Psychological biases and their impact on the modeling of decision-making variables under conditions of uncertainty were explained over the years using conceptual probabilistic hypotheses; But the existence of uncertainty in such models makes these models not have the necessary efficiency; Because uncertainty causes a kind of asymmetry and was not considered in the structure of these models. Therefore, the main goal of this research is to provide a possible framework for clarifying managerial biases in conditions of uncertainty based on the Monte Carlo Markov chain method and the Metropolis-Hastings algorithm. The research method is within the framework of Bayesian approaches and a Chula normal distribution, where the expected value, overall risk, downside risk, 1% of the value at risk and the expected drop for the three modes of rational, overconfident and underconfident managers are based on Markov chain Monte Carlo and Metropolis-Hastings algorithm are randomly calculated using a sample of 150,000 portfolio returns and cash flow growth rates. The results of the research showed that overconfident managers estimate their expected values more than rational managers; But downside risk, value at risk, and expected decline are underestimated decision variables. On the other hand, managers with less confidence than rational managers underestimate their expected values; But downside risk overestimates value at risk and expected decline than rational managers.