Exclusively from Michael Pettis’ newsletter.
For those who are interested in this sort of thing (and I confess I most certainly am), transcripts of the Bretton Woods conference in 1944 have recently been discovered and are being published on Kindle. I think a paper book will come out in 2013.
I have just finished reading through a selection of the debates sent to me by Gao Ming, my former PKU student, now at Harvard, and one part of the debates that I found especially interesting was a debate on workers’ remittances. The USSR had insisted during the meetings that workers remittances be included in the capital account, and so subject to capital controls. Several other countries, led by China, were opposed. The latter won the debate, and today, of course, workers’ remittances are included in the current account.
Among other things this shows just how wobbly some of the capital account/current account distinctions are. I think from a political point of view workers’ remittances should certainly not be subject to capital controls, but otherwise I think conceptually they really belong in the capital account, although because they tend to be countercyclical there certainly are good reasons for suggesting that they should be treated differently from other items in the capital account. I would also argue that interest payments and dividends, which are today part of the current account, should be part of the capital account although again, because they are nominally fixed and not subject to changes in investor sentiment, perhaps they don’t fit wholly comfortably in the capital account.
Finally when it comes to defining the balance of payments components I don’t think all commodity imports should be treated equally. Commodities that are imported for use or for working inventory should certainly show up in the current account, as they do. Commodities that are imported for speculative purposes or for stockpiles, however, should be included in the capital account, since they really are a form of external investment more than a form of domestic consumption.
This is how they would recorded, for example, if rather than import physical commodity for storage a local speculator purchased a commodity-linked note from abroad. There is no real economic distinction between the two transactions, but the former would be treated as a current account import while the latter would be treated, correctly, as a capital account export.
This matters because the numbers can be significant, and so heavily distort the balance of payments numbers. Here, on the subject of cotton, for example, is an article from Bloomberg:
Cotton stockpiles in China, the world’s biggest importer, are set to climb to about 9 million metric tons this season, enough to cover the country’s deficit for the next six years, according to Allenberg Cotton Co.
Inventories are rising as the government boosts purchases to support domestic prices and lift farmer incomes, Joe Nicosia, chief executive officer of world’s largest cotton trader, said at a conference in Hong Kong today. The country may buy 5 million tons for reserves this year, up from 3.2 million tons a year earlier, he said. “As long as China maintains this regime to subsidize cotton farmers, the world will be prone to overproduction,” he said. “Can you imagine a world without China importing any cotton for six years? They hold all the cards.”
…Global stockpiles may total 79.11 million bales on July 31, up 14 percent from a year earlier, the U.S. Department of Agriculture said Oct. 11. China will import 11 million bales, down 55 percent from last year, as its inventories climb 21 percent to 36.61 million bales, the agency said.
When you have stockpiled enough cotton to cover the next six years of imports, it seems to me, most of your stockpile represents a speculative bet on cotton prices. It should be treated no differently than any other speculative bet, and the fact that it warehoused domestically rather than off-shore is largely irrelevant.
It is not just cotton, of course, for which large speculative positions distort the balance of payments numbers. FT Alphaville quotes a Goldman report on copper:
Over-importing owing in large part to ‘financing deals’: Chinese refined copper net imports rose by 73% yoy ytd September, averaging 265ktpm, moving well above average requirements of c.225kt per month. As such, Chinese copper inventories have risen sharply this year (primarily at bonded warehouses), despite a negative Chinese copper import arbitrage (Exhibit 2). Since the import arbitrage is not attracting extra metal (over and above requirements), financing deals, which depend on the positive differential between domestic and foreign interest rates, are the likely culprit (in line with anecdotal evidence).
This, of course, is an old story, and it is not hard to figure out what the consequences of this kind of thing are likely to be. Here is an article from China Daily:
China’s steel industry is a big cause for concern in the fourth quarter due to shrinking demand and heavy losses, according to an industry official. The fears were outlined by Huang Libin, an official from the Ministry of Industry and Information Technology, in an interview with China National Radio.
“The steel sector’s performance has been bad since the beginning of the year,” Huang said. “Their revenues are falling and demand remains weak.” The entire steel sector is now operating at a loss and struggling with problems of oversupply and a broader economic slowdown, he said. MIIT data show that 45 percent of the country’s steel companies suffered losses in the first nine months of 2012.
Clearly there has been too much stockpiling of a wide range of commodities, and just as I have warned for many years, Chinese stockpiling of commodities is a very dangerous balance sheet management given the positive relationship between Chinese growth and commodity prices. It was just a matter of time before a slowdown in Chinese growth would cause a collapse in commodity prices, saddling already-struggling Chinese producers with soaring inventory losses. This seems already to be happening, and of course there is a lot more to come.
Understanding the Balance of Payments
But balance sheet management implications aside, the point is that our understanding of the risks and conditions in the Chinese economy would be significantly enhanced by distinguishing between commodity imports as a current account import item and commodity imports as a capital account export item. The problem with this kind of differential treatment of commodity imports, of course, is that it might simply encourage countries to lie even more about commodity purchases in order to manipulate the numbers. What’s more, if after some period of time a local speculator sells to a local user, the transaction would have to be treated simultaneously as a capital account import balance by a current account import, which might be hard to do as a practical matter.
We are probably stuck with this very distorted way of recording commodity purchases in the balance of payments, and there is not much we can do to change it. It does suggest however that rather than accept the commonly accepted definitions of the balance of payments – or of other things, like GDP, and whether housing must be classified as consumption or as investment – as if they were fundamentally meaningful, we should constantly remind ourselves why exactly we need the information and then adjust the numbers accordingly.
In the case of China, this means that to the extent there has been an increase in commodity imports held for speculative or investment purposes, we should reduce imports and increase the current account surplus correspondingly, in our private calculations. We should also increase our estimates of capital account exports.
The other big adjustment to the current account should reflect the extent to which speculative inflows or flight capital is being disguised as goods and services imports or exports. On Saturday Chen Long in my central bank seminar pointed out that since there is substantial evidence that China was a net recipient of hot money until roughly around the third quarter of 2011, and since then has become a new exporter of hot money, we should consider that China’s current account surpluses were probably lower than reported before the third quarter or 2011, and probably higher than reported since then. This makes sense to me.
The other implication he pointed out is that while the PBoC has been recycling the current account surplus, there are two major adjustments over time that we should consider. Before the third quarter of 2011, China’s recycling of capital included a shift from private sector recycling to PBoC recycling – in other words Chinese institutions were on average liquidating foreign assets (or borrowing abroad) while the PBoC simultaneous accumulated more foreign assets (mostly dollar and euro government bonds) than can be explained simply by the recycling of the current account surplus.
Since the third quarter of 2011, however, there has been a shift in the way China as a whole recycles the current account surplus. The PBoC is liquidating reserves (relative to the current account surplus at first but now in nominal terms too) while other Chinese institutions are acquiring foreign assets.
One way to think of it is that in the past China shifted out of whatever foreign assets Chinese owned abroad (or borrowed privately) into US and European government bonds. Now China is shifting out of US and European government bonds into whatever assets wealthy Chinese are acquiring abroad as they flee the country. The numbers are pretty big, especially if the current account surplus is understated, which it almost certainly is, and so the effect of these various shifts should show up in relative pricing.
What are Chinese currently buying? They are buying homes and real estate in a number of countries, especially Australia, Canada, the United States, and, to the extent that they can get around newly imposed restrictions, Singapore and Hong Kong. They are also buying commodities and commodity-related companies. They seem also to buying a lot of unrelated businesses in places like Australia, which is good for Australian asset prices but perhaps bad for Australian manufacturers.
This has at least one implication. Real estate and commodity prices have been dropping, but this has come in spite of a massive program by Chinese effectively to swap out of US and European government bonds and into commodities and real estate. Where would prices have been absent this Chinese swap? Probably much lower, right?
So what will happen next? The demand for real estate may or may not abate at some point in the future, given the size of Chinese demand to hold assets in a safe place – a demand which is not likely to drop with slower Chinese growth but rather to speed up. The demand for commodities, however, will certainly do so once Chinese long positions, combined with much slower growth, make them excessive.
This can’t be positive for commodity prices. My point more generally is that growth in China is likely to be negatively correlated with Chinese demand for foreign real estate and positively correlated with Chinese demand for commodities. It will also affect other things for which China has effectively been swapping US and European government bonds, after many years of doing the opposite.
By the way, and as a humorous aside, there has also seemed to be a noticeable China-based shift in recent years from US and European government bonds into outrageously expensive Swiss watches (the blingier and more ostentatious, the better, especially if it comes in a large box and with a price tag that is difficult to remove). This accumulation of expensive and loud watches however has become such a scandal in China that perhaps it is time to look for the next big thing in the way of ostentation.