There is always talk of market manipulation when it comes to oil prices. Indeed, yesterday there was an article in the Financial Times about Saudis offering up extra supply to put a lid on prices (thanks to Abnormal Returns for the pointer).
I am more inclined to believe that market forces control the price of energy. This seems like a difficult position to hold at first glance because the majority of the oil produced in the world is controlled by what is ostensibly a cartel, OPEC. However, the pressure by non-OPEC oil suppliers is increased when prices go up and oil does not have a monopoly on energy prices, despite how it might feel when filling up your car.
So today I have prepared an analysis of oil prices versus the dollar. Because a barrel of oil is denominated in dollars, we expect an inverse relationship to exist. I'll be using brent oil spot prices and the UUP index.
As you can see above, 61% of the variation in oil prices is accounted for solely by changes in the price of USD. I suspect that you could explain much of the rest of the variation if you were able to separate the residuals in times of good expectations versus bad expectations, resulting in two lines, looking something like this:
Without looking at a time series it is hard to determine exactly what we are looking at. Here is a graph of the performances of each over the past five years.
Do we have a problem with our thesis that these two data sets are strongly related? The graphs don't seem to reconcile well here. However, with a little data magic (inverting the UUP returns and multiplying them by 5 to account for the lower volatility) we get a new graph.
Voila, something we can work with. The original thesis that these two are related, but with a residual bias (I think I have made that term up) determined by the state of economic expectations seems to bear out.
It seems like the most straight-forward way to turn this insight into profits is by using the knowledge that Oil tends to outperform during periods of increasing economic exuberance and underperform when the market is increasingly depressed. This isn't the same as a cyclical company that is highly correlated to the rest of the market because if the thesis put forth is correct then the relationship becomes disjointed as the market shifts expectations. If you can predict this shift in the optimism-pessimism spectrum there are "alpha" profits to be made. Be careful to note that you don't need to guess when the market goes from optimist to pessimistic or "risk-on" to "risk-off", a shift from optimistic to very optimistic would work just as well.
For the average US, investor this means you have a fairly straight-forward means of profiting. Because we are dealing with commodities, we do not have to be stock-pickers to execute this trade.
Which instruments should we use? For going long the dollar I recommend the ETF UUP, and for shorting it, UDN. I do not recommend using oil ETFs because of the way they are structured. Instead I prefer using futures.
The edge here is all about the timing. If you can predict a shift in market sentiment before everyone else (otherwise it wouldn't really be a shift), you can find profits as the model disjoints. In the case of a shift from less confidence to more confidence one would go long oil futures and long UDN (that is, short the USD), and in the case of a shift from more confidence to less, short oil futures and long UUP (that is, long the USD).