Natural Gas: Commitment Of Traders Analysis

by: Moshe Ben-Reuven


Despite physical indicators showing net NG deficiency in winter, HH prices turned sharply downward.

The Commitment of Traders (COT) data shows significant differences between WTI crude and Natural Gas.

The COT data for NG, extrapolated, indicates a Flip sometime in January, as Long overwhelms Short - both Commercial and Non-Commercial, simultaneously.

With the Flip, there is no need for a Perfect Storm.

1. Background

In a recent article (NG: Predicting, Especially the Future), we have yet again carelessly predicted a sharp rise in the prices of NG this 2015 winter, based on a physical model driven by projected consumption and diminishing reserves.

We seem to be in the midst of an undeclared war by Saudi Arabia on Iran and Russia, so crude oil prices are in near free fall: from over $100 per barrel, to about $55/BBL, just about 50% lower. We should mention that LNG prices in Asia are derived from crude prices; a rough measure is LNG/MBTU is 13% - 17% of Crude/BBL, for the so-called Japan Crude Cocktail, JCC. So, by inference, NG must also be getting cheaper, or should it. That is, if we were exporting any LNG, which we are not.

The weather seemed warmer than average for this side of December in the US northeast. Then came the NG price tumble on Friday, December 19, followed by Monday, December 22. Essentially the price on the February and March futures contracts of Henry Hub NG (NGG15 and NGH15, respectively) came down from about $3.60 to $3.00, some $6,000 loss per contract. The question in everybody's mind is, what the heck is happening? The simple answer is, nothing out of the ordinary.

Figures 1 and 2 provide a snapshot of Friday the 19th for February 2015 NG and WTI/NYMEX crude. Note that the volume is 62 thousand.

Expecting WTI to drop? Here is a small surprise, in Fig. 2. As NG futures tumble, WTI crude picks up $2.77 per barrel, i.e., $2,770 per contract. We note here that traffic is 5 times larger than NG, with volume at 367 thousand.

A few days later, WTI came back down, and went down further, resting currently at $55/BBL.

In summary, winter seems mild, there are publications of NG storage levels being "highest in 3 years", but no weekly storage data in support. Further publications allude to large surplus in NG production, again no concrete numbers. Actually, the recent peak in EIA reported storage is the lowest peak in 3 years. Shale today provides 40% of all NG in the US. With the recent Q3/2014 moves by shale operators out and away from dry gas and into oil shale, as reported, we should expect less, not more NG production from shale. Finally, there is no mention of consumption, or balance. After all, in all winters, consumption exceeds production. We therefore end up with some psychological notions as driver, which are not in agreement with physical reality.

To gain some insight into this interesting market dynamics, we turn to the Commodities Futures Trading Commission, or US CFTC. Specifically we refer to the Commitment of Traders (COT) report, published each normal Friday afternoon, pertaining to trading in the preceding Tuesday of the week. The last such report available at the time of this writing is for Tuesday, Dec. 16, 2014. Gratefully, dis-Aggregated Historical data file is available for the current full year, for download in Excel format.

To get a sense of perspective, we start with WTI oil futures. This field is energy related, and certain cross correlations to NG are inevitable.

2. WTI NYMEX Futures and Price Moves

As mentioned above, on or about July 8, 2014 there started a slide in the price of WTI. Figure 3 shows (EIA Data, 2014) price moves for the front month contract for 2014.

Two striking events are marked in Fig. 3. The first is the onset of the price slide, July 8. This was about the time the US has declared an open direct diplomatic track for negotiation with Iran over its nuclear armaments program, and the date that negotiation was approved by the Iranian Supreme Leader, Ayatollah Khamenei, seen half smiling in the enclosed event photo; Mr. Hassan Rouhani, President, is not amused. Also, not coincidentally, the time when the Hamas war on Israel became a full-blown reality, officially acknowledged.

The US effort to resolve Iran's nuclear armament drive by diplomacy remains extremely worrisome to the Saudis, until now an ally, moreover, a quasi-state in the Union. The house of Saud relied on the US to provide ultimate security, in return for providing unlimited oil at reasonable price. The Saudis know the Iranian rulers will never give up their nuclear ambitions, unless forced. They also know the Iranians will (yet again) lie to appear negotiating, while actually running the centrifuges to the maximum, to get fissionable mass for a few nuclear bombs. The Saudis know the sanctions are working, and that Iran's economy is on the brink of collapse. However, the US President's current approach toward Syria, and to Iran, remains on a track of peaceful diplomacy. Indeed, to show good will, on July 21, the US granted release of $2.8 billion in frozen Iranian assets, and provided 4 more months of centrifugal freedom, to November 24; by which time, an Agreement must be reached. Thus, the trump card against Iran is unused, and the nuclear clock is ticking. Note that July 21 also marks a mini-tumble in WTI oil prices, Fig. 3; in view of the above, not coincidentally. It is also the first time Hamas agreed to discuss a cease-fire, again not coincidentally.

Of course, Iran does not need a nuclear weapon to obliterate Saudi oilfields and Gulf terminals with its conventional SS missiles. Iran has full capability for that, but it needs to actually fire, or to send suicide bombers, or else effective defenses would be deployed to defeat such attack. However, armed with nuclear weapons, Iran does not need to act anymore. A spoken threat would be sufficient. This is something the Saudis, along with other gulf states, cannot tolerate.

So what is Saudi Arabia to do? The only advantage available to it, is petroleum. Its use as a weapon is simple. Drive to lower crude oil prices drastically, and the weaker economies depending on its export will be crippled or collapse. This is far cleaner than a messy shooting war. Just need to open the spigot on the largest world reservoir to the maximum, and the hell with weaker demand. Oh, and to appear environmentally well-meaning, ostensibly declare your target as defeating the US shale-oil producers, a negligible speck of the market.

They taught us in the Army that the attacker should outnumber the entrenched defender 3:1, and should expect significant losses in attaining its objective. Indeed war is painful. Saudi Arabia (CIA 2014 data) has a population of 23.37 million, and (World Bank data, 2014) GDP of $745.27 Billion. It is exporting (CIA) 6.88 million barrels per day (BPD) of crude oil. If the price of oil dropped by $45/BBL for 12 months, the Saudis would sustain the equivalent of an annual GDP loss of 15.2%. Iran, A nation of 80.8 million people, (CIA) and GDP of $368.9 billion (World Bank data, 2014), should be exporting 2.44 million BPD . The same 12 month period and $45/BBL loss, would mean an equivalent 10.9% GDP loss for Iran. One must remember, however, that Iran's economy is already in bad shape, while Saudi Arabia remains highly bankable, and can possibly sustain several years of such belt tightening. Also, the damage on per-capita basis is amplified at least 3.5 times higher to Iran than to Saudi Arabia.

We finally note that on November 24 it became clear that the Vienna P5+1 nuclear talks with Iran failed, while 4 months were centrifuged away. Instead of tightening sanctions abruptly, the US announced an open door policy on diplomatic solution, an agreement to differ, perhaps gradual easement of sanctions; and maybe, cooperation with Iran on ISIL?. On November 25, Goldman Sachs has downgraded its outlook on crude oil from $90/BBL to $70/BBL, which coincided with (if not directly caused) a significant mini-slide in WTI price, see Fig.3.

Now what has the commitment of traders to do with this scene? As we all know, Futures traders are divided into three categories. Commercial traders are those dealing directly with the physical underlying, like oil producers, refiners, oilfield services providers. The Non-Commercial traders, are hedge funds, investment banks, institutions. The first two types are required to file reports with the CFTC, and tend to be big, multi-million interests. The third kind is very small in comparison, private investors, which are termed non-reporting. Each group will include Long (buyers) as well as Short (sellers). The Non-Commercial group has an additional class of "Spread" trades, which contain coupled, equal-number of Long and Short Future trades. At any given moment in the Futures market, the total number of open Long and open Short positions must match perfectly, for obvious reasons. The CFTC COT report provides the number of trades in each category at the weekly Tuesday reporting day's end. This is the basis for all related data used here.

A snapshot of the last COT report available at this writing is provided in Figs 4.a and 4.b, which show the trader composition of WTI at NYMEX, on December 16, 2014, for Long, and for Short trades, respectively.

What is evident from Fig. 4a is, that non-commercial (NC) trades dominate at 64% of the field in Long positions, where commercial (COM) trades are just 31%. Regarding Short positions, Fig. 4b shows 51% held by commercial trades, and the NC and others make up the other half, combined. We also note the total number of open interests was 1.48 million. Evidently, the commercial trades are overwhelmingly on the Short side. It would be interesting to see the recent-historical behavior of Trades on this WTI Future market. Figure 5 provides a timewise distribution.

In Fig. 5 we see the variation of WTI trades for the 12 months 2014. Until July 8, COM Shorts (orange line on top) were between 900,000 and 1 million, while NC Short + Spread (the red and green lines) were rather flat. After July 8, The number of commercial Shorts declines quite strongly, until Nov. 25. In the meantime, NC Shorts are moderately rising. In other words, while COM entities reduce selling in an attempt to support a higher price, the NC Short trades are betting on the WTI prices to fall. We also note a declining COM and NC buy (Longs, dark blue line and light blue) activity after July 8, ending about Nov. 25; these COM and NC Long distributions are in near-perfect sync after July 8.

On November 25 we see a sharp rise in WTI COM Short sales, perhaps in recognition that the WTI price will keep falling further, so, lock in the current price. NC Long and NC Spread (light blue and light green lines) buying are rising after Nov. 25, perhaps assuming the market is about to bottom, which it has not.

What is the value of COM Shorts relative to the total US crude oil market? If we consider US consumption at 19 million BPD, and the daily number of open COM Short positions on Dec. 16, 2014 is 752,236, at 1,000 BBL/contract. Thus by volume we are looking at 3.6% of the total physical market. Not much.

In summary of this somewhat long introduction above, we conclude that the WTI NYMEX Future market may react to, but not in any imaginable way drive, the price or availability of crude oil in the US. This should hold regardless of whether one accepts, or rejects, the foregoing theory of the implicit Saudi war on Iran. We must emphasize that both July 8, as well as November 25, have serious geopolitical implications for the world's premier oil supplier, and that both dates find distinct expression in the WTI price, and the Futures market, as shown.

That said, we are reminded of the quote by Douglas Hubbard in his book "How to Measure Anything". The 2 biggest mistakes in interpreting a correlation: The biggest mistake, assuming that the correlation proves causality; second biggest mistake, that the correlation is not evidence of causation.

3. What about Natural Gas Futures?

And now, something completely different. In a recent article by this author, quoted above, it was noted that certain physical features in the balance between production, storage, and consumption, can be construed to produce a sharp rise in NG prices in the winter, driven by local demand. How is this vision supported, or refuted, by the commitment of traders ? We take a hard look with the available 2014 data from CFTC.

Figure 6 show total open interest in Futures on Henry Hub Inter-Continental Exchange (ICE), compared with COM Short positions, and also overlaid with the Henry Hub prices for the same period.

Firstly, it is evident that we are looking at a very large volume relative to the total annual NG market. The total number of open interests went over the last 12 months from about 6 million, to about 4.5 million. Each contract involves 10,000 Mega-BTU (MBTU). The corresponding total volume on Dec. 16 was 4.467 million contracts, equivalent to 43.67 TSCF. The annual US consumption of NG (EIA, 2014) is about 26 TSCF. So the proportion of open positions to total market by volume is 168%. Very high, and very different from crude oil futures visited above.

What is the composition of NG trades for these futures? Figures 7a and 7b provide a Dec. 16 snapshot, based on the same COT report.

It is shown that the dominant force are the COM Long positions, 69% of all Longs, and the COM Short positions, 71.6% of all Shorts. Non-Commercial interests are dwarfed in comparison. Is there a question as to which of the two, COM or NC, drives the market? Similar to the calculation above, we see that COM Shorts comprise 3.206 million contracts, equivalent to a volume of 31.34 TSCF, or 120% of the total annual domestic market.

In brief summary, it seems that pessimistic commercial interest expect a significant decline in NG price, with Short overwhelming Long positions-until very recently.

Now we turn to the full COT report for the NG HH in 2014, for historical perspective, showing all 5 distinct trade positions. This is shown in Fig. 8.

Figure 8 shows that significantly on October 28, there was a temporary "flip" point, shared by both commercial (COM) and Non-commercial trades. Long and Short volumes intersect, respectively, and then re-separate. By "intersect" we mean specifically that Long overcomes Short. The effect on the Henry Hub ICE prices is shown in Fig. 6 above: a temporary, sharp rise from $3.54 to $4.22/MBTU. This upward trend peaked on November 18, followed by a rapid decline in the price.

Looking again at the late fall of 2014 in Fig. 8, we see that Long and Short resume lower and higher positions after the brief Oct. 28 "flip". To affect an upward change in price direction, a dominant trading class must go through a flip, where the previously low buys dominate the previously high sells. By far, the COM trades dominate this futures market. We drew 2 polynomial trend lines through the COM Short (2nd order) and the COM Long (3rd order). A persistent intersection is indicated sometime in January.

Of course, the information in the COT report is already late when we see it. However, what is important to observe is the trend, through near-term history. Thus, sometimes in January, as both (forward projected) COM and NC cross-overs indicate, the price of NG will rise. A very large winter storm thrown in, with record low temperatures could help, but is not really necessary. This is very different from WTI crude in Fig. 5.

As a precaution, we should remember the wise words of the disillusioned Macbeth (William Shakespeare, Act 5, Sc. 5):

"Tomorrow and tomorrow and tomorrow/ Creeps in its petty pace from day to day / To the last syllable of recorded time. / And all our yesterdays have lighted fools / The way to dusty death. Out, out brief candle--"

Likely referring to the Futures candle charts. Here is wishing all a very prosperous and happy new year!

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.