We’ve read two very interesting and distinct pieces of analysis today that raise quite dramatic questions about China’s GDP data. Yes, we know you probably weren’t too credulous to begin with, but the details of both are interesting.
First, a column by Toshiya Tsugami in Japan’s Asahi Shimbun Asia & Japan Watch. Tsugami compares official data on fixed asset investment and credit. FAI accounts for about half of China’s GDP growth; and Tsugami totals up four years’ worth of this investment from 2009 to 2012, coming out with Rmb110tn. Yet total bank assets are much, much smaller than this figure would imply:
According to the latest statistics by the People’s Bank of China, the total of all bank deposits and bank loans is 94 trillion yuan and 67 trillion yuan, respectively, and among total bank loans, two out of five are short-term lending less than one year. Investors can seek other financing such as corporate bonds outstanding and assets in various investment trusts in China, which are growing sharply these days and are estimated at somewhere between 20 trillion and 30 trillion yuan.
If they were ‘good’ investments, it might make sense, he says:
If these fixed-asset investments have been yielding high returns, making it possible for investors to recoup their investments and repay their debts in two to three years, the money borrowed and invested in such assets in 2009 must have already been repaid and plowed back into new investments in 2012.
If such a virtuous cycle of investment and reinvestment is actually taking place, there is not much need to worry about investment financing.
But what is actually happening is the exact opposite.
Clearly, investments made in the past several years have been producing poor returns because an excessive amount of money was invested.
Clearly. And also, shadow banking has been playing a rising role in credit provision — indeed, it must play a rising role if the growth is to be maintained while official bank loan targets are also met. Tsugami’s estimate of Rmb20tn to 30tn is not a conservative one, either. He also adds that monetary policy has been relatively tight in the past couple of years despite the 2011-2012 stimulus. The whole column is very interesting and worth a read in full.
[Update: Martin Wolf wrote to make the following point:
It appears from the national accounts that most of the finance of China's corporate fixed investment comes from retained earnings, not bank lending. So I am unpersuaded that the Japanese author is right.
Of course local government infrastructure is largely financed by banks (or other credit) but if the corporate sector doesn't rely so much on credit for its FAI, that would explain a bit of the big discrepancy Tsugami identifies.]
Next China GDP questioner of the day is StanChart’s Stephen Green: This is where the provocative 5.5 per cent figure in the post’s header comes from, in case you were getting impatient.
Green’s critique centres on comparing official GDP data with proxies for both investment and consumption. While his favourite proxies indicate growth began to lift in Q4 last year, Green notes that official quarter-on-quarter GDP data show Q3 as the strongest point in the year:
Green notes that while the first prints of China’s GDP numbers are widely believed to be subject to some ‘smoothing’ over of rapid growth or deceleration, the official trajectory still seems to tell a strange story.
He hypothesizes that it might be a problem with the GDP deflator; the number used to get from nominal to real GDP growth.
First, the GDP deflator (which gets much less press than the CPI or PPI) has been capturing more inflation than the other two indices in recent years, as Figure 2 shows. The GDP deflator is a better gauge of overall price ressures, and we should be grateful that the National Bureau of Statistics (NBS) is able to find new sources of
price data that do not get captured in the monthly CPI and PPI numbers. At its peaks in 2007 and 2011, GDP deflator-measured inflation approached 8% y/y.
Secondly, Green believes both the GDP deflator and the CPI underestimate inflation among services. These include housing the National Bureau of Statistics has no data on rents, so the CPI relies mostly on costs of utilities for housing cost inflation. The CPI estimates of rents seem to be based on mortgage interest rate costs and not housing prices — the latter, Green argues, do flow on to rents eventually. Moreover, services such as healthcare and school fees are too ‘informal’ to be captured by the official CPI data.
To find an alternative method for determining the GDP deflator, Green turns to work by Calla Wiemar on China’s savings rate and its relationship to GDP, which covered the period up to from 1987 to 2008.
Wiemer argues that “household services and manufacturing upkeep” component of the official CPI index is a better measure of services inflation. Based on this, Green updates her names for 2010-2012:
We admit that choosing this proxy for services inflation is a little random, but eye-balling the numbers (Figure 4), the results do not look crazy. Growth boomed in 1992 with Deng Xiaoping’s revival of reform, but then slowed sharply with Zhu Rongji‟s retrenchment policies of the mid-1990s. The economy was further weakened by domestic restructuring and the Asian Financial Crisis in 1997-98. Then, with the state-enterprise reforms of the mid-1990s feeding through, China‟s entry into the WTO in 2001, and strong global growth, China’s economy took off again in 2001, peaking out in 2007. (We worry that official growth numbers during late 2008/09 may suffer from overly optimistic industrial output data.) Growth has slowed considerably since the height of the stimulus. Our guesstimates for the past two years look considerably weaker than the official estimates: our guesstimates for 2011 and 2012 are 7.2% and 5.5%, respectively, compared with the official prints of 9.3% and 7.3%.
Green does stress that this is a guesstimate only. But those who follow Green’s work will know that he’s a fairly level-headed and thorough China-watcher.
Here is why Chinese official economic figures can’t be trusted – Asahi Shimbun AJW
How to (almost) understand China’s quarterly GDP data - FT Alphaville
China’s massive credit dependency – FT Alphaville