From Gerd Kommer and Daniel Chancellor
In addition to stocks, bonds, real estate, precious metals and collectibles, commodities are one of the six main asset classes in which investors can invest. In this blog post we will examine whether and under what circumstances a raw materials investment is attractive from a private investor's perspective.
Raw materials can be divided into five main groups: energy (coal, petroleum, natural gas), precious metals (gold, silver, platinum, palladium), base metals (copper, lead, zinc, aluminum and others), minerals (bauxite, phosphate, sulfur and others) and agricultural raw materials (e.g. grain, tobacco, coffee, meat). There are around 40 individual raw materials across all five groups, if you ignore those that are less important in the global economy.
From an investment perspective, precious metals and gold in particular are often classified as an independent asset class, separate from the other “commercial” raw materials that – in contrast to gold – are actually “consumed” or “used” in the economic production process. For the purposes of this blog post, we will follow this convention. The following statements refer to raw materials excluding precious metals. Gold, we have a separate one at the end of 2019 Blog post considered.
If you compare the annual global production volume of the raw materials market with the global market capitalization of the asset classes files, bonds and real estate, the commodity market asset class is very small or downright tiny.
It may sound surprising to some: private investors cannot invest directly in commodities (again with the exception of precious metals). For direct raw material investments – in financial jargon “commodity spot market investments” – the storage costs would be far too high. Try buying iron ore, oil, coal or wheat for 10,000 euros and then necessarily take delivery of it. For private investors, any conceivable return from such a project would be eaten up by the costs of transport and storage.
If you want to invest in raw materials as a private investor, you have to do so via “raw material futures”. Commodity futures are nothing esoteric and ultimately no more complicated than stocks or bonds, but they work very differently.
Commodity futures are commodity futures contracts, i.e. “forward transactions”. In the case of a forward transaction, the transaction conclusion date, on which the price and the other important conditions of the transaction are determined, and the later delivery and payment date of the purchased item fall apart. Actually a banal thing, because most of us have probably made a lot of “future transactions” in our private lives without feeling it as anything special and probably without noticing it.
The first organized futures transactions and markets existed in Japanese agriculture among rice farmers and their buyers as early as the 17th century. Futures markets are just as old and perhaps even older than the first stock and bond exchanges. In a forward transaction, the seller (raw material producer, e.g. a rice farmer) and a buyer (e.g. a rice wholesaler) agree on day T0 to deliver and sell goods (here rice as an agricultural raw material) on a future date, e.g. B. T180 (six months after T0). However, the price is already binding between the two parties in T0, but delivery and payment only take place in T180. This has the advantage for the farmer that he already knows on day T0 what price he will get for his goods in six months. Through the appointment arrangement, the farmer buys “price insurance” and can thus manage and calculate better and more reliably. For example, it is protected from the risk of a potentially existential decline in prices due to a general record harvest. The farmer does not have to spend any money directly on the insurance.
The buyer (the rice wholesaler), on the other side of the futures transaction, is protected from price increases between T0 and T180. This also gives him more planning security and also the risk-reducing advantage that he only has to pay for the rice in T180 when it is physically delivered.
A commodity futures ETF works basically the same way. In this case, the private investor indirectly takes on the role of the buyer of goods. However, the investor deposits the purchase price for T180 (unlike the rice wholesaler) in T0 and receives the return of the money market up to T180 (super-safe short-term government bonds). However, with a commodity ETF, there is not just one 6-month contract, as in our rice farmer example, but many that are continually renewed (sold and replaced) on a rolling basis just before they expire. In addition, the vast majority of commodity ETFs are not just about a single commodity, but rather a logically defined group of commodities (a commodity futures index).
The total return for the buyer of a futures contract or a commodity futures ETF is as follows:
(a) The underlying spot market price changes of the commodity over time. In the long term, this return is likely to be between 0% and 1.5% p.a. in real terms (adjusted for inflation) (before costs and taxes). More on that below.
(b) The so-called “roll return”, which results from the difference between the spot market price on the redemption date of the futures contract and the previously determined futures price. If this difference is positive for the buyer (the spot market price is above the futures price) there is a profit, a positive roll yield. If raw material futures are to be attractive investments, the roll yield would have to be at least slightly positive on average over the long term. However, adjusted for inflation, it has been predominantly negative over the past decade.
(c) The interest income from short-term government bonds, which is generated from the capital deposited by the buyer of the contract (collateral for the contract) for the purchase price.
Commodity futures returns can and often will be significantly decoupled from commodity spot market returns over periods ranging from one year to around ten years. Ultimately because they are two different markets.
The long-term inflation-adjusted price increase of raw materials measured on the spot market is very low; it is not far above zero. What the actual value looks like depends on several factors: Does the data series start before, during or after the industrial revolution? (During this historically unique phase from around 1780 to 1870, raw material prices rose sharply before later falling again.) What is the composition of the raw material index to be considered? When it comes to raw material indices, there is no logical, generally accepted weighting method for the many individual raw materials or raw material groups, such as B. is the case with stock indices or bond indices. Energy raw materials seem to offer slightly higher returns in the very long term than, for example, B. have agricultural raw materials, but in general there is no raw material group or individual raw material for which the rough long-term guideline value of 0% p.a. to 1.5% p.a. does not apply. Oil, the most important raw material in terms of production value, had a real return (price increase) of 1.24% p.a. in USD (before costs) in the 120 years from 1900 to 2019 - so not very much.
These low numbers will come as a surprise to those who believe that raw materials will inevitably become increasingly scarce and that their prices will therefore inevitably rise in the long term (prices are indicators of scarcity). The theory of raw material scarcity is an ineradicable fairy tale that prophets of doom have been spreading for a good 150 years. Some representatives of the ecology movement that emerged in the 1970s agree with this nonsense because it fits their desired worldview.
The first prophet of doom to spread the seemingly plausible thesis of increasingly scarce raw materials at the end of the 18th century was the British economist Thomas Malthus (1766 – 1834). He saw a growing world population and contrasted it with the earth's supposedly finite, limited supply of raw materials. From this he concluded that humanity was heading towards the inevitable catastrophe: famine or war.
The mistakes that Malthus made and that his brainchildren continue to make to this day: period-specific developments in the recent past are extrapolated unchanged into the future, the market price mechanism and the incentive effect of rising prices are not understood, raw material deposits known today are viewed as the only relevant ones, the very likely discovery of large, currently unknown raw material deposits in the future, especially if prices rise, is ignored, as is raw material substitution and saving through technical progress. The fact that agricultural raw materials grow back is “forgotten”. The Earth is viewed as a closed system, which it is not thanks to the gigantic supply of energy from the sun. And in turn it is “forgotten” that the growth rate of the world population has been falling for around 50 years and will probably fall to almost zero in the next 60 to 80 years (see English Wikipedia “Projections of population growth”).
So much for the historical context of a humanity that is confronted with the supposed finiteness of raw materials and their scarcity.
In the table below we look at the historical returns of commodities and compare them to two other asset classes. We differentiate between the non-investable raw material spot market (direct investments in raw materials) as mentioned above and the investable raw material futures market (commodity futures market), here in the form of the best-known raw material futures index GSCI Total Return Index.
You can invest in the commodity futures market “actively” (through certain hedge funds) or “passively” through an ETF that tracks the GSCI index or another commodity futures index.
Table: Real (inflation-adjusted) returns of commodities and selected other asset classes in USD before costs and taxes from January 1970 to November 2020 (50.9 years)
► [A] Bank of Canada Commodity Price Index. A mixed commodity index that tracks the prices of 28 commodities and is weighted by production volume. As mentioned in the text, this index is not investable. ► [B] Standard and Poor’s Goldman Sachs Commodity Index Total Return, the best-known commodity futures index with the furthest data history. Many commodity ETFs in the USA track this index. ► [C] Maximum cumulative loss (real) over the entire period. ► [D] Annualized standard deviation of real monthly returns. ► Data sources: Bank of Canada, Standard & Poor’s, Bloomberg, MSCI, Dimensional Fund Advisors.
What can be read and concluded from the table?
— The commodity spot market (BCP index) has delivered a negative real return (–1.1% p.a.) over the past 51 years. This index is not investable for the reasons stated in the text.
— The S&P GSCI TR index, in which you can invest via ETF, produced a tepid real return of 2.0% p.a. over these 51 years. This result appears completely “unacceptable” when you look at it against the background of the enormous risk that came with it and when you also look at the return pattern over the five individual decades. In each decade the return was lower than the previous one. A worrying trend, the probable cause of which we will return to below. At best, the low correlation to the MSCI World stock index of +0.20 could be seen as a small ray of hope, but this too is put into perspective when one realizes that the correlation has increased significantly in the recent past (not visible in the table). What is particularly negative is that in the past the correlation to the stock market often increased during stock market crises and therefore the hoped-for diversification did not take effect precisely when it was particularly important. Even in the most recent Corona crash, raw materials collapsed as hard as or more than stocks and, in contrast to them, raw materials have only recovered to a small extent to date (January 5, 2021).
— High-quality bonds not only had better returns than commodity futures over these 51 years, they were also significantly lower risk and made a stronger diversification or hedging contribution to stocks, as evidenced by their lower correlation (+0.08) to the highest-yielding of all asset classes: stocks.
Why has the return trend on commodity futures been so consistently downward over these 51 years and over the last 40 years? This could have three causes:
First, the abnormal, enormous rise in raw material prices in the 1970s (until the end of 1980), which occurred as a result of the politically motivated oil shortage by the Arab-dominated and then very powerful OPEC. Secondly, inflation rates have fallen quite steadily and significantly in most developed countries from the early 1980s to today. Third, the so-called “financialization” of the commodity futures market since the mid-1990s.
On the rise in oil prices in the 1970s: The repetition of a real tenfold increase in the price of oil as from 1972 to the end of 1980 seems unlikely in the future.
Regarding inflation: Commodities usually have good returns when inflation rises noticeably or is high. That was the case in the 1970s. Should inflation rise sharply again at some point, commodity futures are likely to benefit from this in terms of returns.
Regarding financialization: This technical term means that from around 1995 onwards, raw material futures investments by pure financial investors (professional and private investors) increased significantly. Previously, only raw material producers and raw material traders were active in this comparatively small market. Now financial investors joined the traders. This reduced the returns on the dealer or investor side and increased the correlation to the stock market. Much has been written in science about financialization in the commodity futures market and its effect on reducing returns and increasing correlations. There is currently no final consensus on this. From our perspective, the majority opinion in academia tends to consider the financialization argument to be valid. We do that too. At the end of this article, three academic articles are mentioned that represent the opposing positions (return-reducing effect of financialization yes/no).
There are currently 24 broad commodity futures ETFs available in the German private investor market, which track several slightly different futures indices. Due to the differences in the indices, the historical returns of the individual ETFs can vary quite significantly, especially over shorter periods of time up to around five years. In individual cases, special features can also lead to different returns, such as currency hedging or the exclusion of individual raw material groups.
So-called raw material certificates pose as raw material investments, but in reality they are bank bonds whose pay-off profile (the return characteristics) is linked to a specific individual raw material or a rather arbitrarily thrown together basket of raw materials. If the bank that issued the certificate fails, there will probably be a large loss on such pseudo-commodity investments. (In the USA and most other Western countries with better investor protection than in Germany, certificates, known there as “structured notes”, are significantly not permitted for distribution to private investors.)
In general, raw materials are used by supporters of the real asset concept (which we see from the investor's point of view consider it rather useless) assigned to the so-called tangible assets. Curiously, however, it remains unclear whether this assignment should also apply to raw material futures.
Private investors who are sympathetic to sustainable investing for ecological or ethical reasons will tend not to be supporters of commodity investments. However, due to space limitations, we cannot expand on this argument here.
The stocks of raw materials companies generate higher long-term returns than raw materials, but are not a good substitute for real raw materials investments because such stocks correlate more closely with the general stock market than with the raw materials market.
In some ways, it helps us all when commodity prices fall. Then we all benefit from it as consumers. Conversely, if they rise, that will tend to help commodity futures investors. Only “tends to” because the futures market often develops differently than the spot market.
Conclusion
Direct investments in physical raw materials are not possible for private investors, with the exception of precious metals.
Commodities - both the spot market and the futures market - have had historically low returns combined with high volatility (intensity of fluctuations). Even in stock market crises, raw material investments are not a reliable hedge.
The return on commodity futures has trended down quite steadily over the past 40 years. The main reasons for this could be falling or low inflation and financialization.
Commodity futures still have an attractively low correlation to stocks. However, in our opinion, this is not a sufficiently strong argument for raw materials as an addition to a global portfolio. The returns on raw materials are too low and the stability of raw materials in stock market downturns is too unreliable.
The return on commodity futures could increase significantly in the future if inflation increases significantly and/or there is a severe global political (not global economic) crisis that leads to fears of supply shortages and hoarding.
Anyone who wants to invest in commodity futures is well advised to follow a few simple rules: Commodities should not make up more than 5% to 10% of the risky portfolio portion of an investor's household and the ETF should reflect a commodity index that is as broad as possible. In our view, ETFs on individual raw materials or groups of raw materials are a bad idea due to their increased risk without additional expected returns.
With regard to the question of whether a moderate allocation of raw materials makes sense in a global portfolio, we consider the arguments presented here against them to be more convincing than the arguments for them.
literature
The first two of the following three essays take a critical view of raw material futures and their financialization, the last essay takes an opposing position:
Erb, Claude/Harvey, Campbell (2016): “Conquering Misperceptions about Commodity Futures Investing”; In: Financial Analysts Journal; Volume 72; Number 4; 2016
Bianchi, Robert/John HuaFanNedaTodorova (2020): “Financialization and De-Financialization of Commodity Futures: A Quantile Regression Approach”; In: International Review of Financial Analysis; Volume 68, March 2020
Bhardwaj, Geetesh/Gorton, Gary/Rouwenhorst, Geert (2015): “Facts and Fantasies about Commodity Futures – Ten Years Later”; May 25, 2015; Internet reference: http://ssrn.com/abstract=2610772