Econometric models with discrete dependent variable in portfolio analysis

Abstract

Purpose: the authors substantiate the need for the use of econometric models with a discrete dependent variable for modeling investment decisions in the stock market. Discussion: Sharpe’s use of one-factor regression models in portfolio analysis allowed to update the theory of portfolio investment with new results. The authors created the risk structural idea and introduced the portfolio beta concept. Likewise the authors marked that in new econometric models the dependent variable is discrete. In many cases, it is convenient to represent the dynamics of the financial assets profitability in the discrete time series form. The authors justifies this way the idea of applying these new models in portfolio investment tasks. Results: the authors considered the possibility of using the binary choice econometric model for modeling the profitability of a market asset. Also the authors obtained the formulas for calculating the yield and variance of the asset on the basis of this model. The authors offer to use this derived formulas in the optimal portfolio of securities construction.