КОРРЕЛЯЦИОННО-РЕГРЕССИОННЫЙ АНАЛИЗ И ПРОГНОЗ ПРОИЗВОДСТВА НЕФТЯНОГО БИТУМА В РОССИИ

The purpose of the study is to identify economic factors associated with changes in the volume of oil bitumen production in the Russian Federation and assess the possibility of using them to predict production output. The information base of the study was compiled by statistical data for 2010-2021, including indicators of oil bitumen production, average product price, dollar exchange rate, foreign trade, oil prices, production capacities, inflation, as well as financing for the repair and construction of highways. The paper uses methods of statistical analysis, correlation analysis, multiple regression and scenario modeling. At the first stage, nine factors are considered, potentially related to the dynamics of production. The calculation of paired correlation coefficients showed the closest positive relationship between the volume of production with the dollar exchange rate (r = 0.87) and the amount of financing for road construction (r = 0.83). For subsequent modeling, the average price of oil bitumen, the dollar rate and the amount of funds allocated for the repair and construction of highways were selected. The constructed linear model has a coefficient of determination R² = 0.944 and is statistically significant in general. Exponential, logarithmic and power forms of dependence are additionally considered. The best values for coefficient of determination, standard error and mean approximation error are obtained for the power model.Scenario calculation with an increase in road construction financing by 10 % while maintaining other factors at the average level showed an increase in the estimated production volume from 6552.3 to 6634.4 thousand tons, or by 1.25 %. The results obtained can be used to tentatively estimate changes in production demand for bitumen products as infrastructure demand changes. The practical significance of the study lies in the possibility of applying the obtained dependencies in the preparation of preliminary estimates of the development of production and the formation of approximate forecasts of industry demand. The results allow the impact of infrastructure costs to be taken into account when assessing output prospects and can be useful to refineries, management bodies and planners.