Natural Gas

Market Snapshot for Gas

Although natural gas inventory reports continued to be bearish through the beginning of May, the price stopped sliding. This was a basic tenet of my articles in April. The degree of bearishness of the inventory reports surprised me in the beginning of May, since there appeared to be an oversupply of 3 BCF/day for a couple of inventory reports! Since then, however, the inventory numbers have been close to neutral and prices have rallied. This is the type of action that one often sees in a deeply oversold market – where bearish numbers provide little new ammunition, and neutral or slightly bullish numbers provide the impetus for a snap back rally.

The Wild Ride

The past 3 or 4 years have been a wild ride for anyone involved in the natural gas market and have provided more than one instance of apparent irrationality in the markets. Today’s article provides one possible narrative of natural gas prices over the past 40 months, and uses an inventory model as a reference point for the underlying supply/demand disposition through the period [Figure 1].

Figure 1: A comparison of price, inventory levels, and supply/demand (2006-2010.) Supply and demand as shown is based on a model that accounts for the effects of heating and cooling degree days.

2007-2008: Commodity Mania

At the beginning of 2007 natural gas was still in a bit of supply glut that was leftover from the drilling boom after Hurricane Katrina. We can see from Figure 2 that, ex weather, supply was greater than demand for roughly the first half of the year and that this dynamic reversed starting in July. Prices rose gradually from the beginning of the year until June and then started down in early June, at roughly the same time that the inventory dynamics started to shift. Prices bottomed in early September and then rallied until July of 2008. What I find notable from this period is that in 2007, a price of eight dollars (per MBTU) natural gas resulted in a glut of supply and six dollar natural gas resulted in a lack of supply. At the beginning of the following year (2008 Q1) supply and demand seemed fairly evenly balanced at eight dollars. However, this balance did not stop prices from moving higher. In fact, as spring time wore on, prices continued higher even as a balanced market turned to one where supply significantly exceeded demand. This is an example of how markets can “remain irrational for longer than you can remain solvent!” Why prices moved up to .5 while marginal costs of production were dropping is difficult to say: perhaps natural gas was caught up in the general commodity euphoria of early 2008. Also, a cold late winter, and a hot early summer had masked the underlying imbalance to some extent. Regardless, the stage was set for an epic bear market in natural gas prices.

Figure 2: A picture of natural gas supply and demand, ex weather, 2007-2008. The red regions indicate times when supply>demand, and the green regions indicate demand>supply. Prices bottomed in September, 2007 (middle of the graph) and then rallied until July, 2008 (right side of graph.)

2008-2009: The Epic Bear

July 2008 was the beginning of a perfect storm for natural gas prices. At the dawn of July, 2008 natural gas was very expensive, trading close to its all time highs of 2005. Shale natural gas was a market factor, but its degree of importance was of significant debate. The commodity market was hot: in fact compared to crude oil, natural gas was still cheap in BTU terms. All of this was about to change. In hindsight, natural gas was a canary in the coal mine for the credit crisis. The bottom dropped out of the natural gas bid in early July, with the price falling 25% in ten trading days. For those who were actively trading natural gas at this time, you probably recall the amazing one-way direction of the fall. There was no bid during this period, no retracement in price – the price went virtually straight down. Amazingly, as prices were falling, companies continued to increase drilling levels – both the lower marginal production cost and the shift in sentiment sent producers into a frenzy of hedging. While ultimately this proved a lifesaver for the companies doing the hedging it caused the imbalance in supply and demand to become more pronounced. The credit crisis was the final nail in the coffin, as industrial demand hit a thermal in September, 2008. By January, 2009 supply was exceeding demand by more than 4 BCF per day! Although drill rates were in free-fall by that time, it took until July of 2009 for supply and demand to be in balance (ex weather.) Even after that prices continued to fall because of the huge overhang in the stock of natural gas inventory. On top of that, the summer of 2009 was a relatively cool summer, so inventories continued to increase even though the underlying demand was actually outstripping supply.

Figure 3: The Epic Bear, 2008-2009. Note the extreme imbalance during the winter 2008-2009. In spite of a cold winter, inventories continued to grow, and prices fell until September, 2009.

2009-2010: The Aftermath

The last 9 months are best viewed in context of the epic bear market that came before it. Although supply did fall below demand for several months during this period, the effect of that was merely to work off the overhang of supply that existed. Also, the continuing development of shale gas led supply to rebound with vigor. Prices bottomed in September 2009 at .40 – only 17.5% of the July 2008 price! Wow. The snap-back rally was pretty fast and furious, but even the January, 2010 price apex of was less than 50% of the price peak of July 2008. During the past 9 months supply and demand has seesawed from being undersupplied (September through January) to being significantly oversupplied again (March through present.) The latest inventory numbers indicate a market that is balanced.

Figure 4: The Aftermath. Prices recovered into the winter of 2009-2010, but supply was quick to come back. Since February, prices have been falling in the face of a renewed bearish imbalance.

What can Supply and Demand Predict About Future Prices?

There is a common element to all three above periods: the balance of supply and demand shifts 2-4 months before a lasting change in the direction of price takes place. The opposite is also true: it takes even longer for a price shift to be reflected in a change in supply and demand balance. In fact, in all three cases, the imbalance continued to get worse even after prices started moving in the other direction. There are a few reasons we might expect these lags:

1. Price trends can take on a life of their own. Traders and hedgers often speculate that prices will continue to do what they have been doing in spite of fundamental reasons that suggest a change.
2. Usually a developing imbalance in supply and demand will be offset by an opposite imbalance in the stock of inventories. For example, if a bearish imbalance develops in supply and demand this most often occurs in response to (and therefore after) inventories that are below average. If the market is focused on the bullishness of the below average stock of inventories, it may take quite a while before the bearishness of the emerging balance of flows starts to sink in. This makes sense from a market perspective, since the risk when inventories are low is that inventories will shrink more. During such a time if supply exceeds demand for a short time it is actually working to balance the market(1) .
3. Weather and seasons play a crucial role in market timing. A very hot summer or a very cold winter can increase demand by 150 BCF and 400 BCF respectively. A severe hurricane season can also have a large affect on supply. The natural gas market is very keyed in to the weather. And while the market responds to inventory numbers contemporaneously, it tends to price weather in advance. That is why prices often rise in September-November (in advance of winter) and fall in February and March (in advance of spring.) These seasonal effects can prolong price patterns or can act as an impetus to a change in price direction.
4. The final reason is that once the price changes directions, it is often pretty far away from the marginal cost of supply. This is an unfortunate outcome because it results in a series of over-corrections, like a drunk driver swerving from the shoulder into the opposite lane of traffic and back again. This is inefficient and the volatility of natural gas prices is one of the factors that dissuades a more serious fuel shift from crude oil to natural gas.

Looking Forward for Gas

The most important determinants in the market for the next 6 months will be

1. The pace of drilling for shale natural gas: Baker Hughes provides a cool graphic on their investor relation page that color codes rigs by type, and the cluster of gas rigs in Haynesville, Barnett, and other shale regions is unmistakable. https://gis.bakerhughesdirect.com/RigCounts/default2.aspx
   
   The pace of future drilling will be determined to some extent by price, but by a large extent by the need to protect lease investments. As several recent news articles have pointed out, NG companies locked themselves into restrictive lease agreements during the 2008 price boom which dictate, among other things, that the companies must drill on the land within 3-5 years or lose the mineral rights on the land. https://www.marketwatch.com/story/shales-a-curse-and-blessing-for-natura...
2. The rate of decline from non-shale natural gas supply. Bill Powers has pointed out that many basins in the US are suffering from the low price and that these fields have essentially gone into a production freefall as drilling has come to an abrupt halt. https://www.financialsensearchive.com/editorials/powers/2010/0517.html
   
   Additionally, there will be a marginal loss of supply associated with the drilling moratorium in the GOM.
3. The last important determinant will be the strength of the domestic recovery, particularly in industries that are heavy users of natural gas. Although the European debt crisis has thrown this recovery into doubt, my opinion is that demand will continue to have the wind at its back for some time.
4. Finally, weather will be a short-term factor. A hot summer or a well-placed hurricane can send the natural gas market into a tizzy. As both of these are forecasted to be factors this summer, it is likely that the market is currently pricing in a premium for this forecast.

As I will discuss in my next article, weather does not form a basis for long-term shifts in price, so if there is a large move due to weather, another dynamic will need to emerge if the shift in price is to be sustained. For this Thursday’s inventory report, my model predicts a fill of 100 BCF. The reduced use surrounding Memorial Day increases the level of uncertainty for this week’s report.

Resources:

(1) Another clear example of all four of these effects is found in the time period leading up to winter in 2005. The market was keyed on low inventories and the approach of winter, and was caught up to some extent in the excitement of all time high prices in natural gas. It essentially took getting whacked over the head for the market to notice that underlying supply was far in excess of underlying demand.