Ensemble kalman filter for crude oil price estimation

Abstract

Oil has become the most important sector for the world economic sector. Since 1983, oil has been the main focus of economists after the known impact of oil prices on the economy of the United States after the Second World War. Oil has also been the key role to a world economy although its nature changes over time. The relationship between capital markets and commodities is one of the most challenging problems for investors. Turmoil in one market can affect other market price indexes. Crude Oil Prices are influenced by political conditions and weather-related factors, which can create an unexpected shift in influencing supply and demand. Oil price volatility can be resolved by estimating world crude oil prices so that economists can predict when world oil prices fall or rise and set policies in the purchase and use of crude oil. Estimates are made because a problem can normally be resolved using previous information or data related to the problem. The Kalman filter is a method of estimating the state variables from a discrete linear dynamic system that minimizes estimated covariance errors. The objective of this study is to estimate the price of crude oil using the Kalman Filter (KF) and Ensemble Kalman Filter (EnKF) method. The simulation results show that the EnKF method has a high accuracy of less than 2% and KF method has accuracy of less than 8%.