Time Lags, Producer Behavior, Supply Response, and the Potential for a Natural Gas Price Spike

Gas-directed drilling activity in the US has fallen by 54% in just over a year, with the Baker Hughes gas-directed rig count now hovering in the low 400s. At the same time, however, US Lower-48 dry gas production has remained between 64 and 65 billion cubic feet (Bcf) per day throughout 2012 (see Figure 1), with an increase to over 65 Bcf per day anticipated in the fourth quarter.

Figure 1: Gas-directed Drilling Activity and Production (Source: IHS, Baker Hughes and EIA, © 2012, IHS Inc)

Many observers look at this and assert that production should have fallen by now. This line of thinking concludes that the gas-directed rig count is no longer a useful predictive indicator of future gas production. Indeed, given the large volumes of associated gas emerging from the tight oil drilling boom, this metric is more difficult to interpret. But it is far from useless.

Cyclical nature of the gas business The gas industry has long been a cyclical business. This was perhaps forgotten during the period from 2000 through 2008, when natural gas prices were generally much higher than in the 90s, and when dry gas production stagnated at around 50 Bcf per day even as drilling activity increased year after year. With the Shale Gale, however, US Lower-48 dry gas production has grown by around 30% since the beginning of 2007. As a result, the cyclicality of the gas business is about to reassert itself.

One of the main drivers of cyclicality is the time lag between drilling activity and the production response. The fall and rebound in production in 2008/09 offers an illustration (see Figure 2).

Figure 2: Time Lag Between Drilling and Production (Source: IHS, Baker Hughes and EIA, © 2012, IHS Inc)

Here, the lagged effect from peak in drilling activity to peak in production was about five months. As drilling activity began to climb again in 2009, the elapsed time from trough to trough was also about five months.

This time difference is attributable to the activities that must be completed between the drilling of a well and placing the well on production, including activities such as completion and connection. This example works well because the fall in drilling activity was very sharp, falling 50% in about six months and almost 59% by the time the correction had run its course. One factor obscuring this effect during the current correction is that drilling activity, which averaged over 930 rigs in October 2011, was well above replacement level, the level of drilling activity necessary to offset decline. IHS CERA estimates replacement level today to be around 600 rigs, a level not breached until May 2012. Extrapolating from that date, production should have begun to fall in October.

One other factor obscuring the underlying correction is the delayed effect of drilling in the Marcellus and Eagle Ford plays. In both, wells were drilled and completed, but not placed immediately on production owing to a lack of infrastructure (see the IHS Insight Pipeline Projects to Drive Increase in Marcellus Production). With perhaps as much as 2 Bcf per day coming on stream, production will increase at the moment when, other things equal, it ought to fall.

Nevertheless, the correction is well under way. IHS CERA expects US Lower-48 dry gas production will peak at just under 66 Bcf per day in November before falling to a low of 62 Bcf per day in late 2013. This will restore market balance and bring with it higher natural gas prices, reduced coal displacement, and an elimination of the surplus of storage inventories above the five-year average.

Here is where it gets interesting. IHS CERA estimates that the marginal cost of natural gas supply today is around $3.50 to $4.00 per million British thermal units (MMBtu), perhaps closer to $4.00 than to $3.50. This represents the full cycle break even cost of the dry gas production needed to supply the market. IHS CERA estimates that the cash price of natural gas will exceed $4.00 per MMBtu by the summer of 2013. The futures market may well anticipate this before it occurs. This will give producers the signal that it is once again time to pursue dry gas prospects.

Time lags at play It probably will not happen that producers begin to drill for dry gas in earnest at precisely that moment. We must now deal with two other time lags. The first is recognition lag. Our research has developed an insight into energy recognition lags by applying research into bias associated with human nature. Irving Janis (Janis, I. L. (1972). Victims of Groupthink: a Psychological Study of Foreign-Policy Decisions and Fiascoes. Boston: Houghton Mifflin.) developed the idea of "groupthink" in analyzing the failure to anticipate the outcome of the Bay of Pigs invasion even though in retrospect the available evidence was clear that the invasion was doomed to fail. This concept suggests that peer pressure to remain part of a group drives people to emphasize information that supports and reinforces the conventional wisdom of the group and to discount or ignore information that challenges the commonly held vision. As a result, the group is slow to recognize accumulating evidence that the world is not going in the direction that they expected-hence a recognition lag. Max Bazerman and Michael Watkins also describe "Vividness Bias" as overweighting vivid information in developing expectations or making decisions. It tends to distort perceptions of risk. For example, since current information is more vivid we tend to overestimate the probability that current conditions will continue. This leads to unrealistic expectations that the future will look like the recent past.

Once the price signal has been given, producers need to understand it. In mid-2013, producers will have suffered through a prolonged period of low natural gas prices. The conventional wisdom will be that drilling for dry gas is not the best way to direct scarce capital, and the notion that natural gas is a low-priced commodity will be a very vivid piece of information. Indeed, we have already seen a pronounced shift in investment towards natural gas liquids-rich and oil drilling, owing to the better prices those commodities command in the market. In this environment, executives of natural gas producers must a) perceive the price signal, b) recognize that it represents a reversal in the recent trend, and c) go through their budgeting process and perhaps to their boards to obtain capital. This will be very difficult to do without several months of higher prices to support a change in resource deployment.

Once producers make the decision to drill for dry gas in earnest, they will need to obtain drilling rigs. With the shift towards oil, it may prove difficult to obtain those rigs. Indeed, a resumption of dry gas drilling may require the construction of new rigs. This provides a second time lag, one of uncertain duration.

Lastly, only after the recognition and drilling rig lags run their course, once gas-directed drilling picks up, the five months between drilling and production will come into effect. IHS CERA predicts that the average cash price of natural gas at Henry Hub will exceed $5 per MMBtu in 2015, and will climb systematically toward that level in the interim. IHS CERA's analysis suggest that there is the potential of price spikes beginning as early as the end of 2013, and continuing into 2015.

Price volatility still a reality This does not mean that the Shale Gale was a mirage. It does mean that it is difficult to line up investment dollars with the natural gas market with precise timing in a business characterized by time lags and in which the "fog of war" can obscure underlying trends. Indeed, the weather is the largest driver of price volatility in the natural gas market, and a significant weather event (e.g., the very mild month of January 2012) can move the market for months if not longer. Volatility may be muted, but it is not gone. The Shale Gale has lessened the probability of a prolonged period of high prices, and it will likely discipline a price spike quickly if one does occur. But it does not mean that a spike cannot occur. Market participants should expect - and prepare for - this kind of volatility.

Posted 4 December 2012

This article was published by S&P Global Commodity Insights and not by S&P Global Ratings, which is a separately managed division of S&P Global.