Pronóstico del Tipo de Cambio USD/COP mediante modelos econométricos y su aplicación en la fijación de precios para exportaciones con pagos diferidos

Abstract

Problem: The USD/COP exchange rate represents a significant financial risk for Colombian companies operating in international markets with deferred payments, generating uncertainty about future cash flows and potential losses in profit margins. Objective: To comparativel evaluate different econometric models for forecasting the USD/COP exchange rate and their applicability in pricing exports negotiated in dollars with deferred payments exceeding 30 days. Method: The USD/COP exchange rate time series was analyzed over the period 2020-2025, estimating four econometric specifications: ARIMA, SARIMA, EGARCH, and the hybrid SARIMA-EGARCH model with generalized error distribution (GED). Model selection was based on AIC and BIC information criteria, error metrics (RMSE, MAE, MAPE), and residual diagnostics, complemented by Monte Carlo simulations and a practical export pricing case study. Results: The hybrid SARIMA-EGARCH model demonstrated the best predictive performance by integrating the temporal structure of SARIMA with asymmetric volatility modeling from EGARCH, producing more realistic confidence intervals. Projections place the exchange rate in the $4.200-$4.300 COP per dollar range in the short term. Empirical validation through the Aroma Andino S.A.S. case study confirmed the model's utility, with forecast errors decreasing progressively from 3.59% at 30 days to 2.10% at 90 days. Discussion: Although the SARIMAEGARCH model constitutes a robust tool for exchange rate risk management, its use as the sole reference for export contract valuation should be complemented with financial hedging instruments, fundamental macroeconomic analysis, and consideration of exogenous variables not captured by univariate models.