EFFECT OF REGULATION ON ANALYSTS' FORECAST ERROR IN THE BRAZILIAN CAPITAL MARKET

Abstract

This article aimed to understand whether regulation affects the analyst's task of predicting information from companies operating in the capital market in Brazil, listed and active with the Brazilian Securities and Exchange Commission (CVM), in the period from 2010 to 2020. The Panel Analysis approach was used, which emerges as a statistical instrument in the evaluation and understanding of temporal dynamics and individual variations in financial contexts and which allows a statistical approach that examines data over time and between different units of observation. This way it can handle longitudinal data, allowing the identification of patterns over time. The sample obtained was related to 176 companies listed on B3, in the period from 2010 to 2020, with 4,031 initial observations. The model also has its analysis segregated by 10 sectors which inform the companies that operate with the Securities and Exchange Commission, namely: (1) Industrial Goods; (2) Communication; (3) Cyclical Consumption; (4) Non-cyclical consumption; (5) Financial; (6) Basic materials; (7) Oil & Gas; (8) Health; (9) Information technology and (10) Public utility. The information was obtained from the Economática database. The results do not allow us to refute the proposed hypothesis that regulation influences analysts' error in forecasting absolute error and positive error. Thus, it points out that analysts' forecasts may have a magnitude of error far from what was realized and that they may present a confirmation bias, which implies overestimating the most regulated companies.