Modelos de series temporales para el pronóstico de la inflación en Ecuador: un análisis de la estacionalidad y fluctuaciones cíclicas.

Abstract

Abstract: Inflation is a phenomenon of great complexity where its analysis and forecasting are fundamental for decision-making and the formulation of economic policies. In this context, the study aims to apply suitable time series models to forecast inflation in Ecuador, specifically the ARIMA and ARCH models focusing on seasonality and cyclical fluctuations. For this purpose, data from the Central Bank of Ecuador from February 2005 to March 2024 were used, and for the methodology application, stationarity tests (Dickey-Fuller and Phillips-Perron) and AIC and BIC information criteria were included for validation, where the results obtained showed that the ARIMA (1,0,1) model demonstrated a good fit and capability to capture short-term trends, and the ARCH (1) model was effective in modeling volatility. Therefore, it is concluded that both models proved to be robust and effective in capturing the seasonality and cyclical fluctuations of inflation, providing a solid basis for economic decision-making and policy formulation in Ecuador.