Forecasting gas component prices with multivariate structural time series models

Abstract

Predicting gas component prices over different horizons is important for both energy producers and consumers. In this study, we model and predict the joint dynamics of butanes, propane and naphtha traded in the north European market. Our approach is to use multivariate time series with unobservable components. We applied monthly data over a 10-year period, from 1995 to 2006, and tested the predictive power of fitted models using various hold out samples. The in-sample and out-of-sample results indicated that gas component prices follow stochastic processes with trend and autoregressive effects that continuously change over time while the seasonal patterns seem to be stationary. The prediction results were compared with random walk for one-step and multi-step forecasts in each of the out-of sample periods. The results are promising and indicate that our model can be used for short-/medium term forecasting of gas component prices. Copyright 2010 The Authors. Journal compilation 2010 Organization of the Petroleum Exporting Countries.

Shale Gas Boom Affecting the Relationship Between LPG and Oil Prices

“…Less attention has been paid to other important petroleum products, and their relationship with oil and natural gas markets. Westgaard et al (2008) and Myklebust et al (2010) examine the price dynamics of propane, butane and naphtha traded in the north European market. They find that prices contain a random walk component making price predictions challenging.…”

Optimal hedging strategies for salmon producers

“…Kåresen and Husby (2000) use a state space model to create joint forecasts for both spot and futures prices in the Nordic electricity market. Myklebust et al (2010) implement a multivariate state space space model to forecast the prices of different natural gas components. For an introduction to time series analysis using state space models, see e.g.…”

Multivariate Forecasting of Crude Oil Spot Prices using Neural Networks

“…Multivariate models use multiple variables akin to econometric models, but do not always assume a linear relationship or use just regression. Some of the commonly used multivariate methods include multivariate time series models [15], neural networks [16], [17], and multi-criteria decision models [18]. Reference [16] used neural networks to predict commercial oil price using country of origin, sulfur content, and density, and [17] used wavelet neural network to predict monthly crude oil price using just the inventory levels -not truly a multivariate methodology.…”