By Abraham Bailin and Lee Davidson
The broad commodity space has realized significant gains over the past several years. Not surprisingly, a proliferation of interest in hard assets and commodity investments followed by both investors and product providers.
Commodity demand has been driven to a large degree by a number of economic and political developments. Historically, commodities have been found to provide several portfolio level benefits: They have provided diversification away from the correlation of traditional asset classes and have served well as an inflationary hedge. Following the market downturn of late 2008 and early 2009, investors scrambled into hard assets for their diversification benefits. Shortly thereafter, the U.S. Fed adopted an extremely loose monetary policy, which bolstered demand for commodities as inflationary hedges.
More than any other commodity, gold has been turned to in times of panic and uncertainty. The case is no different today. Here we discuss the merits and drawbacks of isolated gold exposure and how one might do better to target a broadly diversified basket of commodities.
Preparing for Inflation
Inflation risk stems from the uncertainty of future inflation rates. Unless an investor is stashing cash under his mattress, his purchasing power is generally protected against expected (current) inflation, because it is priced into the current values of financial assets. That said, terribly few instruments exist to hedge the risk of unexpected inflationary shocks.
Historically, investors have allocated portions of their portfolios to real estate, inflation-linked bonds, gold, and instruments tracking a broad basket of commodities in an effort to preserve purchasing power in the event of an unexpected inflationary shock. Given the empirical evidence available, broad-basket commodity allocations have historically been the most reliable hedge against unexpected inflation.
Unexpected inflation is usually defined in one of two ways: 1. the difference between the risk-free rate (usually the 90-day U.S. Treasury bill) and realized inflation in the same time period; or 2. the difference between realized inflation in one time period and realized inflation in the next period.
For the purposes of this article, we use the latter methodology, because we think that the most recently realized rate of inflation is a purer proxy for inflation expectations in the next time period. We measured inflation with the CPI and, consequently, unexpected inflation will be defined as year over year change in the CPI. Therefore, unexpected inflation is high in absolute terms when the inflation rate varies widely from one year to the next.
A correlation of 1.00 means that the hedge and unexpected inflation move in lock step. A correlation of 0.00 means that the hedging asset's movements are in no way synchronized with unexpected inflation. A negative correlation would indicate that as unexpected inflation increases, the hedge's value would decrease. A good portfolio hedge will exhibit positive correlation with unexpected inflation. That is to say that when unexpected inflation increases, a good hedge will increase to a similar degree.
As evidenced by the data in the table above, the broad-basket commodity indexes (S&P GSCI and DJ-UBS Commodity Index) have historically demonstrated higher positive correlation to unexpected inflation shocks than gold. The spot price of gold also had positive correlation to unexpected inflation, but it was much lower.
Correlation does not, however, imply causation. Beyond examining mere correlation metrics, investors should be concerned with whether or not causal inferences can be made about the relationship between unexpected inflation and the hedging instrument. If a causal inference can be made, then the claim that the instrument can hedge inflationary shocks will be further substantiated. That is to say that the relationship is not simply a "curve-fit" coincidence based on backward-looking data. It may have predictive power.
Proof by Way of Regression
We can attempt to determine a causal relationship by studying the returns of each hedging instrument against current inflation rates and unexpected inflation. The goal of this exercise is to determine whether unexpected inflation causes the returns for the hedging instrument to change, and whether this relationship is significant and reliable.
The table below illustrates the relationship between inflation rates, unexpected inflation, and the returns of each of the traditional inflation-hedging vehicles. The coefficients indicate how a change in either the inflation rate or unexpected inflation would affect the returns for the corresponding hedge (either broad commodities or gold).
For example, looking at the S&P GSCI Index, a 1.00% increase in unexpected inflation rates would result in a 6.55% increase over the index's prior period returns. The strength of this relationship is determined by the t-statistic, and in general, a t-statistic greater than 2 implies a strong relationship.
For both the S&P GSCI and gold, the regression yields economically and statistically significant results. Unexpected inflation has historically driven a portion of each index's return. While gold has a statistically significant relationship with unexpected inflation, its relationship is not nearly as significant (nor is the coefficient as high) as it is for the S&P GSCI. Furthermore, repeating this exercise for other broad-basket commodity indices yields the same result when sufficient data is available.
Because the DJ-UBS Commodity Index--in its current form--has a data track record only going back to 1992, a statistically significant relationship can't be confirmed due to a t-statistic below 2.
Is the CPI too Slow?
Based on the above data, a causal inference can be made about unexpected inflation's relationship to both the S&P GSCI and gold. But be careful. These results should not be interpreted to mean that unexpected inflation is the sole driver of commodity performance, but rather, that unexpected inflation is one of several factors that contribute positively to the performance of commodity markets. Based on research conducted by Erb and Harvey (2006), changes in the rate of U.S. inflation accounted for roughly 43% of the variance of returns for the S&P GSCI from 1969 to 2003.
An astute observer may note that the CPI is posted once a month. That means that it is a lagged indicator of inflation. Perhaps, goes the argument, gold performs more admirably in the wake of short-term inflationary increases. That concern is certainly valid.
Much of the volatility that has befallen our market in recent times has been due to pressures that come in the much shorter term--quantitative easing, European bank stress tests, and speculation about the formation of the Eurobund come to mind. These events certainly affect U.S. dollar inflation, but the CPI is unable to comment in the short-term. In order to attempt to address this concern, we wanted to measure how gold and broad-basket commodities would react over a much shorter time frame.
While commodities have proved to react with greater sensitivity to unexpected inflation, perhaps gold will redeem its intangible qualities by showing greater sensitivity to the short-term changes in relative dollar strength.
Testing Gold's Mettle
To qualify (or disqualify) gold's anti-inflationary benefits over the short run, we need to find some measure that is immediately responsive to expected inflation and currency pressures. Much of the interest in gold has been driven by fears of dollar debasement, and so it makes sense to look to the dollar index. The U.S. dollar index measures the strength of the dollar relative to a basket of other globally impactful currencies. Given that the index fluctuates daily, it should serve us well in determining the responsiveness of gold to any inflationary pressures on the currency over the short term.
For our test, we apply the same regression mechanics that we used to test the sensitivity of commodities to unexpected inflation. In this case however, we run the same regression over progressively longer and longer time frames: daily, weekly, quarterly, and annually.
As it turns out, gold is, in fact, a better hedge against short-term fluctuations in the value of our currency than a broad basket of commodities. Identify the Dollar index coefficients. Note that over the daily and the weekly periods, a unitary increase in the value of the dollar index has a larger negative impact on gold than it does the DJ-UBS Commodity Index. As we look to the longer, quarterly and annual, time frames, however, the broad basket provides a better hedge.
These numbers fall in line with our understanding of both the CPI and gold. The CPI measures inflationary impacts across the broad economy. It makes sense that a basket of commodities maintaining a similar level of diversity would consistently respond with greater sensitivity than gold by itself.
On the other hand, much of the volatility that has befallen our market in recent times has been due to currency-centric pressures that arise over a much shorter term. Some of these, like QE, directly impact dollar strength. Others, like uncertainty about the future viability of the euro, have more of a derivative effect, but can nonetheless be priced into the dollar index, in the short run.
As a monetary relative whose tangible uses are few and far between, it is reasonable to find that gold responds with a higher degree of sensitivity over the short term.
Your Choice Depends on the Length of Your Investment Outlook
We have found that over the short run, gold may be a more effective way to hedge dollar debasing pressures than a basket of broad commodities and that the opposite is true over the long run.
At Morningstar, we are strong proponents of long-term investing. For the investor who maintains a broadly diversified portfolio over the long term, the broad-basket commodity exposure should serve just fine to offset any real inflationary pressures. We would not advise using gold alone as a substitute for that broad commodity allocation.
If, however, you are looking to hedge some of the shorter-term dollar swings or event specific risks (like the impact of a Fed announcement for instance), devoting a small portion of your portfolio to a physical gold allocation may make sense.
The Tool Box
There are a number of good options for both broad commodity and gold exposure. For the commodity sleeve of our portfolio, we would look to either PowerShares DB Commodity (DBC) or United States Commodity Index (USCI).
Both offerings look to maximize implied roll yield and thus minimize the contango that has plagued earlier futures-based exchange-traded products. To do this, both funds allow themselves the ability to quantitatively select futures of varying expiration dates. DBC is actually the largest and most liquid broad commodity ETP on the market. It charges an annual fee of 0.85%.
USCI distinguishes itself by using a very novel and academically grounded strategy. The fund maintains a broad basket of commodities, while selecting commodities with the lowest inventories and the highest momentum, on a monthly basis. Either of these dynamic strategies would serve well as a long-term core holding. We prefer USCI's strategy, but its 0.95% management fee is a bit high.
For gold exposure, we look to physically backed offerings. The alternatives are to use futures or equity-based products. Given that the dollar hedging benefits we found came to the physical commodity, we would look to the vehicle that provides the best tracking of spot. Physically backed offerings like SPDR Gold Shares (GLD) and iShares Gold Trust (IAU) hold stores of the actual commodity, so tracking is razor sharp.
On the whole, we would recommend using IAU. Both have great tracking, and while GLD is larger and more liquid, IAU's size and liquidity are extremely deep. We choose IAU on the basis of its management fee, which is a mere 25 basis points. GLD charges 0.40% annually.
Disclosure: Morningstar licenses its indexes to certain ETF and ETN providers, including Barclays Global Investors (BGI), First Trust, and ELEMENTS, for use in exchange-traded funds and notes. These ETFs and ETNs are not sponsored, issued, or sold by Morningstar. Morningstar does not make any representation regarding the advisability of investing in ETFs or ETNs that are based on Morningstar indexes.