Un interessante articolo su un’applicazione empirica di un algoritmo standard di Machine Learning (PCA) ad un problema di mercato: il rischio politico implicito nel 5y Italian Credit Default Swap.

Italian CDS Towards Elections

As Italian elections loom, we provide an empirical application of a standard Machine Learning algorithm (PCA) to the 5y Italian Credit Default Swap in order to highlight the political risk priced by the market.
In particular, we apply Principal Component Analysis on a basket of 10 European CDS to emphasize the common factors driving the sovereign credit evolution, using data from December 2011 onwards. We interpret the part of the CDS price not explained by the PCA common factors, as the implied political risk.

Si prega di compilare il seguente modulo per accedere al download gratuito
“Italian CDS towards Elections”