Previsão do preço de fechamento de uma ação listada na bolsa de valores B3 (Brasil, bolsa, balcão) com análise preditiva e linguagem R

Abstract

The article presents a predictive analysis of the closing price of a share of a company listed on B3 (Brasil, Bolsa, Balcão). The motivation was to find out the potential for appreciation or devaluation of Itaúsa's share on the following day. For this, the Linear Regression models adjusted with the Generalized Least Squares (MQG), Random Forest (MRF) and XGBoost (XGB) method were compared in nine pairs of series extracted from the historical series, in which the first element of the pair is the training series and the second, with observations consecutive to the first, is the test series. The performance metric used was the Root Mean Squared Error and the median of the forecast errors. The MQG model obtained the best performance among the other models. With the chosen prediction model, it will be possible to outline new strategies for investing in stocks.