Swedroe: Don’t Count Out Commodities

March 12, 2018

The asset class with the worst performance over the past 10 years is collateralized commodity futures (CCFs). For example, from 2008 through 2017, PIMCO’s Commodity Real Return Strategy Fund (PCRIX) not only underperformed the Vanguard 500 Index Fund (VFIAX) by 13.7 percentage points, it lost 5.2% a year.

With that performance, many investors have concluded that the premise of investing in commodities was wrong, and that it was wrong to have invested in such a strategy. Before we go into a detailed analysis of the performance of CCFs, we need to discuss a few issues, beginning with the problem of judging decisions based only on outcomes.

Among the most common errors made by investors is to determine the quality of a decision based on the type of outcome. In his book “Fooled by Randomness,” Nassim Nicholas Taleb explained why this is a mistake: “One cannot judge a performance in any given field by the results, but by the costs of the alternative (i.e., if history played out in a different way). Such substitute courses of events are called alternative histories. Clearly the quality of a decision cannot be solely judged based on its outcome, but such a point seems to be voiced only by people who fail (those who succeed attribute their success to the quality of their decision).”

A good example of this involves thinking about life insurance. We buy it to protect against premature death. And we don’t complain about not collecting the premium. We also don’t judge the decision to buy life insurance a poor one because death did not occur. Yet when it comes to investing, we make this variety of mistake all the time. It just appears to be an all-too-common human error, and one influenced by hindsight bias—which is the tendency, after we know an outcome, to believe that it was inevitable.

Logic & Decisions
In her excellent book “Thinking in Bets,” Annie Duke cites the case of Seattle Seahawks coach Pete Carroll and his decision in the 2015 Super Bowl to have quarterback Russell Wilson pass from the Patriots’ 1-yard line, on second down, rather than hand the ball to human wrecking ball Marshawn Lynch, who had just gained four yards on the previous play, with 26 seconds remaining in the game and a chance to take the lead. The play resulted in an interception and headlines like this one: “Pete Carroll Botches the Super Bowl.”

But the facts are as follows. In 2014, rushes from the 1-yard line scored on just 57.5% of the plays, and Seattle had scored on just 33% of such plays that year. Even worse was Lynch’s success rate of just 20%. And over his three-year partnership with Wilson, Lynch had more often failed to reach the end zone (seven times) than he had reached it (five times) when given the opportunity from the 1-yard line.

And we cannot ignore that, in the NFL’s 2014 season, there were two fumbles on runs in that situation. As to passes, the league success rate from the 1-yard line was the same as for the run. However, there were no interceptions, or fumbles, in such situations. And Seattle had scored on half its tries. Without hindsight bias, also known as “Monday morning quarterbacking,” there is no way one could say that Carroll made a poor decision.

Another common error is recency—the tendency to overweight recent events/trends and ignore long-term evidence. This leads investors to buy after periods of strong performance—when valuations are higher and expected returns are now lower—and sell after periods of poor performance—when valuations are lower and expected returns are now higher. This results in the opposite of what a disciplined investor should be doing: rebalancing to maintain the portfolio’s asset allocation.


Returning to our example of the PIMCO fund, PCRIX, while it had underperformed badly in the last 10 years, in the period from its inception in 2003 through 2007, the fund returned 16.9% per year, outperforming VFIAX by 4.1 percentage points a year. If we go back a bit further, using the Bloomberg Commodity Total Return Index to represent CCF returns, over the eight years from 2000 through 2007, the index returned 12.1% per year, outperforming VFIAX’s return of 3.3% per year by 8.8 percentage points per year.

There’s yet a third mistake investors make when evaluating performance of a fund or strategy, and that’s to think of returns in isolation instead of how the addition of an investment impacts the return and risk of the entire portfolio.

Why You Should Consider CCFs
To examine results from the proper perspective, I’ll review the reasons I’ve recommended that investors consider adding a small allocation to CCFs to their portfolios. The reasons for recommending a small allocation were:

  1. One should not expect a higher real return much above the real return on riskless Treasury bills. The reason is that, ex-ante, we should not expect a positive or negative (at least not more negative than the arbitrage would indicate) roll return. Any higher real return should result from higher expected return on the collateral that supports the commodity futures contracts.
  2. The combination of commodities’ high volatility and their low-to-negative correlation with other portfolio assets means a small allocation can have a meaningful impact.

In Appendix H to the 2005 edition of my book “The Only Guide to a Winning Investment Strategy You’ll Ever Need,” I laid out the case for considering an allocation to commodities. It showed the evidence that, on average, commodities had experienced negative correlation to stocks (both U.S. and international) and Treasury bonds. The most negative correlation was with Treasury bonds.

In addition, not only were the returns to commodities positively correlated with inflation (thus providing a hedge against unexpected inflation), they were also correlated with changes in the rate of inflation—rising inflation would help returns and vice versa. This adds to the appeal of commodities as a hedge against rising inflation.

I also specifically warned investors by citing the period 1980 through 1995. I wrote: “Inflation, as measured by the Consumer Price Index (CPI), declined from 13.3% in 1979 to 4.6% in 1989 (and hit a low of just 1.1% in 1986).



  Compound Annual Total Return (%) Annual Standard Deviation (%)
GSCI 10.7 16.8
S&P 500 17.6 12.7
EAFE 22.0 23.4
Treasury Bonds 12.6 15.1

Source: DFA Returns Program


After presenting the previous chart, I continued: “During this period of declining inflation, the Goldman Sachs Commodities Index (GSCI) not only underperformed stocks, but underperformed bonds as well. In addition, during this period the volatility of the asset class was still quite high. However, we should not confuse strategy and outcome. Gaining benefits in periods of unexpected high and/or rising inflation means we should expect the reverse in periods of unexpected low and/or falling inflation. This is what diversification is all about. For the thirty-four-year period 1970-2003, there were only two years (1981 and 2001) when the GSCI-TR and either the S&P 500 or the EAFE Index produced negative returns. It is also worth noting that during the severe bear market of 1973-74, the GSCI-TR returned 75% and 40%, respectively.”

Additional Evidence
Note that I specifically cited two years when commodities and equities had negative returns. Thus, critics who said commodities failed investors in 2008 (as their returns were negative at the same time as stocks, showing there was not negative correlation) failed to understand correlation is only an average figure over time. It doesn’t mean the correlation will always be negative, unless the correlation is a full -1.

I also showed there had never been a year over the same period in which the returns to both long-term Treasury bonds and commodities were negative. This demonstrates the hedging value of commodities not only against shocks to stocks, but against shocks to bonds as well.


Following are the two pertinent tables from my book:


Years of Negative Returns to Long-Term Government Bonds (1970–2003)

Year Return of Long-Term Government Bonds (%) Return of GSCI Index (%)
1973 -1.1 75.0
1977 -0.7 10.4
1978 -1.2 31.6
1979 -1.2 33.8
1980 -4.0 11.1
1987 -2.7 23.8
1994 -7.8 5.3
1996 -0.9 33.9
1999 -9.0 40.9
Average Return -3.2 29.5

Source: DFA Returns Program


Years of Negative Returns to the S&P 500 Index (1970–2003)

Year Return of S&P 500 Index (%) Return of GSCI Index (%)
1973 -14.7 75.0
1974 -26.5 39.5
1977 -7.2 10.4
1981 -4.9 -23.0
1990 -3.2 29.1
2000 -9.1 49.8
2001 -11.9 -31.9
2002 -22.1 32.1
Average Return -12.5 22.6

Source: DFA Returns Program


I then provided a further warning: “As is the case with any risky asset class, investors considering commodities must be very patient and disciplined investors willing to accept very high volatility. The reason is that a risky asset class can provide very poor returns for a very long time. For example, for the 10-year period ending 1999, the real return to the GSCI Total Return Index was just 0.9%.”

Finally, I provided another reason to consider including commodities. I explained: “Equities have significant exposure to event risk. Unexpected events like wars, disruption to oil supplies, political instability, or even a colder than normal winter, can drive up energy prices, acting like a tax on consumption and negatively impacting the economy and stock prices. Similarly, droughts, floods, and crop freezes can all reduce the supply of agricultural products, and strikes and labor unrest can drive up the prices of precious and industrial metals. These various events are also uncorrelated to each other, providing an important diversification benefit if one has invested in a broad commodity index.”

I added that “most shocks to commodities are negatively correlated with financial assets because they tend to suddenly reduce supply rather than increase it. Supply shocks not only lead to inflation (which is negatively correlated with stocks) but also to higher costs of production inputs (putting pressure on profits). Because commodity shocks tend to favor supply disruptions rather than sudden increases in supply, commodities tend to provide positive returns at the same time financial markets are providing negative returns. Thus the more investors are sensitive to event risk the more they should consider holding commodities as a hedge against that type of risk. Therefore, while many investors think of commodities as ‘too risky,’ the more risk averse an investor, at least to event risk, the more they should consider including an allocation to commodities. And the more sensitive they are, the greater should be the allocation.”

In terms of event risk, note I stated that most shocks to commodities are negative. That, of course, means there can be positive supply shocks, such as technology innovations that increase supply (such as, in energy, deep-water drilling and, more recently, fracking).

Summarizing, the reasons for considering a portfolio allocation to commodities had nothing at all to do with an expectation of equitylike, or even close to equity, returns. In fact, at the time, my forward-looking return expectation for the asset class estimated a small premium above the riskless rate, about 1%.*


Quality Review
Having completed a review of the reasons I thought commodities merited consideration, we can now review the quality of a decision to include an allocation to them in the proper light.

First, and most importantly, it’s important to recall the reason to buy CCFs was not to produce higher returns. Instead, it was for two other reasons: a hedge against unexpected inflation and a hedge against a negative supply shock or a positive demand shock (increased demand for commodities).

What actually happened? Instead of unexpected inflation, we had disinflation, with the CPI falling from 4.1% in 2007 to 0.1% in 2008, and with it not exceeding 3% in a single year over the decade.

Over the period 2008 through 2017, the CPI increased at an annual rate of just 1.6%. Recall that commodities are positively correlated with inflation, and moreover with the direction of inflation. Thus, if we have falling inflation, we should expect below-average returns to commodities.

Instead of a negative supply shock, thanks to fracking technology, we had a positive supply shock, which led to a dramatic fall in energy prices. Furthermore, instead of a positive demand shock, we got a negative demand shock (the 2008 financial crisis). If an investor had that perfectly clear crystal ball we would all like to own, given those facts, one would forecast poor returns to commodities.

To review, the purpose of owning commodities was to act as insurance against certain types of risks—not all risks. Judging the poor performance of commodities on the basis that the risks you insured against didn’t appear is the same as criticizing your purchase of earthquake insurance because your house was lost to a flood. If you want insurance against a flood, you have to buy flood insurance. Similarly, if you want insurance against deflation (caused by negative demand shocks), you buy long-term government bonds.

Remember, only fools judge strategies by outcomes without considering what alternative universes might have shown up. CCFs were meant to act as insurance, and the risks they were intended to mitigate didn't show up—between mid-2014 and early 2016, oil prices dropped by more than 75%. But that doesn’t mean the decision to buy CCFs was a bad one any more than the decision to buy stocks on Jan. 1, 2000 was, even though for the next nine years, through 2008, stocks produced a risk premium of -7.6% a year.

More Evidence To Consider
The second factor we need to consider is that, in forecasting forward-looking return expectations for CCFs, the long-term evidence was backwardation (where futures prices are lower than spot prices) and contango (where futures prices are higher than spot prices) were, on average, a wash, producing neither a significant benefit or cost. Thus, that formed the basis for my assumptions about forward-looking return expectations.

There were two possible explanations for why this did not happen. The first is that financialization led to increased demand—estimated holdings of commodity index-related products have increased from $15 billion in 2003 to over $200 billion in 2008. The second is supply issues.

Either way, the result was large contango for most of the period. The large contango acted as a significant, and unexpected, drag on returns. We also had a collapse in prices, so investors didn't need the insurance hedge they purchased via commodities. I would note that, finally, just this last month, we returned to significant backwardation in energy futures, providing a tail wind, instead of a head wind, for returns.

Third, recall the positive correlation of commodities with inflation and the negative correlation of commodities with long-term bonds. That led to my recommendation that, if you are going to invest in CCFs, you also consider extending the duration of your safe bonds, as CCFs hedge unexpected inflation risk and longer-term bonds hedge the risks of deflation that negatively impact commodity returns.


Keeping that thought in mind, also recall that considering the returns of an asset in isolation is a mistake. In addition, recall that the low-tonegative correlation of CCFs to other portfolio assets, along with their high volatility, meant only a small allocation to CCFs was needed to have an impact on the portfolio.

I typically recommended investors consider allocations of 3-5% of the portfolio. And while it’s true CCFs did very poorly, they would have been a relatively limited portion of the portfolio overall. On the other hand, bonds would have been 40% of a typical 60/40 portfolio, and investors would have had higher returns on the longer-term bonds they owned, because owning commodities allowed them to take more term/inflation risk.

Portfolio Comparison
With these thoughts in mind, I’ll examine two portfolios.

Portfolio A is allocated 60% to the Vanguard S&P 500 ETF (VOO) and 40% to the Vanguard Short-Term Treasury Index (VGSH)In Portfolio B, we add a 5% allocation to DCMSX (DFA’s commodity fund, which is lower-cost than PIMCO’s PCRIX, and gives us data since 2011) and also extend the duration of the bond allocation a few years so that the portfolio now looks like this: a 55% allocation to VOO, a 40% allocation to the Vanguard Intermediate-Term Treasury Index (VGIT) and a 5% allocation to DCMSX. Portfolios are rebalanced annually. (Full disclosure: My firm, Buckingham Strategic Wealth, recommends DFA funds in constructing client portfolios.)

From 2011 through 2017, Portfolio A returned 8.7%; Portfolio B returned 8.4%. The insurance an investor bought cost the portfolio 0.3 percentage points in returns.

Now let’s consider a smaller, 3% allocation to DCMSX. Over the same period, Portfolio B returned 8.8% and outperformed Portfolio A, which returned 8.7%.

Ask yourself: If you would have known ahead of time that DCMSX was going to lose 6.7% per year while VOO returned 13.7%, would you have wanted to add DCMSX to your portfolio? Note, on the bond side, that VGIT returned 2.5% while VGSH  returned 0.9%.

Now, let’s go back further in time to include the full history of PIMCO’s fund, PCRIX, using it to replace DCMSX. That allows us to examine returns from 2003 through 2017. Portfolio A is 60% VOO and 40% VGSH. Portfolio B is 55% VOO, 40% VGIT and 5% PCRIX. Portfolio A, without commodities, returned 7.3% and Portfolio B, with commodities, returned a higher 7.6%.

The relative results (showing that the portfolio including commodities outperformed) are very similar if we change the start date of the analysis to either 2004 (6.5% versus 6.9%), or 2005 (6.5% versus 6.8%), or 2006 (6.8% versus 7.0%), or 2007 (6.4% versus 6.8%), or 2008 (6.4% versus 6.6%).

If we move the start date to 2009, Portfolio A without commodities produced the same 9.7% return as Portfolio B did with commodities. And if we move the start date to 2010, the returns are again the same for both portfolios, 8.9%. These examples clearly demonstrate the error of considering assets in isolation.


It certainly is true that commodities have performed poorly over the last 10 years. However, as discussed, judging performance based solely on outcomes is the folly of fools. Smart investors know the only way the quality of a decision can be determined is based on the facts at hand at the time it was made.

The case for including an allocation to commodities was based on the low correlation of that asset class’ returns to both stocks and bonds, in addition to its hedging properties against negative supply shocks to energy and other commodities, as well as against the risk of unexpected inflation. Fortunately, neither of those events occurred. In fact, the opposite did, leading to the poor returns of commodities.

However, that doesn’t mean commodities did not fulfill the role they were asked to play. Their hedging qualities were always there. It’s just that they were not needed. And despite their dramatic underperformance, a small, 5% allocation to CCFs as a replacement for equities had only a slight negative impact. A 3% allocation produced a small positive impact.

The bottom line is that the case for including commodities as a diversifier of risks, hedging against certain types of risk, still holds. If you are concerned about event risks and unexpected inflation, they are worth considering (along with TIPS). However, given their high volatility and propensity to go through long periods of poor performance, followed by short periods of strong outperformance, it’s an asset class best suited for investors with patience and discipline.

Finally, I would add the following caveat: Recent innovations in finance, including the introduction of interval funds, which offer access to investments that don’t permit daily liquidity, have provided investors with options I believe are superior to commodities in diversifying risk. Like commodities, these alternative options all have low-to-no correlation to stock and bond returns, but significantly higher forward-looking return expectations.

Four alternatives investors may consider ahead of commodities are AQR’s long/short Style Premia Alternative Fund (QSPRX) for tax-advantaged accounts, or its long/short Alternative Risk Premia Fund (QRPRX) for taxable accounts (as it is tax-managed and also more tax-efficient due to its use of cash securities instead of the swaps employed by QSPRX), as well as three funds from Stone Ridge: SRRIX (a reinsurance fund), LENDX (a consumer, small business and student loan fund) and AVRPX (a variance risk premium fund, which sells puts and calls on a wide variety of stocks, bonds, commodities and currencies).** (Full disclosure: My firm, Buckingham Strategic Wealth, recommends AQR and Stone Ridge funds in constructing client portfolios.)

*It is important to understand that expected returns are the mean of a very wide potential distribution of possible returns. Thus, they are not a guarantee of future results. Expected returns are forward-looking forecasts and are subject to numerous assumptions, risks and uncertainties, which change over time, and actual results may differ materially from those anticipated in an expected return forecast. Expected return forecasts are hypothetical in nature and should not be interpreted as a demonstration of actual performance results or be interpreted as a target return.
**Provided for informational purposes only and is not intended to serve as specific investment or financial advice. This list of funds does not constitute a recommendation to purchase a single specific security, and it should not be assumed that the securities referenced herein were or will prove to be profitable. Prior to making any investment, an investor should carefully consider the fund’s risks and investment objectives and evaluate all offering materials and other documents associated with the investment.


Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.

Find your next ETF