Estimate of USO (Oil) and UNG (Nat Gas) Long-term Returns
The problem in analyzing these products is that we don’t have much historical data to consider (04/2007 for UNG and 04/2006 for USO), and the data we do have only covers mostly contangoed markets. How to fix that?
Below is my working estimate of UNG and USO, all the way back to 1994 and 1985 respectively. This is based on individual futures contracts, similar in spirit to what I did for VXX here, but using each ETPs specific roll method.
ETP estimates are in red and spot prices in grey.
How accurate are the estimates?
The graphs below show the period of time in which the estimate (red) and actual ETP returns (blue) overlap.
For a number of reasons, it’s impossible to perfectly simulate UNG/USO returns, but based on the close fit, I’m confident that we’re well within the ballpark.
I’m sure this has all been done before (mainly because it wasn’t terribly difficult), but I haven’t seen it. I had intended to release this data for the betterment of the quant community, but it would be nice to sanity check my results. If you’ve seen a similar analysis done elsewhere, please share.
I’ll share some follow up thoughts on this data in the future, but the biggest red flag is this: don’t assume that the degree of contango (and resulting underperformance) in UNG is par for the course. As I’ve talked about vis-a-vis VXX/XIV, longer estimates paint a very different picture.
. . . . .
Filed under: Random Stuff | 1 Comment