Selection of financial indicators using statistical multivariate analytical techniques

Abstract

The purpose of this research is to provide certified accountants with a methodology to reduce the number of indicators of the status of profitability or accounting risk in order to conduct an analysis of the financial disclosure of companies. This is an experimental, quantitative, and exploratory study that aims to incorporate the contributions of the original set of p financial indicators of companies into another set of q non-correlated variables by applying a statistical technique called Analysis of Principal Components (APC). The findings of this study could be used by accountants to: i) simplify further analyses using a lesser number of components; ii) graphically represent individuals on only two or three dimensions, and iii) observe inter-variable relations in a much more objective manner. Each new matrix of measurements, even of the same variables about the same individuals, requires a new APC, so a given status of profitability or accounting risk could not be reproduced, but instead its variables must be identified every time.