I finally caught-up with some reading and finished Goldman’s note on China’s fixed asset investment (FAI). Published mid-February, the conclusion was pretty clear : “Did China invest too much in 2009?” No, “we believe most of such investment is justifiable for current and future economic development and will not cause a structural over-investment problem, as many fear.”
Those were the main arguments:
1) Yes, too much investment were made if judgment based only on reported investment’s share in GDP (+43%), or the share of its contribution to GDP growth (+92.3%).
2) But it is partly due to the fact that the drag from net export was considerable in 2009, otherwise the share would have been much lower. Investment contributed to +8% of the GDP headline (of +8.7%). Massive. The remaining +0.7% came from consumption, at +4.6%, plus net export, which was a negative contributor, at -3.9%.
3) Also, for 2009, FAI was probably over-reported, as many local governments “recorded the total amount of investment projects (including the uncompleted portions), instead of the actual realized amount.” They wanted to look good.
4) On a more structural note, GS also mentioned a statistical bias leading to an overestimation of the investment series. It is due to an insufficient deduction of land costs and price change, which mechanically increases the FAI number. In the case of land costs, they have been accounted for less than 30% of total real-estate capex spending, while property developers reported numbers between 45% and 50%. The difference appears as investment, but nothing has been built/created. The FAI deflator does the same thing by understating the price change. In 2007 for example, “it was at +3.9% despite the commodity price rally and significant price increases in investment goods.” In the end, counting a portion of the price change as FAI.
5) And finally, “the strong recovery in domestic investment was policy induced initially and can be seen as part of the cyclical adjustment measure (…), especially the government led portion, [which] should be seen as a cyclical management tool.”
Now let’s sum up. Those abnormal high 2009 FAI figures are the consequence of:
- Temporary effects (negative export contribution, over-estimation by local governments and temporary boost as part of a cyclical management tool)
- Structural effects (over-estimation due to statistical bias in the reporting of land costs and change in prices)
So GS made some interesting points. The real impact of FAI, net of statistical bias and temporary policies is probably a few points lower. But is that really the core of the issue? No.
The real matter is not how much FAI has really been taking place in 2009. It is a too short time frame. It also does not include the other side of the story, how much was needed, and will actually contribute positively to future growth. And this is where GS analysis is getting pretty light:
“Some overhang capacity exist in some sectors and areas, but we believe it is not structural.” A strong GDP growth, around 10%, will help removing overcapacity much more quickly than in other slower growth countries. “For example, the excess capacity problem populated in 2002 disappeared quickly in most sectors in 2003, some of which turned into bottlenecks by early 2004.” In the end, this crisis was opportunity to catch-up on some investment needs in infrastructures, which will contribute to future growth.
Overcapacity? Easy, high growth will remove them, as it did before. Nothing on the level of urbanization, or investment capacity already existing. High growth baby. There is nothing it won’t cure. Problem: that was true 7/8 years ago. But high level of investments have been made since. And Jim Chanos said it well, the probability of this strategy to work is diminishing every year, as more bridges, schools and factories are being built, urbanization and production capacity increase, and the need for more decreases, at least relative to GDP.
Now there is a second major point: “China still needs to rebalance toward consumption-driven growth”. How? “by implementing measures that increase household real income.” Hard to argue, every one seems to agree. So where is the difference? On whether or not this process has been already taking place.
GS is optimistic, and will give you 20 figures to show how fast the Chinese consumption is growing… but is it enough? No. As Michael Pettis (http://mpettis.com) often says, high consumption growth is not enough to rebalance the economy. You need consumption to grow faster than GDP, in order to gain a higher share of it. Otherwise, FAI in percentage of GDP will continue to grow. And so far, it has probably not been the case.
In concluison, I am surprise that some of the key points, ie. need for more infrastructures, high GDP growth quickly removing excess capacities and rebalancing on its way, have been the least developed by GS. It is interesting to show that the FAI share of GDP in 2009 might have been a point or two lower than it had been stated. But that is not changing the core of the issue:
- After 20+ years of very strong growth and high rate urbanization, how much more does China need?
- Why should we assume that the same will happen to excess capacities, 10 years into one of the strongest investment cycle?
- What would be necessary for the Chinese economy to rebalance, and has it been put in place?
Chanos vs. Goldman : Chanos wins the first round, on the overall logic and global picture. But the match is not over.
Lesmaes 