Purveyors of commodity investing vehicles have long touted the diversification benefits of adding commodity exposure to a portfolio. Indeed, the relatively low correlation of commodities to both stocks and bonds does provide diversification. However, to be truly beneficial to a portfolio, the “diversifiers” should also add to portfolio returns over the long run. For commodities, the long run is often longer than most investors can stomach.

Before the advent of commodity ETFs, commodity investing was outside the reach of the average retail investor. Getting exposure to commodities required opening a futures trading account or joining a private commodity pool. Although those avenues were available, most retail investors chose to forgo those particular options. Instead, stocks (and funds) of oil companies, mining firms, and agricultural companies provided an easier method to gain some exposure to commodities, even though that exposure was indirect and failed to provide the full benefits diversification.

The world changed when the first broad-commodity ETF arrived in February 2006. According to its fact sheet, the PowerShares DB Commodity Index Tracking Fund (DBC) “is designed for investors who want a cost-effective and convenient way to invest in commodity futures. The underlying index is a rules-based index composed of futures contracts on 14 of the most heavily traded and important physical commodities in the world.” Today, it is the largest broad-commodity ETF, with about $1.9 billion in assets.

PowerShares DB Commodity Index Tracking Fund

Although investors could for the first time get exposure to commodities futures as easily as buying a stock, many problems with commodity investing remained. These included the issuing of Schedule K-1s instead of Form 1099s, the inability to track spot prices, fixed-index allocations, and performance.

Since funds like DBC have portfolios of commodity futures contracts, they are not “40 Act” funds regulated by the SEC. Instead, the Commodity Futures Trading Commission has authority, referring to them as commodity pools, and their year-end tax statements arrive as Schedule K-1s instead of the more common Form 1099s. This created unwanted tax headaches for many investors and prompted the invention of new vehicles to help alleviate those headaches.

The Barclays iPath Bloomberg Commodity Total Return ETN (DJP) sought to overcome the K-1 problem by packing commodity exposure as an exchange-traded note (“ETN”) instead of a fund. As notes (debt security), ETNs were bonds that did not own any commodity futures, but the bond’s value was linked to a commodity futures index. However, this attempt at fixing the problem created additional problems. As bonds, ETNs carry the risk of issuer default, and the cost of hedging their positions has caused many issuing banks to stop issuing additional ETN shares. Issuer default is devastating for ETNs, as demonstrated by the collapse of Lehman Brothers and its Opta brand of ETNs. Additionally, when issuers stop providing additional shares, the primary mechanism to control trading prices disappears, rendering them broken products. ETNs are not the solution.

The First Trust Global Tactical Commodity Strategy ETF (FTGC) was the first broad commodity ETF to overcome the K-1 problem. Launched in October 2013, the fund pioneered an investment approach whereby its only direct investment was in a wholly-owned offshore subsidiary, allowing it to be a 40 Act fund that issued 1099s. The offshore subsidiary was then responsible for handling all of the purchases, sales, and rolls of the needed futures contracts. Most newer commodity ETFs have adopted this approach.

Another problem with commodity funds is that investing in futures contracts does not provide the same performance as something tracking spot prices would. That is because all futures contracts have an expiration date, and to maintain exposure to the commodity, the fund must sell the expiring contract and replace it with another contract with a later expiration date. Most of the time, the value received from the sale is less than the amount needed to buy the new one. This pricing structure, known as contango, causes the fund to lose money every time it rolls its contracts and steadily lose ground against spot prices.

Precious-metals ETFs have overcome this problem by physically buying the underlying commodity instead of using futures contracts. This solution works well for gold, platinum, and even silver, but quickly becomes impractical for other commodities. Whereas $10 million of gold bullion does not require much physical space and is easy to transport, $10 million worth of corn, soybeans, or crude oil is a different story. The cost of storing and transporting large quantities of these commodities is higher than the cost associated with rolling futures contracts at a loss. Therefore, the problem of futures contract performance not equaling spot-price performance remains unsolved for broad commodity ETFs.

The introduction in August 2010 of the United States Commodity Index Fund (USCI) marked the beginning of another phase of commodity ETF evolution by becoming the first smart-beta commodity ETF. Instead of blindly buying every commodity futures contract and rolling those contracts on a regular basis, the underlying SummerHaven Dynamic Commodity Total Return Index selects only 14 of the 27 eligible commodities, and the selection criteria include contracts displaying backwardation (the opposite of contango) and one-year price momentum. It’s not a perfect approach, but it is a giant step in the right direction. Since then, many new commodity funds employ a smart-beta approach.

The biggest problem with broad commodity ETFs is visible in the performance graph above—they have not made any money over the last three-, five-, and 10-year periods. Since its introduction in February 2006, DBC has declined nearly 35%, for a -3.7% annual return. I don’t care how low its correlation to stocks and bonds might be, something that loses that much money is not a good portfolio diversifier—it is an expensive total return reducer. DBC and other commodity ETFs had a nice run up in 2007 and 2008, prompting many to extoll the virtues of commodity investing, but it has declined more than 12% annualized (68% cumulative) since its 2008 peak.

Commodity ETFs have been a welcome addition to the ETF toolbox. However, investors need to be aware that not all tools should be used all of the time. They are great trading vehicles, and their low correlation to stocks and bonds can help provide diversification when they are trending in the right direction. However, broad commodity ETFs are not buy-and-hold investments. See and compare more than 140 commodity ETFs in the 2017 ETF Field Guide.

Disclosure: Author has no positions in any of the securities mentioned and no positions in any of the companies or ETF sponsors mentioned. No income, revenue, or other compensation (either directly or indirectly) is received from, or on behalf of, any of the companies or ETF sponsors mentioned.