Wednesday, July 30, 2008

Commodity Speculation Lessons

In my many years of commodity speculation, the most important lesson which I have learned is humility. The second most important lesson is to look beyond the obvious. And the third most important lesson is that price action leads the fundamental news. In this case, I suggest everyone looks to all of Nat Gas, Coal, RBOB and Heating Oil crack spreads, and the shape of the yield curve, for far more interesting market information than whether a particular price point triggers some short term or long term stops.

Nat Gas has given back its entire rally since January without so much as a minor bounce. This would equate to Crude moving to 90$/bbl without any meaningful bounce. Coal has also taken out the price equivalent of Anatoly's stop level on the downside.
The gasoline (RBOB) crack spread has been negative or near zero for some weeks now. This is unprecendented for the summer driving season and shows the extent of true end user demand destruction. Refineries will obviously reduce their runs rather than process crude at a negative crack spread. Crude is a completely useless commodity. It's the products that really matter!
The heating oil crack spread, in contrast, remains extraordinarily wide. Some attribute this to Chinese hoarding post snowstorm/earthquake and pre-Olympics. So the gasoline and heating oil crack spreads are giving us contradictory signals. While I'm open-minded on the outcome (although short right now), if I see the heating oil crack start declining, I'll know for sure that it's game, set and match for this phase of the crude bull market, and I'll aggressively press my shorts.
The Crude yield curve recently moved to contango from backwardation. Producers are now motivated to build even more inventories, and this is a self-reinforcing feedback loop. When the crude curve goes into contango, it is predictive of declining prices in the spot contract for the following three to six months.
The price volatility of crude is around 40% now. So, as a pure probability statement, one can easily envision a 40% decline from the "high" and we'd still be in a "secular" bull market! More importantly, given 40% volatility, one would expect daily swings of around $7 per barrel — just as random noise.
I am unaware of any rigorously backtested studies which show that the COT open interest is predictive in crude. They are only a coincident indicator of trend. Perhaps Anatoly has different studies available that he can share?
Lastly, and most importantly, many commodities do trend for good economic reasons. I suspect even Vic and Laurel will concede this point. One can ridicule the trendfollowers but I reckon the good ones have been long crude from $90 and started exiting when we broke through $135.
Right now the entire energy complex appears to be breaking down after a remarkable multi-month and multi year-bull market. If you are buying the dip for a quickie bounce — good luck! But if you hold your positions for weeks or months, as I do — it's frivolous to declare your entire market view based on whether one particular price prints. Markets are far wiser than that.

No comments: