In recent years Capital at Risk has been brought into the market as a way to minimizing risks in the replacement of the variance in optimal portfolio selection problems. A study was conducted for this work, by utilizing the classical stochastic control methodology on the consequences of using the Capital at Risk measure in a Black-Scholes simple market model and in a Generalized Inverse Diffusion market. Theoretical results were compared to data taken from bolsa de Valores de Colombia, for the cases of Ecopetrol and Isa.