The Mechanics of a Natural Gas Price Recovery

A few weeks ago, we took a look at crude oil fundamentals in order to determine where in the commodity cycle we are today. Because of exponential rig efficiency gains, the traditional method of using rig count to predict production, and then prices, has broken down. Not only that, but because of a mismatch in grades of crude, storage inventories are also not as good an indicator of the supply situation as they have been in the past. We traced the data on production capacity and inventory drawdown, to conclude that we are likely on the brink of supply and demand rebalancing, and therefore  we have begun a sustainable recovery in crude oil prices (at least WTI prices).

But where do we stand in the natural gas markets? Well, first we can be clear that the factors making crude oil cycle predictions difficult also apply to natural gas, in fact even more so. Again because of rig efficiencies and high-grading, the relationship between price and production has broken down entirely. As prices have fallen, production has grown exponentially.

To revisit our methodology, we are applying the concept of a commodity cycle as follows:

  1. Drop in production capacity
  2. Flattening of contango on the futures curve
  3. Drop in production
  4. Drawdown of inventories
  5. Price recovery

Drop in production capacity. Since rig count used to be an indicator for production capacity, let’s look at rig efficiency gains in natural gas.

Source: EIA Drilling Productivity Report

Source: EIA Drilling Productivity Report

Once again just like in crude oil, we see that efficiency has yet to level off. The chart is somewhat misleading, since some of the newest production is attributable to completions on wells previously drilled. The EIA suggests that efficiency growth may be slowing soon, so we will be watching for production per rig to level off before we can expect a drop in actual production.

The difference here between the crude oil and gas markets is the geographic distribution of the production growth. While crude production has grown in every basin we studied, the story is not the same for gas:

Source EIA Drilling Productivity Report. Last 3 months are projections

Source EIA Drilling Productivity Report. Last 3 months are projections

The Marcellus shale has contributed substantially all of the production growth in the US. Back in 2012, the Haynesville actually produced more than the Marcellus, about 10 Bcf/d at its peak in the summer of 2012 while the Marcellus was still around 8 Bcf/d.

So in order to understand what’s happening in the gas market and when “production capacity” will level off, we have to pay the most attention to production capacity in the Marcellus. This means rig count, as well as DUC inventory. Unfortunately, data on DUC inventory is hard to come by as it’s not explicitly kept by any state except North Dakota. Investor calls from Antero, Eclipse, Cabot, and Rice suggest that on the whole, the inventory of behind-casing production is beginning to fall. Antero specifically sees these completions as a capital-efficient way to bring on more production and show growth in a low-price environment. Since the drilling expense is sunk and often tied to rig contracts, the only incremental cost is the expense of the completion itself.

Verdict: Because of continued gains in rig efficiency and uncompleted well inventory, production capacity growth is only now beginning to slow.

Flattening of contango on the futures curve. As we discussed in the crude oil article, an oversupplied situation is usually accompanied by a futures curve in which the farther-out months are more expensive than the near months, incenting operators to store excess commodity. As the oversupply gets larger, the curve gets steeper and storage fills. As the oversupply diminishes, the curve flattens and storage operators are no longer incented to add to inventory, which then begins to draw down. How does that look for natural gas?

Source: CME Group

Source: CME Group

Verdict: Not only is the curve not flattening, but the contango is getting steeper, further incenting operators to add to inventory.

Drop in production. The most recent EIA production data is for February 2016, which showed total US natural gas production near an all-time high at 75.3 bcf/d and up 1.4% from January. The agency predicts that production will begin to drop in May, and Bentek’s estimates ahead of this official data do bear this out, with daily supply estimates in May at 72 Bcf/d and lower.

Verdict: It looks like we may soon see a drop in production significant enough to impact the overall market balance. However, unlike crude, the domestic gas market is already heavily oversupplied and lacks the easy relief valve of displacing imports. Not enough gas export facilities and willing markets are online at this time to soak up the excess supply. Essentially, a production drop is not only needed, but physically unavoidable as the summer goes on and we fill all available US storage capacity.

Drawdown in inventories. Speaking of storage capacity, this summer’s injections are on pace to fill to the 4 Tcf high-water mark. Since the futures curve continues to incent injections, it is likely that every molecule possible will continue to flow into storage throughout the summer. Not only are we not drawing down inventories, it’s likely we will test the limits of US storage fields.

Price Recovery. The overall picture for a sustainable natural gas price recovery in the coming months is not great. However, there are a few very important numbers to watch as harbingers of the recovery through this winter and into next year. First, Marcellus production. As I mentioned previously, the story of US gas production is ultimately the story of the Marcellus shale. When Marcellus production begins to truly fall, it’s likely that spot prices will rise, flattening the futures curve, leading to drawdown of inventory and rebalancing of the market.

Second, I’ve yet to mention gas demand in this article. I’ve focused more on supply since it’s the valve most easily turned by month-to-month changes in drilling and completion activity. But the demand picture for natural gas is extremely strong in the coming years. Northeast utilities are busy converting commercial and residential customers to natural gas, and a significant amount of industrial demand in the form of fertilizer and chemical plants is projected to come online in the next couple of years. Not only that, but retirement of coal-fired power generation and fuel-switching also contribute to increased gas demand.

According to EIA estimates, all of these factors add up to total projected gas demand of 76.5 Bcf/d in 2016, and 77.4 Bcf/d in 2017. Recall that current production (as of February) is just over 75 Bcf/d, with more up to date estimates from Bentek at 72 Bcf/d and lower. Idiosyncrasies of weather and seasonal demand changes aside, this adds up to a significant natural gas shortage by 2017, once storage stocks are depleted back to normal levels. In order to meet this increased demand, prices must rebound to a level high enough for producers to cover full-cycle gas drilling and production costs.