CASPIAN JOURNAL
MANAGEMENT AND HIGH TECHNOLOGIES
USING ARIMA MODELS FOR FORECASTING OF CURRENCY EXCHANGE RATE
 |
Read | Pilyugina A.V., Boyko A.A. USING ARIMA MODELS FOR FORECASTING OF CURRENCY EXCHANGE RATE // Caspian journal : management and high technologies. — 2015. — №4. — pp. 249-267. |
Pilyugina A.V. - Ph.D. (Economics), Associate Professor, Moscow State Technical University named after Bauman, 5 (building 1) 2-ya Baumanskaya St., Moscow, 105005, Russian Federation, pilyuginaanna@mail.ru
Boyko A.A. - post-graduate student, Moscow State Technical University named after Bauman, 5 (building 1) 2-ya Baumanskaya St., Moscow, 105005, Russian Federation, boiko_andrew@mail.ru
ARIMA model and method of its implication for exchange rate of the US dollar to the Russian Federation rouble short-term forecasting are described. Time series classification from stationarity side is given. Procedure of time series stationarity checking using augmented Dickey-Fuller test is explained. It is shown that given time series of monthly average exchange rate of the US dollar to the Russian Federation rouble is an integrated time series of the first order. Differencing operation makes monthly average rate a weak stationary time series. For determining autoregression order QUOTE and moving average order QUOTE optimization mesh search algorithm is used. Bayesian information criterion QUOTE and Akaike information criterion QUOTE are used as optimization criterions. Using these criterions an QUOTE model was chosen. Additionally, an adaptive model identification algorithm was proposed with QUOTE and QUOTE orders and corresponding coefficients refining upon receipt of every new time series sample. However, forecast error estimation showed that for mean average percentage error QUOTE minimization the simplest ARIMA models should be used, more specifically, QUOTE and QUOTE models.
Key words: валютный курс, прогнозирование, временной ряд, стационарный относительно детерминированного тренда ряд, стационарный относительно взятия разностей ряд, модель QUOTE , байесовский информационный критерий QUOTE