China's recovery not ironclad
Commodity producers and traders are reading too much into PMI numbers, and should be looking to other indicators of import demand
Commodity producers and traders have no doubt been cheered by the recent recovery in China's key manufacturing sector, but the boost may be due more to sentiment than actual demand.
This is because there is a fairly weak correlation between movements in the official purchasing managers' index (PMI) and imports of key commodities such as crude oil, iron ore and copper.
There is a far better correlation between China's imports and these commodities' prices. This suggests that while stronger, or weaker, industrial growth helps set the direction for imports, the size of the movement in imports is more related to price.
The official PMI rose to a 16-month high of 51 last month, beating market expectations for a reading of 50.6, with the breakdown showing better conditions across the factory sector, including the export-orders category.
The rise in the official PMI was supported by a similarly positive reading in the HSBC PMI survey, which rose above the 50 line that separates expansion from contraction, hitting 50.1 last month for the first time in four months.
The HSBC index is more biased towards small and medium-sized enterprises, while the official PMI concentrates on larger, state-controlled companies.
Both PMIs appear to be saying the Chinese economy has turned the corner from a weaker start this year and is once again expanding on the back of increased infrastructure spending and better global economic conditions.
It seems logical that this is good news for commodity producers, as a stronger China generally means the world's biggest commodity consumer imports more. But the logic doesn't really stand the scrutiny of the data.
Take iron ore, for example. From the end of the 2008 global recession until the end of 2010, the official PMI was consistently above 50, with the lowest reading at 51.2 in July 2010. But ore imports flatlined for much of that period once the initial rally after the recession was over.
More recently, the official PMI has been meandering in a narrow range near the 50 level, but iron-ore imports have been surging, reaching a record high in July of 73.1 million tonnes.
However, if you compare iron-ore imports to the Asian spot price, it becomes clearer that Chinese buying accelerated after last year's sharp price decline.
It also shows that the weakest month for iron-ore imports this year, namely February, came after the price had rallied almost 80 per cent between September of last year and January.
It is the same story with crude-oil imports. They rose along with the PMI after the 2008 financial crisis.
However, after that, oil imports were weak at the end of 2010, while the PMI was strengthening, although they both rose in tandem towards the end of 2011.
This year, crude imports have been trending higher, reaching a record in July, while the PMI has been tracking sideways.
Looking at crude imports compared with the price of Brent, it seems that higher oil costs lead to lower imports and vice versa.
Oil imports trended higher until September 2010 before moving sideways for about a year, at a time when Brent prices were above US$100 a barrel. Last year, crude imports dropped around September, just as Brent was rallying, and this year the lower oil price that prevailed until recently has led to accelerating imports.
With copper, imports were trending higher in 2011 at a time the PMI was trending lower. So far this year, imports of the industrial metal have been declining while the PMI has been steady.
Comparing copper imports with London benchmark prices shows imports were weaker when prices were strong between July 2010 and July 2011, but imports rose later in 2011 as prices weakened. More recently, weaker prices since April have resulted in gains in imports.
What the data shows is that China's commodity buying is more leveraged to price movements than industrial-output growth, as the Chinese appear willing to use inventories and increase domestic output when prices for imports rise.
A better indicator of likely import-demand growth than the PMI may come from watching inventory cycles, the cost of domestic production for commodities with significant local output, and movements in international prices.