Dependent Default Modeling through Multivariate Generalized Cox Processes
By: Djibril Gueye, Alejandra Quintos
Potential Business Impact:
Predicts when things will break down together.
We propose a multivariate framework for modeling dependent default times that extends the classical Cox process by incorporating both common and idiosyncratic shocks. Our construction uses c\`adl\`ag, increasing processes to model cumulative intensities, relaxing the requirement of absolutely continuous compensators. Analytical tractability is preserved through the multiplicative decomposition of Az\'ema supermartingales under assumptions that guarantee deterministic compensators. The framework captures a wide range of dependence structures and allows for both simultaneous and non-simultaneous defaults. We derive closed-form expressions for joint survival probabilities and illustrate the flexibility of the model through examples based on L\'evy subordinators, compound Poisson processes, and shot-noise processes, encompassing several well-known models from the literature as special cases. Finally, we show how the framework can be extended to incorporate stochastic continuous components, thereby unifying gradual and abrupt sources of default risk.
Similar Papers
Forward Performance Processes under Multiple Default Risks
Mathematical Finance
Helps investors make more money with risky investments.
A Dynamic Factor Model for Multivariate Counting Process Data
Methodology
Helps understand how students solve hard problems.
Towards modelling lifetime default risk: Exploring different subtypes of recurrent event Cox-regression models
Risk Management
Helps banks better guess loan failure risk.