A Stochastic-Goal Mixed-Integer Programming approach for integrated stock and bond portfolio optimization

We consider a Stochastic-Goal Mixed-Integer Programming (SGMIP) approach for an integrated stock and bond portfolio problem. The portfolio model integrates uncertainty in asset prices as well as several important real-world trading constraints. The resulting formulation is a structured large-scale problem that is solved using a model specific algorithm that consists of a decomposition, warm-start, and iterative procedure to minimize constraint violations. We present computational results and portfolio return values in comparison to a market performance measure. For many of the test cases the algorithm produces optimal solutions, where CPU time is improved greatly.

► We consider a Stochastic-Goal Mixed-Integer Programming portfolio model.
► The portfolio selection problem involves different investments and financial factors.
► The formulation is a large-scale problem that employs a model specific algorithm.
► The algorithm consists of a decomposition, warm-start, and iterative procedure.
► We find increased CPU time and model performance from the ad hoc procedure.