Open Access Article SciPap-943
Research of Objective Market Price Factors in the Formation of Prices on the Oil Market
by Yulia Y. Finogenova 1,*, Denis V. Domaschenko 2 and Edward E. Nikulin 3

1 “Risk management, insurance and securities markets” Chair, Plekhanov Russian University of Economics, Stremianny per. 36, Moscow 115093, Russian Federation

2 Research laboratory “Study of the monetary system and analysis of the financial markets”, Plekhanov Russian University of Economics, Stremianny per. 36, Moscow 115093, Russian Federation

3 Research laboratory “Study of the monetary system and analysis of the financial markets”, Plekhanov Russian University of Economics, Stremianny per. 36, Moscow 115093, Russian Federation

* Authors to whom correspondence should be addressed.

Abstract: “Brent” oil prices (BOP) serves as a global standard for commodity market and it strongly influences the world economy. Forecasting BOP presents a significant and at the same time an arduous task. The main question related to “Brent” prices forecasting is the correct determination of the cause-effect relations. In order to conduct the causality analysis, we have employed adaptive-neuro fuzzy interface system based on the if-then rules and a great potential for the determination of causeeffect relations. The modeling has shown unobvious results. Despite the fundamental law, which claims that the balance of supply and demand forms the oil price, we have proved that the fundamental dependencies are not valid for “Brent” oil pricing. We have revealed that precious metals prices (Palladium, Gold, Silver and Platinum) and commodity currencies exchange rates (USD/NOK, USD/AUD, USD/CAD and USD/ZAR) serve as a signal or cause for the “Brent” price changes. Additionally, we have examined the efficiency of the forecasting model in terms of forecasting oil price trends, achieving maximum 62% of accuracy on the daily data.

Keywords: Fuzzy Logic, Ann, Anfis, “Brent” Oil Prices, Causality Analysis, Forecasting

JEL classification:   C12 - Hypothesis Testing: General,   C45 - Neural Networks and Related Topics,   C53 - Forecasting and Prediction Methods • Simulation Methods

SciPap 2018, 26(2), 943

Received: 1 May 2017 / Accepted: 27 June 2018 / Published: 23 August 2018