THE EAST IS RED
In the second week of last October, after the collapse of Lehman Brothers but before the global financial crisis was confirmed to have become an Australian economic one, Rio Tinto sent a survey team of more than 20 people across China to find out what was going on.
The fast deployment was evidence that Rio was more attuned to the dynamics of the Chinese market than either its competitors or its customers.
Rio’s survey team was hired through a Shanghai consultancy called CBI, which markets itself as providing “a wide range of intelligence products”.
Despite the self-puffery, CBI’s mode of operation was not to give brown paper bags (or red envelopes) in exchange for state secrets or even commercial secrets. The contract and terms of reference made that very clear.
Nor did it set out to discover the internal cost structures and production data of Chinese steel mills. That would not be worth the trouble, as China’s 540-odd steel makers are not known for their business planning and the Chinese Government’s monthly steel production data is fast and reasonably reliable.
Rio only had to glance at its own order books or the iron ore spot market to understand that demand had fallen through the floor. Its challenge was to efficiently collate public data and form a view on leading indicators such as lending growth and floor space under construction.
But a large portion of the supply side of the iron ore equation remained a mystery.
Rio knew how much the rest of the world was producing from sharemarket announcements of Vale and BHP Billiton. But unlike steel data, China’s domestic iron ore production figures are worthless.
Most of China’s 5000 iron ore mines keep a low profile for the sensible reason that they are illegal. They lack some or all of the six government licences required and survive by quietly feathering the nests of officials.
A rough Chinese production number can be implied by working backwards from how much iron ore must have been fed into Chinese blast furnaces, adjusting for stockpile movements and then subtracting the volume of imported iron ore.
But Rio needed better than that, so it sent the CBI team to find and knock on the doors of hundreds of mines until they had a large and reliable sample.
CBI had previously done similar work for Rio, which had allowed the miner to model a cost curve that could predict how much Chinese iron ore mining capacity would be shut down when prices fell below certain thresholds. A version of that cost curve was published by Rio’s chief executive, Tom Albanese, in May last year and has since been borrowed, adjusted and relabelled by leading investment bank analysts to become the industry standard.
Within a fortnight the CBI team had confirmed what Rio’s cost curve model had predicted.
China had been producing about 400 million tonnes of iron ore a year when adjusted for equivalence with Australian imported ore (the raw tonnage is actually much larger because Chinese ore quality is so poor).
About 110 million tonnes was dug from large and generally more efficient mines that were either state-owned or “captured” by a particular steel mill and therefore had muted incentives to shut down. But of the remaining smaller, private, higher cost and often illegal mines, which accounted for 300 million tonnes of Chinese production, CBI discovered half of them (by volume) had shut their doors.
So Rio knew the precipitous decline late last year in Chinese steel production would be offset by global miners being able to displace about 150 million tonnes of Chinese iron ore. Regular iron ore mine surveys continued until Stern Hu was led away from his Shanghai home on the morning of Sunday, July 5.
The China Iron & Steel Association had privileged access to data. Indeed, Rio had formed the view that the association was getting real-time updates from whoever was listening to its phones and watching its email traffic (presumably the Shanghai State Security Bureau). But CISA’s conclusions about the market were always driven by politics and wishful thinking rather than clear-eyed analysis.
CISA continually asserted that Chinese domestic miners would somehow roar back to life so China would not need Rio’s iron ore. And it refused to comprehend the rising steel production trends made plain by official Chinese Government data.
To this day CISA claims China will produce 500 million tonnes of steel this year, even though production rose to a new record above 50 million in July. Through little more than multiplication, Rio predicts a full-year result of about 570 million tonnes.
At each round of iron ore price negotiations - and this year there were surprisingly few - Hu and his negotiating partner would simply point out the window at what was going on in the real world and hold their ground.
The Rio team would say the market was strong; CISA would deny it, and Rio would turn out to be right every time. For CISA it was beyond humiliating.
Rio’s negotiating advantage over its Chinese customers was and continues to be that it collates information systematically and analyses it intelligently. This advantage is compounded because Rio is dealing with a steel association that acts as if it has no idea how far China’s steel and iron ore industry has moved beyond its powers of command.
It would be strange if Rio were ”stealing” state or commercial secrets because there is no Chinese mill, mine or official who has confidential information worth knowing.
Rio Tinto’s intelligence advantage over the Chinese steel industry was more powerful and less sinister. Hu is nevertheless paying the price.
0