Projeção do índice s&p 500: uma aplicação do método de regressão dinâmica

Abstract

This study aims to analyze the historical evolution of the S&P 500 and to forecast the future behavior of the index, covering the period from January 2010 to May 2026. The S&P 500 is one of the leading indices of the global stock market, comprising large-capitalization companies of major worldwide relevance. The analysis was carried out using the Dynamic Regression method (DR), which applies a set of diagnostic tests to reach the final model. The data were obtained from Stooq, for the S&P 500 series, and from the Federal Reserve Economic Data (FRED), for the variables Federal Funds Rate (FEDFUNDS), Consumer Price Index (CPI), and the ten-year U.S. Treasury yield (GS10). The estimation identified a dynamic component in the evolution of the index, expressed by its relationship with its own value in the previous month, as well as a positive relationship with the CPI. The variables FEDFUNDS and GS10 were not statistically significant and therefore were not included in the final equation. The model showed a good fit to the data, with an adjusted R2 of 99.33% and a mean absolute percentage error (MAPE) of 3.15%. Based on the

estimated model, the index was projected for the following twelve months through the auto- projection of the CPI, resulting in an upward scenario for the S&P 500.