Utility maximization in a stochastic affine interest rate and CIR risk premium framework: a BSDE approach (Q6098178)

This paper studies the problem of portfolio optimization in a market with stochastic interest rates, for agents reporting expected utility preferences, within the framework of affine models for the underlying processes using the methodology of backward stochastic differential equations (BSDEs). This approach allows for the study of the non Markovian case, as well as cases such as the Heston or the Vacisek model. The analysis focuses on the case of the power and the logarithmic utility functions. For both cases the corresponding BSDE that characterizes the optimal strategy is constructed and particular solutions are obtained in terms of auxiliary ODE systems that are explicitly solved. Under certain assumptions it is shown that the obtained solutions are the unique solutions of the corresponding BSDE. The BSDE related to the power law utility case is nonlinear (and explicitly solvable upon reduction to a set of nonlinear ODEs reducible to the Riccatti equation) while for logarithmic utility case it is linear (and explicitly solvable upon reduction to a set of linear ODEs). The paper concludes with a numerical illustration of the results.