A “Plug-and-Play,” Goal-Based, Dynamic Asset Allocation Model with Genetic Algorithms

This work aims to reconcile theory and practice by proposing a dynamic model of asset allocation that allows, with the help of genetic algorithms, the creation of reasonable portfolios for investors chasing a variety of objectives over multiple time horizons. Our results are interesting and confirm the efficiency of the model, due to its good convergence towards the space of best solutions in a limited computation time. The methodology may be applied in the fields of: 1) wealth management, to define approaches that are useful for investors whose wealth must be invested to offset future liabilities; and 2) pension funds, to define—based on an individual’s pension needs—personalized investment dynamics that are useful for meeting retirement needs.