Use of principal components regression in modelling the return on assets of the enterprise

Abstract

Purpose: The article presents the feasibility and possibility of assessing the investment attractiveness of an enterprise based on the econometric modeling of net profitability of assets.
Discussion: With a wide variety of approaches to identifying and differentiating factors and indicators of the company's investment attractiveness, authors underlined the key role of financial indicators in assessing investment attractiveness and proposed a system of financial ratios of the enterprise - predictors of net return on assets as its main indicator. Under the collinearity of prognostic factors, authors suggest principal components regression which enables to obtain better prognostic characteristics, maintaining reliability and informative modeling.
Results: authors proposed a tool to analyze contributions of predictors of the net return on assets to the assessment of the investment attractiveness; its quality was tested using standard Fisher and Student tests, according to the standard model error and the Akaike and Schwarz information criteria. Results of the empirical evaluations carried out using the Gretl software confirmed the feasibility of this approach to assess the investment attractiveness of the enterprise by potential investors, shareholders and owners to manage the efficiency of the use of capital.