a strategy helps.
Gas markets in the United States are complicated, dynamic, and evolving. They offer significant commercial opportunities for some companies, commercial hazards for others.
Many companies find it difficult to estimate the price they will receive for gas the next year, month, week, or day. Even with two futures markets, companies still find it difficult or expensive to hedge price risk completely.
Today, understanding price behavior and knowing how to respond has commercial value. Moreover, in an evolving energy marketplace accessible to an increasing number of companies, market shares are sliced by competitive strategies supported by almost constant involvement in market activity.
Cash vs. Futures Prices
The most important relationship for a futures contract market probably lies in the expected near equality of futures and spot prices for the same delivery period at the close of trading of the futures market. If these two prices are not equal, trade between markets will soon bring them back into alignment. Traders, for example, will buy the commodity on the spot market and then sell through a futures contract when the futures price exceeds the spot price by more than the transaction cost required to complete the trade. However, gains from such risk-free trades accrue only to companies that constantly track the market.
Markets for natural gas are different from most other commodity markets. Spot markets for guaranteed deliveries during a delivery month stay open after the futures market closes. This feature can create hedging problems for a company that uses the cash and futures market; it will close out its futures position prior to completing the cash deal.
A trader obtains a perfect hedge only if the futures price equals the cash price when the cash contract is completed. When prices are different, the trader may incur an additional cost from having used the futures market. For example, if the futures price lies below the cash price, a buyer that has taken a position in the futures market will incur a loss equal to the difference in prices. In essence, the trader buys gas at a higher price and sells gas at a lower price. When the amount of gas hedged reaches to 100 million cubic feet and the difference in prices equal $0.10 per thousand cubic feet, the loss grows to $10,000.
Timing Between Markets
Each month offers at least two distinct trading periods for obtaining gas deliveries. These distinguishable trading periods require separate analyses, since participants will likely have somewhat different needs. These different needs may lead to different relationships between prices.
Bid week. Guaranteed deliveries for an entire delivery month are completed during bid week. This period lasts only a few days after the futures market closes, and never continues into the delivery month. The average price associated with this period is designated the "average bid-week price."
On average, a close relationship exists between the futures price and the average bid-week price1 published in Inside FERC.2 In fact, the average bid-week price can be expressed as the futures price plus a small random amount. This random amount remains unpredictable (em sometimes negative and sometimes positive (em and carries an expected value of zero. Thus, if a company closes out its futures contract position near the termination of the futures contract market and regularly completes its cash transaction during bid week, then, on average, it obtains a perfect hedge. In other words, over time, positive and negative differences between cash and futures prices balance out.
Nevertheless, irregular or infrequent users of the futures market still run significant risk, since the futures settlement price at the close of trading of a futures market and the bid-week price at the Henry Hub can differ significantly. In the last several years, for instance, the difference has run as high as $0.10 MMBtu.
After Bid Week. Marketing and other companies with active trading units regularly engage in both bid-week market and market after bid week to take advantage of commercial opportunities afforded by engaging selectively in both markets. The average of the deals completed during and after bid week is designated as the "average overall spot-market price." This price and the futures price don't converge, on average, but do share a systematic relationship.
An average relationship can be estimated by subjecting price data from the New York Mercantile Exchange and Natural Gas Week to regression analysis.3 The average overall spot market price can be expressed as $0.26 MMBtu +
(0.84 x futures price/MMBtu) +
a random amount. Based on data for June 1990 through September 1995, a high futures price indicates a low spot price: When the futures price is $2.30/MMBtu, the expected spot price is $2.19/MMBtu ($2.19 = $0.26/MMBtu + 0.84 x $2.30/MMBtu). Thus, when the price is high, a producer should sell as much gas as possible during bid week, rather than after.
Nevertheless, dangers remain hidden in such average relationships. Factors that influence the market can change systematically in unexpected ways. For example, from October 1995 through January 1996, temperatures lingered consistently below normal. Weekly storage levels consistently fell farther and farther from levels for the same week in the previous year. The perception of dangerously low storage levels and the fact of consistently lower-than-normal temperatures induced a trend-like behavior, with higher prices in the market after bid week. For example, the average overall spot-market price for December was $2.45/MMBtu; the futures settlement price at closing day came in at $2.24/MMBtu.
Despite such periodic anomalies, the expected average relationship between prices may manifest itself again. The average systematic relationship may reflect 1) regular differences in the quality of the information available during the two trading periods, 2) different needs satisfied by the market
during and after bid week, or 3) a different mix of industry participation with different trading capabilities in both markets.
Traders can make use of other, similar average relationships to gauge the viability of using an indexed contract to represent a market price, or of using the futures market to hedge price risk at locations far from the Henry Hub. Such average relationships should be used gingerly in establishing contractual relationships, however, since they may fracture when market influences change.
In general, a strong systematic linear relationship exists between bid-week prices in Oklahoma4 and the Henry Hub, making it relatively easy to hedge risk for deliveries to a pipeline in Oklahoma using the futures contract price and an adjustment for the particulars of the linear relationship between the two price series. Not too surprisingly, the expected average relationship broke down during December 1995 and January 1996, when the actual price received in Oklahoma was much less than the expected price: $3.13/MMBtu, compared to an actual price of $2.00/MMBtu. Such experiences may motivate some companies to build reopeners into their contracts for periods when short-term price volatility increases significantly.
Beyond the Henry Hub
The price difference between locations for many commodities is equal to a relatively constant amount (sometimes called "basis") that largely represents the cost of transportation between locations. For gas, the situation is more complicated. At many locations, buyers can obtain gas from either producing or storage sites.5 Moreover, a difference in price between two hubs (em one in a primary production area and the other near a primary market area (em can represent not only the cost of transportation, but also differences in supply and demand conditions for natural gas and for pipe and storage capacity between locations.6 In fact, it is difficult to separate out any transportation cost embedded in the cost of the commodity at these locations.
Table 1 gives some idea of how the cost difference between locations changes over time. The related Figure 1 reports the difference in gas price between an area near the location where a delivery through a futures contract is made, and other locations in the United States.7 These locations lie in South Louisiana near Henry Hub, where deliveries are made through the New York Mercantile Exchange futures contract, and in Texas near Waha Hub, where deliveries occur through the Kansas City Board of Trade futures contract. Although the middle value (median) of the difference in cost is often small, the range (the difference between the maximum and minimum values) is consistently large and the difference takes on positive and negative values in several instances.
Interestingly, the average difference and the range appear smaller for the Texas location than for the Henry Hub for four of the other five outlying locations. In fact, statistical analysis suggests that for four of the five outer locations, the Texas price would be more effective for indexing a cash contract.
The highly variable difference in prices is not surprising since demand and supply conditions for the commodity and for pipeline capacity changed markedly during the 1990s (em particularly the cost of transportation, as represented by rental fees for space on pipeline systems for both short- and long-term contracts. This fact is important because the price of gas at other locations includes not only a commodity cost, but may also include some embedded transportation costs. Many shippers were able to obtain deep discounts, at times, on the cost of pipeline space, because excess pipeline capacity was available to many markets. The amount of excess capacity also varied over time, as demand shifted in response to changes in the weather. Then, in 1993, the industry switched to a new rate-setting method (straight fixed-variable) that calculated rental fees for first rights to space (firm transportation service) according to the fixed capital costs of pipeline systems. After 1993, companies gained the right to reduce the cost of transportation through a capacity-release market, where firms release and sublet rights to unused pipeline space.
Implications
for Trading Centers
The development of a second futures market for delivery at the Waha Hub in West Texas and plans for additional futures markets in locations with access to Western and Canadian markets are motivated, in part, by the large price volatility shown between locations. This variability also supports the development of market hubs as trading centers.
Past analysis has identified several distinguishable markets for natural gas in the United States, not one.8 In fact, the formation of market hubs as trading centers was inspired, in part, by the desire to reduce informational and operational inefficiencies. Hubs organize trading activity at locations where prices on gas, storage, pipeline, and other services are available to all participants in the hub market, and where daily trading might be active enough to provide liquid markets. In fact, as trading at hubs substitutes for transporting or trading gas between locations, the amount of gas that remains subject to locational risk actually declines.
Companies with rights to hub services are well-positioned to trade rights to gas and to transportation and storage capacity. As their current and expected future requirements vary from planned amounts, hubs could provide these companies with a means of getting back into balance or of transferring an unneeded service to another hub participant for a fee. This type of activity encourages the development of forward markets for gas.
Moreover, futures markets could face competition were many hubs to develop into liquid forward markets (i.e., markets where bid prices are close to offer prices). Participants could fix prices through a liquid forward contract rather than attempting to fix price through a combination of a liquid futures contract(s) and an expected spot contract(s) for some future time period in which, as previously pointed out, the futures contract and the spot contract trade at different times for the same delivery period.
Understanding price behavior provides strategic value to distribution companies and other large users of natural gas. Active involvement at a hub can serve as a core activity for a company developing strategic approaches to the use of price and other market information.
Finally, active hubs with broad industry participation can provide the same needs for coordination and information provided by alliances among local distribution companies. The difference is that coordination and information are supported by active, open, and relatively transparent contracting in markets, rather than by relatively fragile agreements between the members of an alliance. t
John Herbert is a senior economist at the Energy Information Administration, and an adjunct professor of statistics at the Virginia Tech Telestar Graduate Center. The views expressed are those of the author and do not necessarily represent those of EIA.
E-mail address: jhh1@msn.com
1 This analysis uses the settlement price of the futures contract on the final day of trading. An alternative would be to use the average of the settlement price on the last several days of trading of the futures contract. Both choices lead to similar results.
2 The first-of-the-month price indices, published by Inside FERC, an affiliate of McGraw Hill, largely represent prices during bid week and are popular for indexing contracts in the gas industry.
3 This estimate is a convenient example. A variety of relationships could be examined, and a variety of variables used for the futures price indices.
4 Inside FERC first-of-the-month price of spot gas, delivered to pipelines for Natural Gas Pipeline Co. of America in Oklahoma. (The "Mid-continent" location in Table 1.)
5 More active use of salt storage facilities also influences the cost of gas and pipeline transportation. Active use of these facilities exploit short-term price variability by making the decision to inject and withdraw gas from storage dependent on changes in the level of price throughout the year.
6 In primary market areas like Ohio, local supplies are available from nearby production and storage facilities. Thus, it is difficult to treat a certain area exclusively as a production area and another as a consuming area, as is all too frequently done.
7 Use bid-week spot prices between the Waha and Henry Hubs and other locations in computing price differences between locations for several reasons. At such locations, the bid-week price is approximately equal to a futures market or a forward market price, if a futures or forward market exists at the location. Statistical analysis shows that the futures settlement price at the close of trading of a futures contract provides a good estimate for bid-week spot prices at the Henry Hub, particularly over the last several years. Thus, the bid-week cash price at an active, liquid market is used as an estimate of a futures market price.
8 See, Emile J. Brinkmann and Ramon Rabinovitch, "Regional Limitations of the Hedging Effectiveness of Natural Gas Futures," The Energy Journal, Vol. 16, 3, 1995 113-124; and J.H. Herbert and Erik Kreil, "U.S. Natural Gas Markets-How Efficient are They?" Energy Policy, January 1996. See also, National Energy Board, Natural Gas Market Assessment, "Price Convergence in North American Natural Gas Markets," December 1995.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.