Commodity ETFs: The Last Frontier

May 01, 2018

[This article appears in our May 2018 issue of ETF Report.]

If there’s one question I get asked by people new to ETFs more often than any other, it’s, are there really any good ideas left for ETFs?

Usually, I give a bit of a pat answer: Sure, there’s always a new headline-grabbing theme. Or the inevitable march of factors coming out of academic financing. Or what about the mess that is fixed-income indexing?

But if I’m thoughtful about it, I think commodities remain the grand challenge of indexing. The big thorny problem—like P vs. NP in math—is, what’s the right way to index commodities?

Consider Equities
What’s remarkable about equity indexing isn’t the rash of new ways to skin cats, it’s how much commonality there is. In the FactSet “fit” methodology, which measures how different funds are from a neutral benchmark, the median ETF out of nearly 800 U.S. equity ETFs has a fit of 90—extremely close to the broad market. Sure, there are indexes that are deliberately different—like the Reverse Cap Weighted U.S. Large Cap ETF (RVRS), which does the opposite of what the S&P 500 does—but for the most part, everyone agrees what the baselines are in the first place.

But that’s just not the case in commodities. The basic premise of most index methodologies can be boiled down to one of two key ways of selecting and weighting securities:

  1. Their “importance” in capturing an asset class
  2. Some belief system on what will go up

The examples are pretty clear: Market cap in equities is a measure of “importance.” So is issuance in bond indexes. Anything that varies from these “importance” metrics is generally doing it for specific reasons, designed to improve the risk/ reward trade-off versus the naive defaults.

No Consensus In Commodities
But in commodities, there’s really no agreement on what “importance” means. Some indexes, like the S&P GSCI, focus on production—how much oil comes out of the ground versus how many soybeans are grown. Some rely on formulas that add a measure of liquidity and limit the over-investment in any one sector, like the Bloomberg Commodities Index. Others just equal-weight.

But there’s not even common agreement on what to include: Some include gold; some don’t. Some include livestock; some don’t. And this is all made more complicated by the decision regarding how to invest. Since you can’t park a tanker at BNY’s custodial offices, the vast majority of commodity funds use futures. That means dealing with contango and backwardation, and selecting what contracts to own, and how to roll them.

So if some enterprising young academic is hankering to make a name for themselves, that’s what I’d focus on. Commodities is one of the only places where I think we not only need new mousetraps, we can probably build better ones.


Find your next ETF

Reset All