It’s been a brutal summer for natural gas investors. Much has been written about the enormous discount for natural gas when compared to oil in energy terms. And, remember, natural gas energy comes with smaller carbon and pollution footprints, so if we lived in a truly rational world (a big “if”), gas arguably should be priced at a premium to oil, not at a discount.
Instead, natural gas is dirt cheap and making new multi-year lows as the summer ends. There’s been plenty of hot money piling into long plays on both the commodity and into the UNG ETF but so far it’s been a bad investment. Right now, it almost looks like natural gas is on the way to becoming free. Yet production is hovering near all time highs. Something is wrong with this picture, but it takes looking a little closer at the data to see what’s really happening.
Here’s the U.S. monthly natural gas production data for the last three decades up to June of 2009:
A naïve look at this graph of U.S. natural gas production could be used to argue that we are in a new era for natural gas where increased production is the norm. There is certainly more reason to be optimistic about U.S. domestic gas production than there is for oil. In fact, I sketched out the arguments for this “new gas era” several months ago in an article in the Journal of Energy Security. There has indeed been an increased focus on drilling for natural gas, and recent higher prices have stimulated the development of better techniques, particularly in the increased use of horizontal drilling and fracturing to produce gas from tight formations.
However, as I said in that article, I am deeply skeptical of the “new era” argument, and a closer at the production data bears out this skepticism. Instead, gas production looks likely to fall very soon. In fact, viewed carefully, production is already falling. To see this, the key is to look not at the raw production data, but instead to graph out the year-over-year change in production. This simple change, looking at the year-over-year change in production, does two important things:
1. It is the simplest way to seasonally adjust the data, which is important because there is considerable seasonal pattern in natural gas production (this seasonal pattern had been less pronounced lately, but it is still significant). To the eye much of this pattern looks like noise, and a year-on-year plot lets you “see through the noise” and gives you a much clearer picture of what’s happening in the production data.
2. Graphing the year-over-year data is a simple way to graph the rate of change (the first derivative, in the mathematical sense, rather than the financial sense) and also lets us see the second derivative easily, too, which is just the slope of the year-over-year plot. (Let me here divulge a little secret of quantitative financial analysis: it is this second derivative, the acceleration, that is often most important in predicting future prices.)
Here, then, is the year-over-year natural gas production data up to the month of June 2009:
The data are much tighter and the current trend is now immediately clear. Natural gas production is not continuing unabated while prices sink to new lows. In fact, the seasonally-adjusted rate of change of gas production has been decreasing for about a year and is now right at zero. Even more important, the data give every indication that U.S. natural gas production is likely to begin falling very soon. This is likely because the deceleration (the 2nd derivative) of year-over-year gas production is strong, and notably stronger than recent production decelerations. (By the way, after mentioning the 2nd derivative again, it is appropriate to give a shout out to the math teachers of the world: your calculus teacher was right when she said calculus really did matter in the real world….).
The steep deceleration of gas production is not actually very surprising. Much of the recent spike in production has come from new gas from tight geological formations. There is a lot of this stuff, but it is expensive to produce and—even more importantly in this discussion—the production from wells in tight formations typically falls off much faster than production from conventional wells. Once you’ve produced the gas from the area you’ve laboriously horizontally drilled and artificially-fractured, the lack of natural permeability of the rock turns off the spigot. But recent prices have cratered gas drilling activity, so the upshot is that there are a lot of spigots that are turning off right about now.
So what does this mean for natural gas prices? Production is, of course, only half of the equation; the other half of the equation is consumption. Where consumption goes depends a great deal on the health of the overall economy. But if you are reasonably optimistic about the economy, falling natural gas production is likely to at least put a floor under gas prices. Whatever the economy does, given the story the gas production data are telling, one would certainly expect the ratio of natural gas to oil prices to move back into its more normal range, reflecting more closely their relative energy values. This would imply a large relative increase of gas prices, compared to oil prices.
Things can change, of course, and commodities are especially unpredictable at times; but given the data above, it is becoming very hard to make a case for even cheaper natural gas. In fact I believe that the tide may turn sooner and more aggressively than many market participants expect.
Disclosure: The author currently has long option positions on UNG, as well as long positions in the energy-correlated securities SU and PBW.
