A simulation environment for discontinuous portfolio value processes (Q2722288)

This paper deals with the general structure and analytical framework behind an internally developed simulation environment linking the classical defaultable bond analysis to movements of risky market representative indexes such as the JP Morgan emerging market bond index. The development allows an accurate representation of the factors affecting portfolios subjected to jumps. The system implements a simulator of market shocks driven by a canonical Wiener process and by a Poisson shock arrival process whose parameter is allowed to vary over time. It supports a variety of applications from the estimation of the non-parametric portfolio shortfall distribution (relevant for capital allocation and credit risk management) to the market monitoring for potentially unstable markets and, ultimately, is used as a generator of return processes for the solution of optimal portfolio problems in risky markets, using a 3-d efficient frontier. A novel characterization of the intensity rate of the Poisson process, modelling the arrival of shocks to the market, as a function of a credit spread curve estimated in high-risk emerging bond markets, is introduced. The procedure is described and tested on the August 1998 Russian crisis whose impact on liquid equity markets is also estimated.