Credit Default Swap Spreads Modeling and Forecasting with a Stochastic Square-Root Three-Factor Model

In this study, we consider the CIR3 model, a three-factor stochastic model with correlated trends and volatilities for modeling and forecasting credit default swap (CDS) spreads. After recalling existence and uniqueness results, together with a generalized Feller condition to ensure positivity, we use a Lamperti-type transform to rewrite our SDEs system in a form in which the stochastic part of the leading process is uncorrelated with those of its mean and volatility processes. Finally, we calibrate the model through the estimating function approach for ergodic diffusions and simulate the CDS prices by discretizing our (transformed) system. These findings contribute to a deeper understanding of stochastic models with correlated trends and volatilities, with applications in pricing, trading and risk assessment.