5. What History Tells Us: Previous Examples of Rig Lay Downs

Summarizing to this point, the bearish and bullish factors affecting the natural gas market are:

BEARISH
1. incredibly high storage levels
2. consistently bearish weather since February
3. emergence of a new resource (shale gas) with lower costs and high initial flow rates

BULLISH
A. low price (which is below the marginal cost of all but the most cherry new wells)
B. the past few weeks of EIA inventory data
C. the change in the supply and demand disposition over the past 4 months

One might think that the battlefield is level at this point, and the incredible price volatility over the past 2 weeks affirms that market participants are even more uncertain than usual about price. However, previous examples of supply gluts and rig lay downs point to a decidedly bullish outcome.
In Figure 6, we model the last three instances of major rig lay downs.
Figure 6

The first thing to note about Figure 6 is that the percentage of rig lay downs is greater in our current episode than in the previous two episodes (roughly 60% vs. 45%). Therefore, even if drill rates in late 2008 were substantially greater than what was necessary to maintain supply, we have now certainly fallen to a drill rate that will not maintain supply.
In the next two figures the thing to note - and this is crucial - is that in both 1999 and 2002 rig counts continued to fall even after the price had reached its nadir (Figures 9 and 10).
Figure 7


Figure 8


The fact that rig counts did not start rising until 8-12 weeks after the price nadir has serious implications for the current volatility of natural gas prices. If market prices don’t start rising until storage flow is clearly negative (that is quantity supplied is obviously less than quantity demanded), and if rig counts fall even after prices start rising, then the market is set up for a potentially serious shortage. I don’t think it is a coincidence that in both cases (1999 and 2002), prices spiked by more than 500% within 20 months after the rig cont low. Extreme volatility in input prices has been shown to create a drag on the economy.

{(footnote 7: Among others, see Bernanke, 1983; Irreversibility, uncertainty, and cyclical investments. Authors comment: The problem of volatility has no obvious solution and without the futures market, the volatility would likely be greater (and the chances of shortages higher.) As much demonizing as there has been of speculators in the past year, this is clearly a place where speculators can improve Pareto Optimality of the economy. By taking a risk to buy at market bottoms, speculators can smooth what would be an even greater disruption of the natural gas market. It is also true in both cases that rig counts did not start falling until prices had fallen significantly from their peaks, so speculators have a function in both rising and falling markets.)}

To examine the two cases in Figures 7 and 8, it appears that in 1998-1999 producers reduced rig use so long as prices were below $2.50. In the second case (2001-2002), producers reduced rig counts so long as prices were below $3.25. In both cases, these were approximately the same price that producers started to reduce rig counts. The significance of this can be seen by looking at the current situation (Figure 11). Rig counts started to fall in late 2008 after prices had fallen from $13.75 to approximately $8. However, prices have now fallen on top of that by more than 50%! This suggests that prices might have to rise substantially (perhaps greater than 100%) before rig counts start to increase.
Figure 9


Once again, to balance the argument, we must consider shale natural gas. If the shale natural gas resource is large enough, the industry will be able to offset the depletion in the entire resource base by only adding more shale wells. If the marginal cost to drill these wells is significantly below $8 and they can be brought on fast enough, then it is less likely that we would see $8 in the immediate future.
To conclude this section, in previous episodes of natural gas gluts that resulted in laying down a substantial portion of rigs, the rig count did not start to increase until 8-12 weeks after price had reached its ultimate nadir. Since breaking contracts is very expensive, gas producers will operate at a loss (selling gas for less than the cost to produce it) so long as it is less than the substantial loss incurred by capping a well or not using already leased equipment. This retards the rate at which they would otherwise lay up their rigs. On the opposite side, a company will not take a rig out of storage until they are sure the project is profitable. Therefore there is a gap in the price between which companies have neither incentive to lay down nor start up rigs. This is what leads me to predict that supply could continue to decrease for at least another 4 months (2 months until rig counts to start to increase, and another 2 months until supply stops falling.) It is worthwhile to note that in each of the previous episodes, rig counts did not start to rise until prices had achieved the level where rig counts had started to fall. While shale gas could certainly alter the dynamics, this suggests that rig counts may not start increasing until prices reach $8.