Pronóstico de la tasa de cambio en Colombia (TRM) utilizando los modelos ARIMA-GARCH durante el período 1992-2025

Abstract

The economic figures forecast as exchange rate or the consumer price index are an essential tool for economic authorities, based on economic research analysts' expectations make decisions with effect in all society. This research has as objective to determinate a useful econometric model to forecast the Colombian exchange rate USD/COP (TRM). For this reason, it developed quantitative research applying ARIMA-GARCH models using time series data for TRM between 1992 and 2022. The results suggest that the model presented in this work achieves adequately filtered the information contained in lags using Q-Stat test, without simple autocorrelation problems at 5% in statistical significance. This permits us to establish that the estimated parameters for forecasting the TRM are significant. Consequently, it is concluded that the model has predictive capacity to forecast the exchange rate between the Colombian peso and the US dollar in the short term.