The article introduces a new method called Cohesive Value at Risk to measure credit risk in portfolios. By using this method, researchers were able to create a portfolio optimization model that outperformed traditional Value at Risk models. The new approach takes into account the unique characteristics of the security market, such as excess kurtosis and heavy tail distributions, to select the best portfolio using linear programming. Empirical studies showed that portfolios optimized using Cohesive Value at Risk yielded better results compared to those optimized using traditional methods.