Linearity Verification of the Effect of Selected Financial Indicators on Bond Rating Assessment

The paper is focused on examining and modelling the relationship between selected financial indicators and credit rating assessment. The influence of rating agencies is significant, mainly because their risk assessment is incorporated in regulatory provisions in many countries. As a result, it is essential to understand rating, the ways how it is measured and main factors of rating assessment. In this study, we identified and chose the potential financial indicators for classification purposes and then we subjected them into linearity verification. Firstly, we found the appropriate fractional polynomial functions and we conducted the sensitivity analysis to identify the influential observations driving the non-linearity. Afterwards, we carried out the stability analysis to assess the sensitivity of non-linear functions selected on small change of data to conclude if the (non)linearity is recommendable