We consider Bayesian estimation of Value-at-Risk (VaR) using parametric Product Partition Models (PPM). VaR is a standard tool to measure and control the market risk of an asset or a portfolio, and it is also required for regulatory purposes. We use PPM to provide robustly Bayesian estimators of VaR, remaining in a Normal setting, even in presence of outlying points. We consider two different scenarios: a product partition structure on the vector of means and a product partition structure on the vector of variances. In both frameworks we obtain a closed-form expression for VaR. The results are illustrated with an application to a set of shares from the Italian stock market. The methodology and the obtained results are described in details in Bormetti et al. (2009).
Bayesian Analysis of Value-at-Risk with Product Partition Models / Delpini, Danilo; Bormetti, G; De Giuli, M. E.; Tarantola, C.. - (2009), pp. 151-156. (Intervento presentato al convegno S. Co. 2009. Sixth conference. Complex data modeling and computationally intensive statistical methods for estimation and prediction tenutosi a Politecnico di Milano, Milano nel 14-16 Settembre 2009).
Bayesian Analysis of Value-at-Risk with Product Partition Models
DELPINI, Danilo;
2009-01-01
Abstract
We consider Bayesian estimation of Value-at-Risk (VaR) using parametric Product Partition Models (PPM). VaR is a standard tool to measure and control the market risk of an asset or a portfolio, and it is also required for regulatory purposes. We use PPM to provide robustly Bayesian estimators of VaR, remaining in a Normal setting, even in presence of outlying points. We consider two different scenarios: a product partition structure on the vector of means and a product partition structure on the vector of variances. In both frameworks we obtain a closed-form expression for VaR. The results are illustrated with an application to a set of shares from the Italian stock market. The methodology and the obtained results are described in details in Bormetti et al. (2009).I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
