As loans are the main earning assets for banks, Iranian banking sector has always been active in financing businesses through this means. Loans are considered the dominant means of finance in Iran but the evaluation process on allocation of loans is mostly retrospective. This has caused a considerable amount of default during decades. In this paper, we have focused on introducing a method of optimisation for commercial banks to create their optimum loan portfolios. To achieve this goal, we have examined different methods and we have used data on 21 industries, loan data from one of the large scale Iranian commercial banks and proposed a new method with a fresh approach to risk and return. The findings of the present study shows that the semi-variance or undesired variance method and the adjusted standard deviation to estimate the banking optimised portfolio is the most accurate technique.