premium