Optimizing Credit Portfolio Using a Quadratic Nonlinear Projection Method

A novel optimization framework through quadratic nonlinear projection is introduced for credit portfolio when the portfolio risk is measured by Conditional Value-at-Risk (CVaR). This framework is formulated in terms of the Lagrange multiplier method subjected under an artificial quadratic error term, which is comparable to the amount or cost of total portfolio adjustment, as the necessary constraint. The route toward the optimal portfolio state can be searched from the initial portfolio state via a continuation process through the maximally three-parameter space described by the total portfolio budget, the increment of the total return or the tolerance of the additional risk, and the total portfolio adjustment cost.