The present article solves the classic problem of optimization of investment portfolios, using the model of average-variance and proposing a way to calculate the volatility through the generalized autoregressive conditional heteroskedasticity (GARCH) models. The problem is solved through a bio-inspired metaheuristic, called artificial bee colony (ABC), whose objective is to reduce the computational execution times present in other solutions. The results were counteracted by a previous work, solved with Lagrange multipliers, finding a similar investment boundary, but with a notably lower reduction in execution time. Finally, reference is made to future work within the area of computer finance. (C) 2017 Universidad ICESI. Published by Elsevier Espana, S.L.U.