چکیده:
Future oil prices are one of the important factors that influence the policies and planning of governments, international organizations, and companies. Modeling oil prices to forecast future oil prices through econometric methods can be an important way forward, such that specialists have been able to gain valuable experiences in this matter. In this article, the performance of the international oil market has been examined to identify the necessary variables for modeling. Then, the data were examined in terms of statistical properties and the long-term relationship between ITW oil prices and other variables was estimated. Then, autoregressive, Maximum Corrective Error (MCE), and Autoregressive Distributed Lag (ARDL) models were estimated to forecast ITW oil prices. Finally, prediction accuracy measurement criteria were used to identify the desired model with the minimum error. The results show that the autoregressive distributed lag model provides the best prediction accuracy.
خلاصه ماشینی:
Then, the data were examined in terms of statistical properties, and the long-term relationship between ITW oil price and other variables was estimated.
Subsequently, autoregressive, error correction (MCE), and autoregressive distributed lag (LDRA) models were estimated for forecasting ITW oil prices.
Kaufman (1991) considered the oil price as a function of fundamental oil market variables, including OPEC supply, the number of days of storage level coverage 1, and the percentage of production to total OPEC production capacity.
Trend of oil supply and demand in the DCEO region (Refer to the page image) Chart 4 shows the trend of change in inventory levels in million barrels per day.
Variables and Data Dependent and explanatory variables identified in the oil market: (TW) :Real oil price; (YD) :Number of days of cover; (OS) :OPEC oil supply level; (ND) :Oil demand level of non-DCEO countries; (RSU) :Real effective dollar exchange rate; iD 4,3,2-i seasonal variable.
In the SLO approach, the ITW oil price is considered in the form of a linear-logarithmic pattern as a function of OPEC supply, non-DCEO demand, adjusted number of days, and the real effective dollar exchange rate, and the estimation results are provided in Table No. 1.
Given the points mentioned, for seasonal forecasting of ITW oil prices, autoregressive, error correction (MCE), and LDRA models, which differentiate between short-term and long-term relationships, are used.
The error correction and LDRA patterns for forecasting are expressed as follows: Except for the lagged residual obtained from the long-term relationships between oil price and independent variables, Y is the vector of independent variables.
(
, 
, 
)
