The markets are moving fast, and I have a lot to account for, starting with my suggestion that investors buy USL instead of USO.

As Jim so nicely pointed out in his blog on Wednesday, my suggestion back in February that investors buy USL and not USO was wrong. Anyone who read my blog, did their own research and then followed through on that trade got killed. For that I apologize. Please know that I do not wear it lightly.

The flaw in my analysis was that I did not think contango would collapse as quickly as it has. When I wrote that blog, the near-month oil futures contract was trading at a 13% discount to the next-month contract. Today, that discount has narrowed to just 1%.

That collapse—one of the most violent in history—benefitted USO, which tracks the near-month futures contract. And it hurt USL, which buys longer-dated futures. The math can be complicated, but basically what happened was that the near-month contract that USO tracks had been trading at a discount ... and that discount disappeared. 

I knew this was a possibility. A few people even suggested it in comments to my blog. I was betting that any collapse in contango would take place over months, not weeks. Contango was a systemic drag on USO, so every month it persisted would harm USO's performance. If the collapse in contango had been spread out over the space of a quarter, investors in USL would have outperformed, because the drag would have outweighed the collapse of contango. But contango collapsed over the course of a few weeks, so it overwhelmed the contango drag.

The question for investors is, what happens next? The answer is, I don't know. Will the oil futures curve continue to reverse? Will oil switch from contango to full-on backwardation? Will it stay the same? Will contango come back? I'm not sure. It is a complicated market.

Toward this end, a reader asked on Jim's last blog why USO doesn't seem to track the price of the oil. The writer—zonacorp—noted that oil was trading at $51/barrel and USO was only trading for $29.61. 

The answer, zonacorp, is exactly the effect I mentioned earlier: contango. There are three factors driving returns in USO: the change in the spot price, contango/backwardation and the interest income that USO receives. USO has under-performed spot oil because oil has been in contango for most of the time that USO has been trading.

A lot of people, like zonacorp, ask me what the best way to invest in oil is. What they mean is, how can I track the spot price of crude oil? The answer is that you can't: Aside from buying oil and storing it in your backyard, the best you can do is futures or futures-based ETFs. But people don't want to deal with all the contango, backwardation and other complications.

The suggestion I would make is to think about what bet you really want to make. I get the feeling that most of the people buying oil are making a bet that it will rise over the next year or two. One way to capitalize on this is to purchase long-dated futures contracts. You can see all the NYMEX oil futures contracts here.

You can buy a December 2010 contract for about $65/barrel. If you think oil will be above $65/barrel by 2010, this is one way to profit from that. You can go out further, too: December 2016 oil futures are trading for $79.93.

Note that liquidity in these longer-dated futures contracts is very, very thin, so trade carefully.

If you want to stay out the futures market, an alternative would be the MacroShares Up Oil Trust (NYSE: UOY). Although it's difficult to say how UOY will trade over the next few years, the security matures in December 2013. When it does, investors will be paid the full value of the February 2014 oil contract. So if you're willing to hold until then, you can count on that. (One caveat: The security will hit a ceiling and be liquidated at full value if oil tops $200/barrel.)

The oil market is complicated.