How commodities move: The economics of moving bulk

3.1 Freight works like a tariff, and its fall globalised one market after another

Before the mid-1800s long-haul freight was prohibitive for anything but luxuries such as spices, coffee and precious metals. Radetzki and Wårell describe two spurts that changed this, each globalising markets that had been local or regional. Globalisation here means converging prices across regions, not only bigger flows.

The first spurt was steam, on rails and at sea. Bairoch's figures for shipping cotton and wheat from New York to Liverpool, in constant 1910-1914 dollars per ton, fell from 55.1 in 1825 to 15.7 in 1857, 8.6 in 1880 and 3.5 in 1910. Between 1880 and 1910 transatlantic freight on wheat fell from 18% to 8% of its US price. Odessa lost its grain trade to American wheat and Russian railways, and Britain's bread from home-grown grain fell from two thirds of consumption in the 1850s to a fifth by the 1880s. Refrigerated ships in the 1880s added meat and fruit.

The second spurt was the specialised bulk carrier, triggered by the 1956 Suez crisis and felt from the 1970s. In 1960 nearly all traded iron ore and coal moved in ships under 40,000 deadweight tonnes. By 1988 ships above 100,000 tonnes, which had not existed in 1960, carried 70% of iron ore and 40% of coal. Seaborne iron ore rose from 23% of world output in 1960 to 36% in 1990, and seaborne coal from 2% to 9%. Over the 1870s to 1990s, bulk transport costs fell by almost 90%.

Now almost anything with a perceptible value now trades globally, down to scrap metal and garden stone. Gas, as later sections show, is the main exception.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities in the Global Economy · pdf pp. 23-28

3.2 Bigger ships cut the cost per tonne, and Suez gave the push

Yergin explains why the Suez crisis changed ship design. After 1956 oil companies no longer trusted the canal, and pipelines had proved equally vulnerable when Syria cut the Iraq Petroleum Company's line. The safe alternative was the long route around the Cape of Good Hope, which made sense only with far larger tankers. Japanese yards, using better diesel engines and steels, built ships much of the industry thought impractical. By 1967 supertankers five times the 1956 size were in service, and six Japanese-built ships of 300,000 deadweight tonnes, seven times the standard 1956 tanker, were about to start shuttling between the Gulf and Europe.

The economics of scale are geometric. Carrying capacity grows with the volume of the hull, while steel, crew and engine power grow more slowly, so cost per tonne-mile falls as ships grow. The limits are draught and length at ports and canals.

The ship classes reflect those limits. Smil notes that most dry bulk carriers by number are under 55,000 deadweight tonnes, but Panamax ships (60,000-80,000 tonnes), Capesize ships (80,000-200,000) and very large carriers (over 200,000) hold most of the capacity. Capesize ships, too large for Suez, sail round the capes; Smil gives a typical capacity of 140,000 deadweight tonnes, Humphreys a typical cargo of about 175,000 tonnes. Vale's Valemax carriers take 400,000 tonnes and can berth at only a few ports, such as Dalian and Qingdao. In oil, the very large crude carrier (VLCC) holds up to 350,000 tonnes, about 2 million barrels. Conway notes that a laden VLCC is too deep for Suez, so it pumps part of its cargo into a pipeline at Ain Sukhna and reloads on the Mediterranean side.

• Yergin (1991), The Prize · pdf pp. 857-858, 966
• Radetzki & Wårell (2021), A Handbook of Primary Commodities in the Global Economy · pdf p. 25
• Smil (2016), Still the Iron Age · pdf pp. 125, 183
• Humphreys (2015), The Remaking of the Mining Industry · pdf p. 74
• Conway (2023), Material World · pdf pp. 276-277
• Blas & Farchy (2021), The World for Sale · pdf p. 105

3.3 Freight rates swing hard because the fleet cannot grow quickly

In the short run the supply of ships is almost fixed. A new vessel takes years to order and build, and in a boom few ships sit idle. Demand for ship capacity, measured in tonne-miles, can jump at once: China buys more ore, a European drought pulls in American grain, or a closed canal lengthens every voyage. A fixed supply meeting a shifting demand produces violent price swings.

Geman documents the 2003 squeeze. China raised its iron ore imports by nearly 40 million tonnes, or 33%, in a year. The cost of shipping ore and soybeans nearly tripled. The spot rate to hire a typical bulk carrier rose from $20,000-28,000 a day in January 2003 to about $75,000 at year end, while firms that had bought forward freight in 2001 or 2002 had locked in about $30,000. Around 40% of cargoes move in ships hired on the spot market. Delivery times for new bulk ships stretched from 18-24 months to as much as three years.

Radetzki and Wårell show the Baltic Dry Index soaring before 2008 and collapsing after. They cite Wårell's finding that the 2007-2008 freight spike was a major cause of the breakdown of annual iron ore benchmark pricing. Humphreys explains the link: under benchmarks, Asian mills gave Brazilian ore concessions to offset its longer voyage, which spot prices delivered to China made hard to sustain.

Traders hedge with forward freight agreements, private contracts on a route's rate. Pirrong describes Trafigura selling one as soon as it charters a ship, to protect against falling charter rates. Traders rarely own the ships themselves: a competitive charter market makes ownership unnecessary, whereas terminals, which have no such market, they often own.

• Geman (2005), Commodities and Commodity Derivatives · pdf pp. 28-31
• Radetzki & Wårell (2021), A Handbook of Primary Commodities in the Global Economy · pdf pp. 28-29
• Humphreys (2015), The Remaking of the Mining Industry · pdf p. 126
• Pirrong (2014), The Economics of Commodity Trading Firms · pdf pp. 28, 44-45
• Morgan (1979), Merchants of Grain · pdf p. 277
• Blas & Farchy (2021), The World for Sale · pdf p. 39

3.4 Pipelines, LNG, rail, river and containers divide the work

Gas is the commodity transport cost has never fully globalised. It is so bulky that moving it takes a large share of the delivered price. Until the 1980s pipelines dominated, with costs rising with distance and more steeply under the sea, so three separate markets formed: North America, Europe with Russia, and Japan, Korea and Taiwan supplied by LNG. Until the mid-1990s East Asian prices were twice those in the United States. Liquefaction compresses about 1.4 cubic metres of gas into a kilogram of LNG, which is what makes shipping feasible. A pipeline ties one seller to one buyer for decades and is exposed to sabotage, as Nord Stream showed in 2022. LNG costs more per unit but can be redirected: when Russian pipeline flows to Europe fell from 191 billion cubic metres in 2019 to 51 in 2024 (Energy Institute data), American LNG filled much of the gap.

Inland, rail and river decide which interior deposits reach the coast. Rio Tinto runs 1,600 km of track linking its Pilbara mines to port. Morgan describes New Orleans as the neck of a funnel through which Midwestern corn and soybeans arrive by barge, truck and rail, with the big merchants owning or leasing most of the eight great elevators along the lower Mississippi. When the Soviets bought heavily in 1972, thousands of grain trains jammed the Gulf ports, and in 1973 small elevators could not get hopper cars.

Containers carry the high-value end. A 20-foot box holds up to 21.7 tonnes, and container cargo has been the fastest-growing part of seaborne trade. Coffee, cocoa and cotton, worth thousands of dollars a tonne, can bear box rates and gain smaller lots and easier tracing.

• Radetzki & Wårell (2021), A Handbook of Primary Commodities in the Global Economy · pdf pp. 26-29
• Conway (2023), Material World · pdf p. 278
• Smil (2016), Still the Iron Age · pdf pp. 125, 183, 185
• Morgan (1979), Merchants of Grain · pdf pp. 228, 431-432, 441
• Pirrong (2014), The Economics of Commodity Trading Firms · pdf p. 8
• Energy Institute, Statistical Review of World Energy web ↗ · Russian pipeline gas exports to Europe, 2019 and 2024

See it in the table

Check your understanding

Why do dry bulk freight rates rise so sharply when Chinese ore imports jump?

In Lundgren's figures, what happened to the freight advantage of Swedish ore over Brazilian ore in the German market between 1960 and the early 1990s?

Pirrong argues that commodity traders seldom own bulk ships but often own terminals. Why?