|
After the stabilization of inflation rates in Brazil, Monetary Correction for financial data was extinguished. Since then, the distinct reflections of prices variations on monetary and non-monetary variables presented on financial statements are no longer considered. From 2004 to 2013, however, the cumulative inflation reaches 226%. In this situation, questions about the reliability of the greatest tools we have in the finance research world - financial data of companies - emerges.
Our objective is primarily to examine whether the recognition of inflationary effects cause significant differences in financial indicators commonly used. Moreover, aim to analyses if this difference also causes changes in estimators and significance of statistical models present in empirical studies of finance.
After the stabilization of The Real Plan, Brazilian inflation rates reached stable levels and the law number 9.249/95 vetoed the use of any monetary adjustment policy for issued balances sheets. After this milestone, some studies as the ones made by Bonizio e Vicente (2001), Gabriel, Neto e Corrar (2005), Ambrozini (2006), Oliveira, Marques e Canan (2007) and Moribe, Panosso e Marroni (2008) dedicated themselves to analyze the impacts that the end of Monetary correction had on accounting data.
We made two different samples: one composed by financial data originally issued by companies and the other one by financial indicators corrected based on the precepts of the extinct Monetary Correction.
First, our study compared financial indicators frequently used on research and financial evaluations, made with the two different databases (original and adjusted). Moreover, we compared panel regression models estimated with original and adjusted variables, to check for changes in coefficients.
Statistical tests showed that financial indicators such as ROI, Asset Turnover, Debt and Market-to-book are significantly higher when we do not consider the effects of inflation. In addition, the panel regression models, when adjusted, had higher predictable power (greater Rē) and representative changes of significance on the variables and coefficients.
|
