Using Econometric Modeling in Likelihood Assessing of Investment Activity Risks

Abstract

The article substantiates the necessity of inclusion of random future variables in econometric models. It is considered various concepts of types and descriptions of their probability different from the frequency probability. The example of the use of the subjective Bayesian probability to assess the risks in insurance activities is provided. It is considered how to improve the assessment of the parameters of auto-regression models by including f-lag (future expected) variables. The assessment model of insurance processes allows to estimate probabilities of future expected events and their influence on errors and risks of management decisions by subjective statistics methods and Bayesian networks.