Commodities are very volatile investments, which may be appropriate only for professional investors. Like all other asset classes, timing is what matters most. In this article, which appears in today’s FT, I outline three factors which are supportive of commodity investments in the period ahead. First, this is the stage of the economic cycle when commodities normally out-perform. Second, the shape of the futures curve in many commodities has shifted recently towards backwardation, which is normally good for total returns in the asset class. And, third, the introduction of commodities into a standard portfolio of equities and bonds is likely to bring diversification benefits, even though these have not been apparent in recent years. The main risk to the bullish case is that China may slow more rapidly than is generally assumed, under the impact of tighter monetary policy.
I do not want to repeat all the arguments made in the FT column, but I would like to comment further on a couple of points. What about the argument that commodities and equities are now so closely correlated that there is little point in adding commodities into a portfolio of assets?
The graph shows that the correlation between commodities and other assets is not stable over time. It is not sufficient simply to look back over past periods and then conclude that commodities are positively correlated with equities, or indeed negatively correlated, because these conclusions depend on the time period selected, and the state of the global economy in the relevant time periods.
In the past three years, commodities have been positively correlated with equities. The key decision has been whether to hold either of these two assets relative to the “riskless” asset, government bonds. During these years, relative asset returns were driven almost entirely by changes in global risk appetite. By contrast, in the 1970s, commodities were negatively correlated with all other assets, and the key decision was whether to hold real assets (commodities) or paper assets (stocks and bonds). These relative returns were primarily driven by supply shocks in the oil market. Before investors can decide on the role which commodities should play in a portfolio of assets, they need to consider what type of shocks are most likely in the period ahead.
As explained in today’s FT article, I am worried about the possibility of a commodity price shock in the next couple of years. This would produce negative correlations between commodity returns and other assets, so commodity investments would be one of the few good hedges against this possible shock.
The next question is how to “buy” commodities? This is not by any means as simple as buying equities and bonds. The most common way of gaining exposure to commodities is to purchase products which track a commodity total return index, such as the S+P GSCI. This may not always be advisable. These indices rise in line with an increase in spot commodity prices, but they also stand to gain or lose from the “roll” return which is derived from investing in futures rather than holding physical commodities, which of course are difficult and expensive to store. In recent years, this roll return has been persistently negative, though the shape of the futures curve has improved a lot in recent weeks.
Mainly because of the negative roll return, there has been a strong tendency for “passive” commodity indices to be out-performed by more actively managed commodity products. These include “enhanced” index products, which attempt to avoid negative periods of “roll” returns when futures curves are shaped in an adverse way. They also include commodity products which are protected from large 2008-style collapses in spot prices. And finally they include portfolios of active commodity funds which attempt to provide skill in asset class timing for the commodity complex as a whole, and for one commodity against another. Evidence suggests that an appropriate mix of these actively managed commodity products has produced better total returns, with less downside risk, than relying on the passive index products which became widespread in the late 1990s and early 2000s.
It is sometimes a good idea to implement asset allocation decisions through low cost “passive” vehicles, but in this case many active managers have been able to out-perform the passive approach. These actively managed products may not all be readily available to individual investors – which is another reason to leave this volatile asset class to the professionals.
