Don Dion, publisher of Fidelity Independent Adviser and related newsletters and manager at Dion Money Management, is generally one of the smart guys in the room. But his recent call on MLPs seems off the mark.

Don’t get me wrong, you could do a lot worse than to follow Dion's advice. But his take on what's going on in the United States Natural Gas ETF (NYSEArca: UNG) seems a bit wrong-headed.

On Monday, Dion posted an article over at TheStreet.com boldly titled "A Natural Gas ETF to Play Contango."

Based on the headline, I assumed this was either:

• A way to make money if natural gas goes up.

• Or, a way to make money if natural gas contango disappears.

Instead, it's an introduction to the JPMorgan Alerian MLP Index ETN (NYSEArca: AMJ). To be fair, it’s a good introduction to AMJ, the only good way to play oil and gas master limited partnerships. I’m actually a fan of AMJ, because it does open up an asset class traditionally hard to get into.

But the disconnect between the premise (an alternative way to get natural gas exposure) and the fund he’s talking about is profound. Dion’s claim that it “offers investors exchange-traded exposure to the natural gas marketplace without the problems of United States Natural Gas (NYSEArca: UNG)” is a bit like saying General Motors gives you exposure to the industrial metals markets—not strictly wrong, but completely irrelevant.

The problem is simple: Oil and gas partnerships are in the transportation business. They get paid for moving oil and gas around. They are operations companies that make money by managing infrastructure well. While there are exceptions, for the most part, they are ambivalent to whether natural gas is expensive or cheap.

In fact, you could make the case that if natural gas gets even more expensive, MLPs will actually suffer because less gas will be consumed.

In reality, however, the historical performance of MLPs is entirely uncorrelated to gas and oil. Wachovia put out an excellent 100-page research report on the topic last year, and here are its findings:

MLP Correlations With Other Asset Classes

Data: Wachovia

To pick out just one data point, the three-year correlation of MLPs with natural gas was 0.13. (Which is to say, nothing.)

Of course, you don’t need to look at fancy-pants correlation math to see this—you can just look at the chart. This is the performance of UNG vs. the Alerian MLP index over the past few years:

In fact, what you see so far in 2009 is the strong negative correlation: The MLP index is up 25% for the year, whereas UNG is down over 60%.

Of course, since the MLPs themselves are traded, there are multiple arguments you can make as to why they went up. But it strikes me as reasonable that it’s predominantly for the right reasons—they’re huge income generators in an environment where yield’s hard to find.

But are they a play on natural gas contango?

Sorry Mr. Dion, there’s just no real evidence to support that theory. Natural gas could double in price and swing into backwardation and most MLPs wouldn’t really care. Unless there’s a massive disruption in actual natural gas supply and demand (and there really hasn’t been during this tremendous rundown), MLPs will keep doing what they do: print money slowly.