The concept of adding commodities exposure as a way of reducing portfolio volatility may seem counterintuitive, but the main drivers of commodities prices often vary from those of other asset classes, particularly bonds and securities. This makes commodities important as both diversifiers and volatility reducers in a well-designed, risk-adjusted portfolio. Advisors and investors can turn to the commodities sector as a potentially effective means of achieving higher risk-adjusted returns in a portfolio.
Benchmark Design And The Futures Curve: Contango And Backwardation
When first developed and introduced in the 1990s, multicommodity indexes/funds became popular (and remain so today). This is due mainly to their ability to give nonfutures investors direct exposure to commodities through indirect futures holdings via an index or fund.
These funds are sometimes referred to as "first generation" funds due to their concentration of holdings in the very front of the futures curve. This is because their focus is generally more on achieving direct exposure to the commodities and less on how efficiently the benchmark might perform over time versus a given investor's time horizon and expectations.
These funds can be affected by issues unique to futures pricing called contango4 and backwardation,5 which can have unpredictable effects on investor returns over time. Contango and backwardation are generally more of a concern in the energy, agricultural and industrial metals sectors than they are in precious metals. The latter category has minimal storage and carrying charge costs associated with it compared with other commodities, where storage, processing, spoilage and other costs affecting contango and backwardation can be substantial.
As the issues of contango and backwardation have come to the fore of the collective thinking of the investor community in first-generation funds, a number of "next generation" funds and ETP products have been introduced. These ETPs have benchmark holdings or trading methodologies specifically designed to mitigate (as much as possible) the potential effects of contango and backwardation. However, bear in mind the fact that contango and backwardation can never be fully eliminated. Both are a fundamental element of commodities trading and are an important factor when choosing a commodities-based ETP design that is right for one's needs.
Generally speaking, investors and asset allocators who want to add beta to their portfolio using commodities will be buying and holding their selected commodities exposure long term. This means that a next-generation commodities ETP, designed to mitigate contango and backwardation, may be a more appropriate choice than a (possibly more widely recognized) first-generation ETP. Conversely, an investor looking at shorter-term exposure to a commodities sector (or sectors) may be better served by a first-generation ETP. Your investment's time horizon is particularly important when implementing beta optimization (alpha) strategies through ETPs.
Any combination of ETP choices can be effective, but in all cases, the benchmark design and holdings of the ETP are what will drive returns more than any other factor.
Choosing Commodities Exposure For Beta Diversification
How might an investor go about choosing exactly what type of commodities exposure to include in the beta portion of a portfolio? A good place to begin might be with some basic statistical analysis. Generally speaking, correlation, when used in conjunction with portfolio analysis, determines if the prices of assets included in the analysis move in the same direction, whereas regression6 analysis examines if the movements of one asset can be explained by the movements in another asset.
Many professional investors are well aware that price movements in precious metals are generally not well correlated with the performance of stocks and bonds. However, it often comes as a surprise to investors that many commodities in the agricultural sector have price movements that are less correlated with stocks and bonds than do other well-known commodities. Figure 2 illustrates the 20-year price correlation of 13 major commodities to the S&P 500.7
Precious metals and energies can be effective beta-enhancing components in a diversified portfolio, and the popularity and presence of these two commodities sectors in many portfolios is widely known. But analysis shows that over time, the price movements of energies seem to be more closely correlated with stocks and bonds than are the price movements of agricultures and precious metals. Regression analysis, which relates to correlation by showing how much the movement of something (in this case, specific commodities) is directly affected by the movements in a comparable benchmark (in this case, the S&P 500) can often explain correlation patterns. In fact, longer-term regression analysis of 13 major commodities consistently shows gold, sugar, soybeans and corn as the least directly affected by the performance of the S&P 500 Index.
Specifically, when one performs a regression analysis of the S&P 500 Index over five-, 10- and 20-year periods8 against the 13 commodities available in the single-ETP format, those that are the least tied to the S&P 500 are consistently precious metals and agricultural commodities. (See Figure 3.) In fact, sugar is historically less tied than gold over the last 10 years, and soybeans are less tied than gold over the past five years.
As previously discussed, investors seeking exposure to commodities have easy access to a variety of commodities-based ETP products. They can choose among several multicommodity ETPs or they can customize their exposure using single-commodity funds. Again, be certain to study the underlying benchmark of your ETP choices, because benchmark design more than any other factor will be the true source of your expected performance.
Optimizing Commodities-Based Beta Through Seasonality
Seasonal patterns in many commodities—especially within the agricultural sector—are often (but not always) used by investors to optimize the beta component of their portfolio. The basic patterns and cycles of life on planet Earth, primarily those based upon growing seasons and usage patterns, can be employed effectively to optimize the timing of when investors increase or decrease beta exposure to certain commodities. These patterns are literally cosmic in nature; the positioning of planet Earth at times of solstice and equinox is ultimately what affects growing seasons and usage patterns. These cannot be affected by investors, but they can be used effectively by investors seeking beta-enhancing strategies.
Absolute-Price Patterns And Variance From Annual Average Prices
There are two important considerations when trading commodities for beta optimization (alpha), especially those commodities most affected by fundamental seasonal patterns. Most commonly utilized by investors (especially those exclusively seeking alpha) is the concept that a commodity or market can create relatively predictable seasonal patterns of absolute-low and absolute-high prices. Less commonly utilized, but of immense importance to asset allocators and/or investors seeking long-term beta exposure, is the concept of variance from annual average prices.
Conversely, sugar often (but not always) reaches both its absolute-price high and its widest positive divergence from annual average prices around February or September. These might be opportunities for those with an overweighted sugar beta component to reduce some of their holdings; the same timing might allow an alpha trader to exit long positions initiated earlier.
These macro-patterns are also often found in the corn and soybean markets. Like sugar, corn and soybeans tend to bottom or negatively diverge from the 12-month average price at harvest time, as illustrated Figures 5 and 6. For corn and soybeans, the most significant global harvest occurs between October and December, coincident with peak harvest time in the Northern Hemisphere. Seasonal patterns establishing absolute-price tops and positive variance from average annual prices tend to be more difficult to identify in the corn and soybean markets, but they generally occur between March and July.
Wheat markets have significant seasonality also, with dual peaks in late summer and midwinter, and dual lows in late spring and late autumn. Wheat is harvested multiple times a year: Winter wheat is harvested in late spring, while other varieties are harvested in summer and into autumn, illustrated in Figure 7.
Seasonality in the core agricultural commodities of corn, soybeans, wheat and sugar may be more supply driven than demand driven. This could be due to the fact that agricultural commodities have a relatively inelastic demand pattern, driven by a wide variety of uses including food, animal feed, fuel and various industrial products. But other important commodities, like oil and copper, have different seasonal tendencies that may be influenced more by usage patterns than by supply patterns. For instance, construction activity is often more intense in warmer months; hence copper's seasonal pattern of often establishing price/variance from average lows during the Northern Hemisphere's winter months and price/variance from average highs in the summer months, as shown in Figure 8.
Motor fuel usage—and therefore refinery utilization—also tends to be higher during the Northern Hemisphere's summer season, perhaps influencing crude oil's seasonal tendencies of midwinter lows and midsummer highs, as illustrated in Figure 9.
Matching Investment Objectives To ETP Design
The wide variety of benchmark designs within the commodities ETP sector can be a benefit to investors, but only if each investor chooses the proper benchmark according to his or her own investment needs. Investors need to understand the benchmark design of the ETP considered, including the holdings and the benchmark objective.
For instance, if all an investor wants is to have immediate, short-term exposure to a given commodity or sector in order to trade alpha, then a benchmark that concentrates its positions in the front of the futures curve may be best suited to that particular investment objective. Such a design will likely capture the short-term price movements of the commodity, allowing the capture of alpha in the investment.
If one wants to have long-term exposure to a particular sector or commodity to achieve beta, then a benchmark design that mitigates contango and backwardation concerns within the futures curve might be a better selection. Efficient long-term exposure to the commodity through mitigation of potentially negative impacts of contango and backwardation is a more important consideration for a beta trader than that of capturing near-term price movements.
The objective and term of an investor's investment holdings are of critical importance when selecting the appropriate commodities-based ETP. Financial advisors as well as the sponsors of ETPs themselves are good sources of information when researching which investment selections might be most appropriate.
Commodities are being used very effectively as beta diversifiers in many portfolios. As addressed in the Morningstar study cited above, the inclusion of commodities in a portfolio can reduce volatility over time without negatively impacting overall returns. Investors now have access through a wide variety of ETP products to a range of principal commodities. These ETPs can differ significantly in benchmark design, which is the main driver of returns to the investor. Of the major commodities represented by single-commodity ETPs, long-term time horizon snapshots of correlation and regression studies show agricultural (specifically corn, soybeans and sugar) and precious metals (specifically gold and silver) as the two commodities sectors having the lowest relationship with the S&P 500. Additionally, investors with basic knowledge of fundamental seasonal patterns within the commodities sector can often optimize their beta exposure and/or capture alpha by using an appropriately chosen commodities-based exchange-traded product.
1 Beta is calculated using regression analysis, and you can think of beta as the tendency of an investment's returns to respond to swings in the market. Source: http://www.investopedia.com/terms/b/beta.asp#axzz1tWSehkhK.
2 Alpha is a measure of performance on a risk-adjusted basis. Alpha takes the volatility (price risk) of an investment and compares its risk-adjusted performance to a benchmark index. The excess return of the investment relative to the return of the benchmark index is the investment's alpha. Source: http://www.investopedia.com/terms/a/alpha.asp#axzz1tWSehkhK
3 Morningstar, "Benefits of Including Commodities in a Portfolio – Lower risk and higher return 1991-2010." Originally published on 3/1/2011.
4 Contango: A condition in which distant delivery prices for futures exceed spot prices, often due to the costs of storing and insuring the underlying commodity; opposite of backwardation.
5 Backwardation: A market condition in which a futures price is lower in the distant delivery months than in the near delivery months.
6 Regression: A statistical measure that attempts to determine the strength of the relationship between one dependent variable (usually denoted by Y) and a series of other changing variables (known as independent variables). Source: http://www.investopedia.com/terms/r/regression.asp#ixzz1tXVbXPIy
7 For this purpose, the correlation analysis for each specific commodity is spot continuation (generic futures contracts) as defined by and sourced on Bloomberg: "Generic contracts, such as US1, US2, US3, ..., are constructed by pasting together "rolling" contracts, according to the pre-selected roll types on the commodity default page. The generic contract uses the value of a particular contract month until it "rolls" to the next month in the series. You can access a generic contract by replacing the month/year code with the number 1, i.e. A 1
8 For this purpose, the regression for each specific commodity is spot continuation (generic futures contracts) as defined by and sourced on Bloomberg: "Generic contracts, such as US1, US2, US3, ..., are constructed by pasting together "rolling" contracts, according to the pre-selected roll types on the commodity default page. The generic contract uses the value of a particular contract month until it "rolls" to the next month in the series. You can access a generic contract by replacing the month/year code with the number 1, i.e. A 1
9 Seasonal commodity graphs are included with the permission of John Bernardi/Newedge USA. Graphs and underlying data prepared by John Bernardi/Newedge USA as of April 2012.