China's economic data is dodgy, but it doesn't matter for commodities
Very few analysts believe the output data provided by the National Bureau of Statistics
There is widespread scepticism about the accuracy of mainland China's economic growth numbers, but the real question shouldn’t be whether the data is credible, it should be whether it matters that it isn’t.
The official year-on-year gross domestic product (GDP) growth figure of 6.9 per cent for the September quarter was widely condemned by mainly Western economists as a fiction bearing little resemblance to actual economic activity.
At the World Commodities Week conference in London last week, the consensus was that mainland GDP growth was more in the region of 5 per cent to 6 per cent, and that the official number owes more to political expediency than to economic reality.
While this may indeed by true, the issue really should be whether absolutely accurate data is necessary when trying to get a handle on China.
If you move from the economic world to the commodities world, China watchers have grappled with dubious statistics for so long that they have become inured to them, to the point where some barely reference official numbers.
For example, very few analysts believe the output data provided by the National Bureau of Statistics (NBS).
Iron ore mined in China has fallen by 9 per cent to 1.02 billion tonnes in the first nine months of the year, compared to the same period in 2014, NBS data shows, while steel output has declined 2.14 per cent to 608.9 million tonnes.
When you take customs data that shows iron ore imports were flat at 699 million tonnes in the first nine months, it becomes hard to reconcile iron ore imports and domestic output with steel production.
Of course, some of the discrepancy can be explained by changes in inventories, or changes in the quality of ore being mined domestically and that being imported, but the point is the numbers rarely tally up accurately.
It's the same story with crude oil, where the non-disclosure of changes in inventories means estimating actual demand, and the flow of crude into strategic storage, becomes a game of best guesswork.
Think of a commodity, and very quickly you can find experts in the area who question the official numbers.
Output data for aluminium is too conservative, consumption data for copper is skewed by the practise of using the metal for financing, cotton imports don't adjust for quality, and so on.
A bright spot in Chinese data has traditionally been the customs data, which enjoys a reputation for being timely and relatively accurate, but even here there are likely to be problems, with one analyst at the World Commodities Week summit saying iron ore imports have been overstated.
On the macro level, there is likely to be political interference in the process, leading to GDP numbers that seem to closely match the official growth target.
But as you go further down the chain, the inaccuracy probably stems more from problems in collecting data across such a vast economy, coupled with the providers of that data seeking to please superiors by telling them what they want to hear, rather than what is actually happening.
In some ways this would help explain why steel and aluminium numbers seem to understated.
The authorities in Beijing want to lower output in over-supplied sectors such as steel and aluminium that are also heavy polluters.
Individual steel plants and aluminium smelters thus have an incentive to under-report their actual production in order to show compliance with the centre's wishes, and can do safe in the knowledge that their excess output can “disappear” into the shadow economy if needed.
The picture that emerges is that Chinese data is at best an indicator of broader trends rather than a definitive snapshot of the activity it is supposed to measure.
In this sense it is useful. Even the GDP chart shows a clear down trend from 2010 that fits in with the overall theme of China trying to rebalance its economy towards a more consumer-led model.
The iron ore numbers do confirm that steel output has flattened out over the past 18 months, while the crude numbers do show that substantial volumes are flowing into storage tanks.
While it would be marvellous if Chinese data could be as close to statistically perfect as possible, this is unlikely in the short to medium term as the Chinese themselves likely don't see a need to reach Western levels of accuracy.
Rather, as the commodity world has shown, you can live with imperfection as long as you can discern the broader trends, and when they are starting to turn.