Natural Gas Traders Ignore Statistics

A detailed, statistical analysis of natural gas supply suggests that we currently have the largest surplus of natural gas in the past 13 years. Prices, to date, have remained stubbornly high, but factors are accumulating that will drive prices significantly lower.

On Friday, January 19, 2007, natural gas prices rose 10% in one day. The most-cited explanations were a bullish outlook espoused by leading oil service company Schlumberger (NYSE: SLB), a bullish one-week weather prediction by independent forecaster DTN Meteorlogix, and proposed legislation to roll-back exploration incentives. However, the current gas supply situation paints a very different picture.

As of January 12, 2007, the Energy Information Administration of the Department of Energy (see link below) reports that the US has 2,936 billion cubic feet (Bcf) of natural gas in storage, well-above the average storage level for this time of year, or an average of 2,445 Bcf over the past five years. On the surface, this suggests that we have plenty of gas in storage to meet winter demand, but a closer look at the data suggests a much deeper supply/demand imbalance.

Analysis of storage data over the past 13 years reveals a staggering statistic--our current storage levels are 2-3 standard deviations above the average. Over the past 13 years, the average storage level for the second week of January is 2,234 Bcf with a low of 1,525 Bcf in 2001, a high of 2,648 Bcf in 2002, and a standard deviation of 302.5 Bcf. This means our current storage level is 2.3 standard deviations above the average. Limiting the sample to the past 5 years (which is common practice among energy analysts) reveals more alarming statistics: an average of 2,466 Bcf with a low of 2,195 Bcf in 2003, a high of 2,648 Bcf in 2002, and a standard deviation of 156.4 Bcf, which means our current storage level is 3.0 standard deviations above this average. (A short statistics refresher: 3 standard deviations above the mean is a 1:200 outlier. 67% of observations fall within +/- 1 standard deviation, 95% fall within +/- 2 standard deviations, and 99% fall within +/- 3 standard deviations from the average).

Winter weather can not correct this imbalance alone. The mild winter experienced by most of the US is already factored into the current storage numbers. Since a weather prediction was cited as a reason for the surge in prices, lets look at some statistics for the remainder of the winter. According the same data stream from the DOE, winter demand for natural gas ends about the last week of March. In the past 5 years, the US has drawn a maximum of 1,499 Bcf of gas from storage between the second week of January and the last week of March (in 2003), a minimum draw of 880 Bcf in 2006, with an average draw of 1, 227 Bcf. Applying these 'estimates' to our current supply yields a projection of reaching the end of March with 1,709 Bcf in storage if we have an average draw from here, 1, 437 Bcf if we draw the most we've drawn in the past 5 years, and 2,056 Bcf if we draw the least. The average storage level for the last week of March over the past 5 years is 1,228 Bcf. Even if we draw the maximum of the past 5 years, we end March with 17% more gas in storage than average (39% above the average with an average draw, and 67% above the average with a minimum draw). It's going to take a lot more than a couple weeks of cold weather to work off this surplus.

An important secondary analysis of storage levels reinforces the extent of oversupply. While gas storage levels follow seasonal patterns of injection and withdrawal, they are also cumulative in nature--i.e. if we end this withdrawal season with a surplus, we are likely to enter the next withdrawal season with a surplus. Closer examination of DOE data vividly illustrates this point. A mild winter in 2006 resulted in the lowest 2nd half draw of the prior 5 years. We ended the winter of 2006 with a record 1,695 Bcf of natural gas in storage. So from a storage perspective, we began this winter well-supplied. The winter of 2007 has been even milder than that of 2006 and guess what?--we're poised to end this winter with another record amount of gas in storage. This cumulative effect of oversupply takes a LONG time to correct, especially when prices remain stubbornly high.

There is a rational explanation for why storage and supply levels remain elevated--high prices. Despite what you read in mainstream media and hear from energy executives, the data indicates that there is a plentiful supply of natural gas, AND this high level of prices is unsustainable. Every oil and gas producer is a profit maximizer (including all the OPEC countries). When prices are high (or more precisely, higher than their marginal cost of production), they will continue to send the products to the market. This is why OPEC production quotas do not work. An analogy would be to argue that an individual investor should not sell stocks when the market is in free fall (like in 1987) because their selling adds to the downward pressure. That individual knows that everyone else is selling, and despite the popularity of the term, he's not going to 'take one for the team.' OPEC, oil producers, and individual investors are not team players, they are all out to maximize their OWN profit, not the profit of the market as a whole. So even though inventories are swollen and prices have fallen, this imbalance will continue to grow until the price falls to that magical marginal cost of production. The marginal cost of production is less than half the current $7 per thousand cubic feet-Mcf-of natural gas. With fat profit margins like that, you can't blame producers for continuing to pump as much oil and gas as they can, hence the swelling supplies.

The unprecedented storage levels also negate terrorism and hurricane risk premiums. Think of natural gas like any other product. Inventories are used to protect against supply disruptions, by definition. In fact, our normal cycle of injecting gas to storage in the fall and withdrawing during the high-use winter months is the balancing factor for the market and has been since the beginning of the industry. (I'll skip the obvious question as to why prices jumped 10% in one day during the winter because high-use is predicted....) There is a more significant implication and that has to do with risk premiums. Perhaps our supply system is more vulnerable to supply shocks due to Gulf hurricanes or to terrorist acts. I think most would agree (and the weather data confirms) that the 2005 Gulf hurricane season was one of the most devastating in history. Yet if we examine the 2005 natural gas storage and withdrawal data after the hurricane, we BARELY see a statistical deviation. By the first week of November 2005, we had 3,229 Bcf of natural gas in storage versus the 5-year average for that week of 3,112 Bcf. One of the most disastrous hurricane seasons on record for that region did not impair our ability to inject gas for winter use. Even if it were able to stop injections completely, we currently may have a large enough buffer to make it through another season with NO injections. Now for terrorism, to think that terrorists could harm our energy infrastructure more that Katrina and Rita combined is utterly ridiculous. If we can survive a more severe beating that those hurricanes, we can certainly survive (with ease) hundreds of shoe-bombers. The point is that inventories are doing what they are supposed to--mitigate against supply risk, therefore, an additional price premium for those risks is unwarranted.

All commodities since the beginning of time have travelled the same 'coiled-spring' path. Due to extreme competition, commodity markets are notoriously well-balanced under normal conditions, but a sudden shock--2% deficit or a 2% surplus--can send prices skyrocketing/plunging until production and inventories counter the risk. We endured the front end of the supply shock after the hurricanes (needlessly, it turns out, see data above) and prices skyrocketed to $15/Mcf. Now we are experiencing the inevitable plunge because it is clear that we have reserved (inventoried, produced) more than sufficient quantities for all known risks. In a balanced market, a deviation as small as 2% was historically sufficient to invoke a dramatic price response, the current deviation is many times that amount. And we are still ADDING to the excess because of the high prices. This 'coiled spring' will be just a devastating to the price of natural gas as past commodities--gas will keep flooding the market until the marginal price of production is met, then it will severely overshoot on the downside just as it did to the upside at the end of 2005.

The slowing US economy receives honorable mention as a short-term factor. Its clear from the data that the largest physical risk to our natural gas supply/demand balance is none other than Mother Nature. More precisely, it is temperature deviations during the winter/peak period. A couple of degrees warmer than normal sends us immediately into an oversupply situation. The effect of winter weather is greater than that of horrible hurricane seasons, terrorist acts, or even short-term price moves. However, the economic expansion of the largest energy-consumer, the US, deserves mention. The inventory build over the past two years occurred during a time of unprecedented economic expansion (5% annual growth rate at beginning of 2006). Whenever the economy returns to a more moderate expansion near its historic 2% rate, gas supplies must shrink even more to maintain balance. Stated another way: had we not experienced rapid economic expansion in the past two years, the supply/demand imbalance would have corrected much more quickly.

The accumulation of negative factors for natural gas prices is overwhelming. We currently are 2-3 standard deviations above average storage levels, the peak/winter demand period is half-over (and has been the mildest of the past 5 years), and economic expansion is slowing. In this context, its puzzling that some were willing to pay 21% more for natural gas today than they were 3 days earlier.