Oil and Natural Gas: Are the Prices Linked?

May 2008

For energy observers, there is an ongoing debate about the degree to which natural gas and oil prices are linked. We thought it might be interesting to look at this relationship and discuss some of the factors that lead to this debate.

The chart below, showing correlation over a 16-month period, shows instances when the two commodities were positively correlated (rise and fall together) as well as times during which they were negatively correlated (move in different directions). Market action this year has shown a positive relationship between the price of crude oil and that of natural gas. Overall this is rather consistent with historical data which shows a mean positive correlation between the two. However, a correlation coefficient of less than 0.40 means that changes in oil price account for less than 40% of the changes in the price of natural gas. The conclusion is that oil prices influence gas prices, but there are other important drivers, too.

An argument for a causal relationship between price changes of the two commodities is that natural gas and oil are substitute products for consumers who can easily switch between the two fuels for their business activity (e.g. dual-fired power plants). For example, if the price of oil were to rise because of unexpectedly decreased supply, then the consumer could switch to natural gas. As a result, the price for natural gas would be bid upwards because of the sudden increase in demand. Conversely, the same idea would hold if it had been the price of natural gas that had risen because of an unexpected decrease in supply.

Nevertheless, it is unclear whether this type of switching is a dominating factor for a relationship between the two prices. According to the US Department of Energy, roughly 66% of natural gas fired power generation plants in the USA that are capable of fuel-switching were built before 2000. However between 2000 and 2006, only 26.9% of natural gas-fired plants built in that period are able to switch to crude oil.

Proponents of a price relationship between the commodities have also considered mean-reversion as a way to get a sense of future trends in price. "Mean-reversion" is the tendency for commodity prices eventually to come back toward their long term mean. Since 2001, the ratio of daily near month crude oil prices to daily near month natural gas prices has varied considerably with the mean ratio being near 8. The chart shows that when the ratio has risen to 12 or 14, it has eventually fallen back toward 8. The rise in oil prices has again pushed the ratio above 12. Mean-reversion suggest this won't last. At current gas prices, oil prices would have to fall to about $100 a barrel to restore a price relationship closer to the historical mean. If oil remains near $130, then gas would have to rise to more than $16 to restore the historical relationship. However, a few studies have questioned the continued existence of the traditional mean-reversion relationship or the stability of the ratios over time.

We have also observed that the relationship between the two is biased towards crude oil as a change in its prices is more likely to affect natural gas prices than vice-versa. Crude oil markets are more established worldwide while those of natural gas tend to be continental. Significant global events (e.g. invasions, terrorism) can affect oil prices in the USA, which in turn can affect natural gas prices by virtue of the aforementioned argument for fuel-switching. On the other hand, local events that affect continental gas supplies (e.g. hurricanes) would have a relatively small impact on global oil prices.

To date, foreign natural gas markets have not had close connection to North American gas markets. However, the growing market in liquefied natural gas (LNG) will move natural gas towards being a more global commodity, and may increase its tendency to move with oil prices.

Thus on one side of the debate, we have the argument that the two fuels are substitute products for each other. On the other side of the debate, we have the argument that one product is traded globally and thus affected by global events while the other is regionally-focused and more influenced by regional supply and demand.

Another explanation for the relatively high correlation in North America between crude oil and natural gas prices this spring is that the supplies for both are tight but for different reasons. On one hand, the economies of China and India are competing ever more for crude oil, while on the other hand, the US drew down much of the surplus in its natural gas inventories because of a harsh winter. In addition, LNG imports to the USA this spring have been lower than last year's thus contributing to the tight supply. 

Fuel switching would tend to support close correlation in gas and oil prices in the short term. Other factors may tend to support the observation that gas prices tend to follow oil prices over the longer term, but with a lag. For example, high oil prices tend to motivate oil and gas exploration companies to shift exploration and drilling dollars away from gas prospects and towards oil prospects. A slowdown in gas-directed drilling will slow the rate of growth of gas supply, and over time, lead to a tighter gas market and higher gas prices.

These are only a few considerations in the debate over the relationship between oil and gas prices. There are several unique, fundamental issues that influence the prices of both oil and natural gas. The existence of these influences complicates the price correlation between the two commodities. While there has been a tendency for a positive correlation, other factors have occasionally exerted greater influence than the others. This has resulted at times in an apparent disconnection (i.e. negative or low correlation) between changes in crude oil prices and those of natural gas.