Monday, April 02, 2012
TMM were somewhat chuffed to see their data mining and manipulation last week pretty much get the China PMI spot on. But what is the point of getting the number right if the market doesn't respond as expected? We know that China is still on holiday and so hasn't had a chance to act but the fade in Asian hours from opening highs in nearly everything "risk" is somewhat frustrating. Why does everything have to be soooo HARD? Can't we just for once have a "there's-a-good-figure-market-goes-higher" event? This morning though, trying to suggest that the data is good appears to be as difficult as arguing the origins of fossils with a creationist.
The heart of this morning's scepticism with respect to the NBS PMI centres around that old chestnut, seasonality. The trouble with this argument is that it's not exactly as if the realisation that many data series exhibit seasonality occurred suddenly overnight. Indeed, one would expect that expectations for the number embedded such seasonal adjustments, thus expecting on a seasonally adjusted basis that activity *fell*. This clearly has turned out not to be the case. Secondly, as TMM pointed out in their piece last week, there is a good amount of evidence that since 2005, the seasonal factors have been dynamic. Specifically, the 2005-2007 period appears optically to be very seasonal and the 2010-2012 period appears to not be that seasonal, if anything looking a lot like the movements in the PMIs from the rest of world. Which brings us to the 2008-9 period which is known to have caused echo effects in the X-12 seasonal adjustment process elsewhere. Indeed, the very large moves in those years mean that simply comparing the average March vs. February change is somewhat naive and biased. Next, the divergence between the HSBC and NBS PMIs, which can arguably be accounted for by considering the changing composition of Chinese growth becoming more orientated to inland cities over the past year or so.
While this is rather dry and geeky, TMM have become utterly fed up with the Mickey Mouse laziness exhibited by both punters, economists and the media as far as the analysis of the Chinese data goes. TMM's analysis that attempts to take note of the above issues by using a dynamic seasonal adjustment approach, suggests that activity moved higher by around 0.7ppts. Bears: just get over it.
But whatever TMM think or say, a mass escape of bears at the bear house in the national bear institute, Ursine city, look less bearish than the commentaries landing on their desks with respect to China this morning. But the price falls we are seeing so far are more associated to European peripherals and especially the European banks rather than Asia. Do we therefore suggest that the China PMI has been put on the backburner and a new bear toy box opened with respect to European PMIs and slowdown? Very Seasonal that. April - Don't we historically start taking a pop at Europe during April?
As TMM mentioned last week there is a regional flavour as to opinion on China so perhaps we just need to get through a complete trading day cycle before we get a true representation of reaction, but TMM are going to stand by their guns as far as China goes and if we are going to start seeing some spurious euro trashing we will play this as a regional cross trade rather than a global play. But for now .. hold fast boys.