Abstract

AYALA CASTREJON, Rey Francisco  and  BUCIO PACHECO, Christian. ARIMA model from 2016 to 2017 term implemented to the peso/dollar exchange rate through temporary sliding windows. Rev. mex. econ. finanz [online]. 2020, vol.15, n.3, pp.331-354.  Epub Feb 10, 2021. ISSN 2448-6795.  https://doi.org/10.21919/remef.v15i3.466.

A wide variety of forecasting in the peso/dollar exchange rate through ARIMA(1,1,1) model from 2016 to 2017 has been done in this work, such model was implemented on the peso/dollar exchange rate is estimated in many different ways through temporary sliding windows. Also, problems of structural change are identified and an optimal adjustment is proposed to the ARIMA(1,1,1) model which allows to give a better forecasting. The empirical evidence highlights the complexity in giving a forecasting with data which specific characteristic is to change through the time, and also with structural change. Procedures are suggested in order to improve the forecasting, for example the use of temporary sliding windows and the structural change proposition. As a conclusion, the 30-day obtained forecast through sliding windows and sliding growing right windows is viable, because with a 95% confidence interval there are 12 out of 30 registers in the rank of real value of the peso/dollar exchange rate.

Keywords : Peso/Dollar Exchange Rate; Forecasting; ARIMA Models; Structural Change.