The latest ‘problem’ with China’s economic output data revolves around the calculation of the GDP deflator. Here’s the issue: effectively, the assertion is that China’s deflator simply tracks producer prices, and thus when import prices slide, the deflator understates domestic inflation and therefore overstates real GDP. In the simplest possible terms: when commodity prices are falling, China (and other EMs) may be routinely overstating GDP growth. FT noted this discrepancy last quarter and indeed, China’s statistics bureau actually went out of its way to refute the suggestion that its deflator math was deficient. “In general China’s GDP deflator hasn’t been underestimated, nor has GDP growth been overstated. Both objectively reflect the real situation,” the NBS said in a statement.
Yes, “in general”, China’s economic data reflects “the real situation,” but that’s not very comforting when the future of global trade effectively rests on whether or not Beijing’s transition from an smokestack economy to a consumption and services-led model turns out to be a complete “made in China” disaster.
That’s why the deflator issue matters so much.
If the country is effectively destined to overstate GDP growth when import prices are falling and understate growth when import prices are rising, then the market will be misled about the health of the world’s second-most important economy just when it needs the truth (i.e. when the world is struggling with lackluster demand and a deflationary supply glut) and perhaps just as dangerous, the numbers won’t reflect the degree to which China’s economy is overheating in boom times.
Here’s further color on this issue from SocGen’s Albert Edwards:
The slew of economic data out of China this week had economists chuckling into their GDP spreadsheets. No-one I meet really believes the economy is growing anywhere near the 7% the Chinese Statistics Bureau insists is the correct number. But that doesn’t really matter. What matters is that the Chinese policy makers are throwing the proverbial kitchen sink at a spluttering economy and a faltering stock market (the latter having been stoked up to help the former). So far investors have failed to appreciate the futility of their efforts as centrally planned economies and markets will surely fail sooner rather than later.
When I read a quote in the FT from one economist that “The government could not have hoped for a more perfect set of data”, I did actually laugh out loud. But National Bureau of Statistics (NBS) spokesman Sheng Laiyun, in a robust statement on Wednesday, said that “China does not underestimate its GDP deflator and we don’t overestimate our GDP”. This is in response to some interesting articles that suggest that China’s GDP deflator has been underestimated for various technical reasons, and by underestimating GDP inflation the Chinese NBS is ‘inadvertently’ overstating real GDP growth (H/T to Zero Hedge). Certainly, the sharp Q1 dive in GDP inflation into outright deflation was significant as far as we are concerned for it helped explain the authorities’ shift to a much more aggressive easing mode. In Q2, GDP inflation popped back up to rise by 0.1% yoy instead of the 1.1% decline in Q1 (see chart below). Although higher, that is still not a ‘good’ outturn.
And more from Bloomberg:
One reason being touted to explain the gap is that China miscalculates the so-called GDP deflator, a broad measure of prices in the economy.
Capital Economics Ltd. argues that China’s GDP deflator is underestimated in periods when import prices are falling less than producer prices, hence the boost to real GDP.
“It’s an esoteric point, but one with big implications: if the deflator is understated and nominal GDP growth is not, real GDP growth will be reported as higher than it really is,”Mark Williams, Chief Asia economist at Capital Economics in London, who formerly advised the U.K. Treasury on China, said in a note.
In other words, China isn’t netting out the changes in import prices when measuring overall price changes in the economy.
Skepticism over Chinese economic data isn’t new and economists frequently question whether quarterly GDP accurately captures what’s happening on the ground.
For their part, officials from China’s National Bureau of Statistics defend their numbers and differ with the analysis by Capital Economics.
Still, Capital Economics is standing by its analysis. The firm agrees that China’s economy may have stabilized or even accelerated after a sluggish period, only at a much slower pace.
It reckons that in the second quarter deflators for primary sectors like agriculture and services rose while the deflator for secondary industry tracked producer price inflation. And unlike the first quarter, import prices didn’t register big moves.
“We still believe there’s a problem,” Williams said in the note. “Accordingly, as in Q1, we think that real GDP is being overestimated by one-to-two percentage points.”
And finally, some short commentary from Citi:
The growth rate is again over-stated, in our view
The GDP deflator rebounded from -1.1%yoy in 1Q to 0.1%yoy in 2Q, while headline inflation was largely stable, but its gap with our estimates dropped from 1.4ppts to 0.4ppt. The growth divergence between the service and manufacturing sectors, however, has been the key surprise to our growth forecast. Power production growth was barely positive at 0.5%yoy in 2Q, and industrial profit growth was down by 0.8%yoy in Jan-May. By printing out another 7% growth, it again shows that the Chinese authorities can tolerate little volatility, not only on GDP growth, but also on FX, rate, property and equity markets, which could come at a cost of enlarged volatility in the future.
Until this issue is addressed, China’s GDP data will be subjected to what we’ve dubbed the “deficient deflator” theory each and every quarter and as noted on Thursday, because the deflator math issue points to a specific deficiency in statistical analysis (as opposed to relying on sweeping accusations about a generalized and endemic lack of transparency) it seems to merit a response from China’s National Bureau of Statistics.
And by that we mean a real response. Saying that the data “generally” reflects a “real situation” hardly suffices.
