Prices and cycles: Why commodity prices swing so hard

2.1 Steep curves: demand that will not budge and supply that cannot

Three properties combine. First, demand for raw materials is price-inelastic in the short run. The material is a small share of the final product's cost: doubling the price of wheat or lead barely changes the price of bread or a car battery, so buyers keep buying. Second, supply is inelastic once capacity is close to fully used, which in competitive markets it normally is. Apart from annual crops, new capacity takes years. Third, for minerals, demand is highly sensitive to income. Most metal goes into construction, capital equipment, vehicles and consumer durables, the sectors that swing hardest over the business cycle.

Put those together and the picture in Tilton and Guzmán follows: a steep demand curve shifting left and right with the economy across a supply curve that turns vertical at capacity. Prices doubling or halving within a year or two is ordinary. Radetzki and Wårell give annual averages: nickel went from about $4,870 a tonne in 1987 to $37,230 in 2007 and back to $9,595 in 2016. The monthly high, in May 2007, was higher still: $52,161 (Humphreys).

Agriculture differs in an instructive way. Demand for food and fibre changes little over the cycle; the shocks come from supply, through drought, frost, pests and disease. So when a harvest fails the price rises, which partly offsets the lost volume in farm income. For miners price and volume fall together in a recession, so revenue and especially profit are far more volatile. Producer pricing does not remove the instability; it shifts it into rationing and swings in sales.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 109-111
• Tilton & Guzmán (2016), Mineral Economics and Policy · pdf pp. 108-119
• Humphreys (2015), The Remaking of the Mining Industry · pdf p. 105

2.2 Inventories amplify demand, and panic buying can double a shortage

Stocks sit at every stage between mine and consumer, and each holder wants stocks in proportion to sales. That turns small changes in final demand into large changes in raw-material demand. Radetzki and Wårell's example has three stages downstream of the raw-material supplier, each keeping stocks equal to half a period's sales. Final demand rises 5%. The retailer buys 5% more for sales plus 2.5% to top up stock, 7.5% in all. The stage above sees a 7.5% rise and orders 11.25%. The next orders 16.9%. A 5% rise at the shop becomes an almost 17% rise at the mine, and the same cascade runs in reverse in a downturn.

Expectations add a second layer. If buyers think prices will rise, they buy now, and that buying is itself demand. In normal times inventories damp volatility; in a scare they become its engine.

The oil market of 1979 shows how far this can go. Yergin reconstructs it. After Iranian exports stopped, free-world output in early 1979 was about 2 million barrels a day below late 1978, a loss of only 4-5% against demand of about 50 million. Yet companies, utilities and motorists, fearing a repeat of 1973, built stocks: drivers who had run on a quarter tank now kept three quarters. That added about 3 million barrels a day of demand above consumption. The total shortfall became 5 million barrels a day, about 10% of consumption. In Yergin's account it was this stock building, more than the lost Iranian oil, that carried prices from $13 to $34 a barrel.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 110-111
• Yergin (1991), The Prize · pdf pp. 1046, 1063-1064, 1080-1081, 1181-1187

2.3 Shocks hit a system with little slack: weather, war and policy

The effect of a shock depends on where the market sits on its supply curve. With spare capacity and ample stocks it is absorbed; near capacity with thin stocks it explodes. Radetzki and Wårell note that booms need tight capacity and thin stocks, conditions that long periods of low prices create by discouraging investment and stock holding.

Weather is the classic agricultural shock, though the spread of farming across more regions has reduced its global effect. Policy shocks are often bigger. In 1972 a Soviet buyer, Belousov, negotiated separately with each US grain house; none knew how much the others had sold. He bought almost 20 million tonnes of grain and oilseeds, including 11.8 million tonnes of wheat, nearly 30% of the US harvest. Kansas milling wheat rose 60% in ten weeks and, Blas and Farchy report, tripled within a year. In 2010 Russia's drought and an export ban lifted wheat 15% in two days.

The 2022 shock combined war, energy and fertiliser. Natural gas is the main feedstock and cost of nitrogen fertiliser; Russia and Belarus are large potash producers; Europe relied on Russian pipeline gas. The tiles record the result. European gas, which averaged $1.58 per million Btu in May 2020, reached $70.04 in August 2022. Urea went from $201.9 a tonne in May 2020 to $925 in April 2022, potassium chloride from $220.4 in April 2020 to $1,202 in April 2022, and US hard red winter wheat to $522.3 in May 2022. Gas moved most, but potash, which needs no gas, rose more than urea: supply fears mattered as well as feedstock costs.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 111, 127, 137-138
• Blas & Farchy (2021), The World for Sale · pdf pp. 42-43, 236
• Conway (2023), Material World · pdf pp. 23, 140
• USGS Mineral Commodity Summaries 2026, Nitrogen (Fixed)-Ammonia; Potash web ↗ · Salient statistics; World mine production

2.4 Holding prices still: buffer stocks and the tin crisis of 1985

Governments tried to stabilise prices through international agreements for sugar, wheat, coffee, cocoa, rubber and tin. The tools were export quotas and buffer stocks that bought below a floor and sold above a ceiling. Radetzki and Wårell identify two flaws. Stabilising price can destabilise income: with supply shocks, a fixed price means revenue falls exactly when a harvest fails, removing the natural cushion. And nobody can identify the long-run equilibrium around which to stabilise, so the agreements drifted into defending prices that were too high.

Tin is the cautionary tale. The International Tin Agreements, running from 1956 to 1985, used a buffer stock and producer export quotas to keep the LME tin price within a band and, Tilton and Guzmán note, at times above its competitive level. In the early 1980s the International Tin Council defended a price well above equilibrium. Export curbs were not enough, so the buffer-stock manager kept buying. To stretch his money he pledged the swelling stock as collateral for loans and used them to pay margins on futures purchases. In October 1985 the money ran out. He stopped operating and defaulted on his futures commitments, leaving debts above $1 billion.

The LME, which then had no clearing house, suspended tin trading. When it reopened in June 1989 the price was far lower, and the exchange was reorganised around a clearing house. Tilton argues two decades of artificially high prices had encouraged capacity and discouraged use, depressing tin for five to ten years afterwards. By the early 2000s no commodity agreement still tried to manage prices.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf pp. 31, 151, 275-276
• Tilton & Guzmán (2016), Mineral Economics and Policy · pdf p. 101
• Tilton (2003), On Borrowed Time? · pdf p. 58

See it in the table

Check your understanding

In 1979, why did a loss of 4-5% of world oil supply lead to a rise in price from $13 to $34 a barrel, according to Yergin?

What was the immediate cause of the International Tin Council's collapse in October 1985?

Why can stabilising the price of a crop destabilise farmers' incomes?