Research examining centuries of commodity price data has tended to sketch a pattern of 15–20 year super-cycles (a period of rising prices), followed by a slide in prices over the following 10–15 years when excess investment leads to a flood of excess supply.
The industrialisation of the US in the late 19th century, and the German build-up prior to World War I, drove an upswing in commodity prices in the early 20th century. The next upswing came after World War II because of the rebuilding of Europe and as the Baby Boomer generation entered the workforce. More recently the industrialisation of Japan, and the population growth and urbanization in many other emerging Asian economies, also contributed to an upswing in commodity prices in the 1970s and the early 1980s.
The most recent upswing in commodity prices had its origins in 1998, when prices approached their lowest for 20 years (equal to depression levels, when adjusted for inflation). This was followed by the emergence of China as an industrial and economic powerhouse, which coupled with a lack of investment in new commodity supply, drove commodity prices up into yet another super-cycle in the first decade of the 21st century.
After peaking in 2011 commodity prices have grounded out a long inexorable decline for the following decade. Excess supplies brought on-line late in the super-cycle became surplus to the overinflated expectations near the peak.
But how does the commodity super-cycle work in practice? Well, both demand and supply for commodities are inelastic in the short term. Essentially this means that it can take quite a long time for consumers and producers to react to pricing signals. When oil prices spike, for example, motorists may have no choice but to continue to use their car to get to work, but eventually they may be able to invest in a more efficient car. Meanwhile, commodity producers take time to invest and then bring on new supplies, i.e., exploring for oil and then eventually bringing the crude to market.
If commodity prices rise, there is a significant delay until production begins to respond. Critically, it takes time for expectations to adapt and even longer for supply to respond to higher commodity prices. Given that the typical time horizon of a major mine can be 20–30 years, with high initial capital outlay and traditionally slow capital return, the planning process is very risk averse. However, once the economic status warrants a mine being brought into production then it typically takes approximately 7–10 years to take the discovery of a new deposit through to production (although lead times can vary greatly between metals).
Two economists outlined put some structure to the theory of commodity super-cycles. The first, Joseph Schumpeter, described periods of short-term volatility that can change the otherwise overriding trend in commodity prices. This change may be driven by an exogenous or unpredictable factor, such as rebuilding a country following a war. But as I outline this cycle has tended to play out in full over a period of 30 years.
Even within these long periods of rising or falling prices there may still be lengthy periods when commodity prices run counter to the longer-term trend. In addition, not all commodities will necessarily behave in the same way during the period of a commodity cycle. While energy and metal markets will undergo long elongated cycles, agricultural markets will be much more closely linked with the much shorter crop cycle.
The second, Nikolai Kondratiev outlined long waves spanning 60 years, closely related to demographic cycles. His commodity super-cycle was based on the idea that economic activity gradually rises and is coupled with low interest rates and rising prices. However, a turning point is reached where asset price bubbles start to form, interest rates rise and economic growth slows. The final phase of the cycle involves recession or depression, and an unwinding of the excesses of the earlier economic boom.
Crucially, two commodity super-cycles (2 x 30 years) have typically coincided with one Kondratiev wave (60 years). The last time the two cycles peaked together was around 1980. Before that it was the 1920’s, the mid 1860’s and then the 1810’s. If the pattern continues then the next peak will not occur until around 2040.
If we are at the beginning of the next super-cycle then investors will be eager to know what will drive it. As before, structural underinvestment will provide the foundations. Our economy may well be increasingly experienced in the digital world, but the physical infrastructure that supports it has been largely forgotten, and underinvested.
Going in to the 2020’s demand for commodities is likely to be largely driven by government policy. This will occur through addressing environment degradation (promoting the decarbonization of the economy), social inequality (income and wealth transfers to those with a higher propensity to spend) and versatility and adaptation (infrastructure improvements and diversification).
Demand will also be supported by positive macro-tailwinds. The depreciation in fiat currencies is likely to increase demand for ‘hard’ assets, which could include both base and precious metals, but also energy which has proven to be a strong hedge against inflation in the past. In addition, many emerging economies will be experiencing a surge in young people entering working age. This will increase demand for infrastructure, housing, and transportation.