Analysis Of Beta Coefficients In The Brazilian Stock Market Using Fuzzy Linear Regression Methodology

With the aim of using all the information provided by the market to determine the systematic risk, we intend to continue the study of Terceno et al. (2011, 2014) using fuzzy linear regression to calculate the sectors betas of the Brazilian Stock Market. The analysis with fuzzy regression can be applied which crisp data, uncertain or with a mixture of both. The objective of this work is, precisely, to compare the obtained results using the fuzzy regression with crisp data and uncertain data. After that, we make a comparison with the results obtained by using ordinary least squares. The comparison allows us to determine which of the systems allows a better adaptation of reality. As we will show, fuzzy regression is in many ways more versatile than conventional linear regression because functional relationships can be obtained when the independent variables, dependent variables, or both, are not crisp values but intervals or fuzzy numbers.