Bullish on Natural Gas? There's a Smarter Way to Play

| About: The United (UNG)

Every time I browse Seeking Alpha I see the same thing: authors screaming about the new commodities bubble inflating again. But I suppose that this time it's different.

Let’s take a look at forward curves of major commodities. Check wheat, for instance. Last summer, when a huge bubble appeared and front month went through 1300, we saw solid backwardation. Why? Because those conditions were specific; it was wheat that caused prices to move higher. Traders supposed that the following summer, farmers, excited with spot prices, would plant much more grains. That’s why calendar spreads moved so quickly and reached such huge backwardation levels.

Now we see that curve is nearly flat. Even more important, spreads didn’t move at all during the spring, when front month went up more than 30%. The answer is simple. We now see another reason and it is the world’s excessive liquidity, so the forward curves in all commodities are now much less specific for any particular commodity. Even metals, where steepness of curve was always significant, now is flat. But there is still an exception.

A small glance at oil and natural gas. While I completely agree with most authors telling us about the lack of fundamental reasons for such growth on WTI, I’d like to stress the difference between them. Natural gas was for a long time a satellite for oil in terms of forward curve. I saw many times how small movements by anyone in this pair were repeated by the second one with some lag.

But now we see a significant discrepancy: WTI is on the run with a nearly flat curve, new lows every month and a huge contango in NG. Oil flatness is understandable – oil is a global market, so opportunities for quasi-arbitrage appear in steep contango. Everybody witnessed just a few months ago how big companies freighted vessels to store WTI and constantly rolled already sold futures on oil over and over again, gaining up to 150% per annum. The American natural gas market, by its nature, is local and NG is expensive to store, so such possibilities are much more difficult to realize. So at the moment we see growing natural gas inventories that push nearest contracts to ridiculously low levels and at the same time very high estimations in more distant years. The situation is unique, so let’s to try to discuss how it can be used for trading.

I guess that there are two most probable outcomes. In the first one, inventories in NG continue to grow up, causing all curves to slide down, because in one day traders will realize that these inventories will not begin to decrease in the foreseeable future due to fallen demand. That will force them to sell distant contracts more aggressively, narrowing the contango.

In addition, gas producers will start more intense hedging due to a very simple reason – just 25 percent of them are profitable if the gas price is less than 3.5 and selling distant futures contracts on the falling spot market is the only way for them to save the margin.

The second and more improbable outcome, in my opinion, is uplifting of first months. What do we see in every major move up in first contracts? Narrowing contango again. Get the idea?

Calendar spread on NG. I’d like to emphasize that this game is quite correlated to just buying front month but it is much more safe and traders can pocket hefty profits even in case of total oversupply of NG.

And as the old saying tells us, the devil is in the details, so it’s important to understand that we are not going to play on seasonality. This spread should be in the same month but different years. Traders should also avoid rolling such spreads, so my personal choice is Oct09/Oct10. Looking at NG calendar spreads' historical volatility, I suppose that 3 months is more than sufficient for this unique situation to resolve. Right now this spread is at record levels for the decade, but if we take into consideration the low spot price for NG and calculate the ratio between these two contracts, it will be an all-time record – a whopping 1.5.

Chart of spread

I hope that you catch the idea!

Disclosure: Long NGV9, short NGV0.