Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Goldman: PBoC using swaps to limit reserve growth

by Brad Setser
February 18, 2007

Goldman seems to have confirmed something that I — and, for that matter, Stephen Green of Standard Chartered and Louis Kuijs of the World Bank — have suspected for a while: China's "true" reserve growth is higher than the reported number ($247b).

Goldman via Reuters and Pakistan's Daily Times:

To shield the banks from currency risk at a time of a rising yuan, the People’s Bank of China (PBOC) may have arranged a total of $40-50 billion in currency swaps with Bank of China, China Construction Bank Corp and Industrial and Commercial Bank of China at an implied yuan appreciation rate of 2-3 percent a year, the bank said. Goldman’s research is the latest attempt by PBOC-watchers to explain an apparent paradox: given China’s current account surplus last year of about $230-250 billion and foreign direct investment inflows of $63 billion, why did the PBOC’s official reserves rise in 2006 by only $247 billion? 

The gap between China's reported reserves and the sum of its current account surplus and net FDI inflows is even bigger if you strip out estimated valuation gains from the reported $247b increase.  China holds some euros, and those euros rose in value in 2006.  If China has about 70% of its assets in dollars and say 20% in euros, 5% in pounds and 5% in yen, valuation-adjusted reserve growth was more like $220-225b.  Even with $40-50b in swaps, the 2006 increase in valuation adjusted reserves ($220-225b), off-balance sheet reserve growth ($40-50b) and transfers to a large reinsurer ($4b) add up to between $265-280b.  That still seems to be a bit less than the sum of the current account surplus ($230-250b) and net FDI inflows ($60-65b).   There may be a few more dollars out there in someone's hands …

The use of swaps also explains why the banks have been willing to buy dollar bonds when private Chinese citizens's haven't shown comparable interest in foreign assets: they aren't bearing the currency risk.   Goldman again:

Liang and Yi said it was hard to argue that banks chose to buy FX bonds on commercial grounds because they have limited FX operations and, since August 2006 when the yuan’s pace of climb started to quicken, the trade has been losing money. Furthermore, the banks’ decision to add aggressively to their FX assets contrasts with a steady drop in Chinese households’ holdings of FX assets. Just this week, Bank of China said it had closed one of the first funds through which retail investors could buy overseas bonds because of heavy redemptions.

Lang and Yi note that the total foreign assets of China's banks increased by around $70b in 2006.  $40-50b of that was financed by swaps — i.e. it effectively was borrowed from the PBoC.  Most of the remainder, I would guess, comes from the roughly $25b ICBC and the BoC raised in their Hong Kong IPOs.  Goldman should know that data especially well.  To the best of my knowledge, the PBoC hasn't allowed China's state banks to convert their Hong Kong dollars into RMB

Kudos to Goldman's China team; this looks to be good work.


  • Posted by Emmanuel

    Madness, absolute madness. Financial independence is a myth in China. This take on the health of Chinese, er, “SOBs” (state-owned banks) might be extreme, but there is truth in here somewhere. I guess these “SOBs” should be thankful that they can at least buy FX assets yielding better than sterilization bonds, but still. It boggles the mind what they’re doing there to keep a lid on revaluation.

    Message to the Communist Party: Revalue for you can’t fight fate.

    Gleaming new corporate headquarters notwithstanding, don’t underestimate the backwardness of China’s banks. The Bank of China does not have a centralized computer system to keep track of its client accounts. Corruption is ubiquitous. Its prospectus outlined fraud amounting to $737 million. Last April, it fired 75 bank officials for corruption and one former branch head was even given a (suspended) death sentence. And if you believe that the fish rots from its head, consider that the CEO of each Chinese megabank is a Communist Party crony.

  • Posted by Guest

    “…Despite accepting that corporate governance has improved immensely in China in the past few years, the country still ranked almost last in Asia. Only Indonesia scored lower…”

  • Posted by bsetser

    SOBs are usually acronymized as State commercial banks (SCBs), though at least one (ABC) isn’t yet very commercial. the others do have rather substantial market caps …

    the basic deal seems to have been in return for the state taking old bad loans off the books and providing a nice spread on new lending (the gap between the cap on deposits and the floor on new lending), the SCBs cannot lend as much as they want and have to use a growing fraction of their balance sheet to buy various sterilization products, whether bills or fx swaps. the swaps both take RMB out of circulation (the SCBs give the PBoC RMB for dollars) and let the banks invest the dollars (hopefully getting some experience in the process) without worrying about the XR risk …

  • Posted by HK

    Brad–This is very, very bad for China.

    I recall that Keynes recommended the British Treasury to intervene in the foward exchange market rather than in the spot exchange market in order to defend the pound. Thailand, in 1997, heavily intervened in the foward market in order to defend the Baht without apparent foreign excahnge reserve loss (at least for the time being).

    China is doing similar things in a different context, but the intention apprears to be the same; defending the excahnge rate without apparently increasing foreign exchange reserves. This does not help them resolve the fundamental problem at all.

  • Posted by Aaron Krowne

    This is desperate — all the hallmarks of the usual official-level scrambling before an international financial regime falls apart inelegantly. Brace for impact…

  • Posted by Karl

    Rodrigo de Rato, Managing Director, International Monetary Fund
    said in septembre 2006
    “Multilateral Consultation focuses on narrowing global payments imbalances while maintaining robust global growth. Fund staff have already held bilateral discussions with the participants in the consultation—China, the euro area, Japan, Saudi Arabia, and the United States, and over the coming months we will hold roundtable meetings”

    i doubt he was thinking about such kind of solution ..

  • Posted by Guest

    “…When China’s markets reopen on February 26, the bubble is likely to inflate further… and Beijing will have to deal with a bursting of the Chinese stock markets. If the bubble pops before the 17th National Congress scheduled for this fall, it could weaken President Hu Jintao’s ability to appoint the successor government he would like. If the burst does not come before the fall, it will likely be a dramatic plunge and could wipe out the savings of many within China’s new middle class…”

    “…Investors are losing some of their enthusiasm for China and India, the world’s two fastest-growing major economies, a Merrill Lynch & Co. survey showed. More money managers planned to cut stakes in Chinese companies on a net basis for the first time in at least four months… according to the Feb. 14 survey…”

  • Posted by bsetser

    Selling RMB forward (buying the $ lent to the banks back) isn’t quite like selling dollars you may not have forward … the risks are different. Tis the same point that is made with real cash: there is no upper limit on how many reserves you can accumulate, so long as you can sterilize them …

    That said, it looks like China’s 06 reserve growth was a lot closer to $300b than $200b, even with a very controlled pace of appreciation for most of the year and int. rates well below US rates. China may have detered “speculators” but it only encouraged its exporters …

    the real question is can China continue to increase the pace at which at adds to its foreign assets without something really breaking — i still think total foreign asset accumulation in 07 (banks, SAFE, state investment company) will be closer to $400b than $300b. setting apart whether it can do this without creating problems elsewhere, there is the issue of whether it should have so much of its national wealth tied up in assets that should depreciate over time (dollars and euros).

  • Posted by Dave Chiang

    From Canada Global Research, the end of US Dollar hegemony for the United States

    ” The world has changed, especially with the rise of the huge Asian economies of China and India. We must now search for the principles and mechanisms that can work in a world no longer dominated by the Western powers where domestic production is stagnant, and where financial bubbles distort measures of real value. China is also using its US dollars to make loans to African nations and is gaining economic leverage on that continent as well as in Latin America. It appears that US Dollar hegemony has finally backfired and is starting to undermine national security. This is one reason observers assume the US military views China as a potential adversary. “

  • Posted by Joseph Wang

    The Vardy article is total non-sense in that it confuses numbers left and right. $2 trillion is roughly the total lending of the big four banks, and so Vardy is assuming that all of the loans that the banks are issuing will go bad.

    Corruption is not a major reason why the Chinese banks had problems in the first place. $1 billion is a drop in the bucket compared to the total NPL problem.

    The basic reason for the NPL problem was that the banks were asked to finance unemployment and social welfare benefits, as unprofitable state-owned enterprises were closed. They are now closed, and the SOE’s are profitable.

  • Posted by Joseph Wang

    The Shanghai stock market has crashed before. It will crash again. It’s not a big deal.

  • Posted by Dave Chiang

    Businessweek’s Credibility Takes Hit With China “Bust” Article

    What they call an upcoming “bust” for China is called a “market correction” at home. “Bust” sells magazines. Like most of the US mainstream media, Businessweek is very pro-India and rabidly anti-Chinese. Since there is a good pool of talented English writers and the labor is relatively cheap there compared to the West, Businessweek has also outsourced a lot of jobs to India since year 2000. – Dave C.

  • Posted by Guest

    “…In a paper released in recent weeks Chicago based economist David Hale says he expects a revaluation that will surprise on the high side, “with some movement perhaps around the time of the Chinese New Year in mid-February, or the 17th CPC Congress…”

    As compared with his Spring 2003 view:

  • Posted by Guest

    “”Bust” sells magazines”

    maybe to you – along with headlines like “end of US Dollar hegemony for the United States”

  • Posted by Guest

    “…”The industry suffers institutional loopholes in preventing and combating corruption, especially the collusion between government officials and businesses,” said Wang Guangtao at a recent conference on building a clean and honest government…”

    “…Siwei and others warn that up to 70 percent of the listed companies on the Chinese exchanges are worthless…”

  • Posted by moldbug

    jwang: profit and loss can only be calculated in monetary terms. And the number of new RMB born every minute is enough to turn a lot of losses into profits.

    Unfortunately in an open-loop financial system it is not possible to calculate or even define the rate of monetary dilution. But if you accept Chinese official figures for money supply growth – what is it, 15%? – and redo those balance sheets in fully diluted yuan, the East looks pretty red all of a sudden.

  • Posted by Gcs

    brad writes:

    “there is the issue of whether it should have so much of its national wealth tied up in assets that should depreciate over time ”

    this accumulation is not result of a store of wealth strategy
    but a forex strategy

    this is a cost of doing international trade

    excess reserves in a pegged to currency grow as a passive result of success at maintaining an export enhancing
    low ball pegged forex rate
    higher reserve growth largwer reserve size greater exposure to deval risk are all positive signs (if negative consequences) of this higher then otherwise growth of exports

    fair is foul and foul is fair McBrad

  • Posted by bsetser

    GCS — as you know, tis quite possible for trade (X+M) to grow even in the absence of official asset accumulation; China just has put itself on a trajectory where unbalanced trade growth implies ongoing asset accumulation … tis a policy choice, not a simple by product of the growth of trade.

    though at this stage changing course looks like a mighty tricky job — far harder than in 04.

  • Posted by Gcs

    trade trade off

    big risky deval prone reserve build up

    but its the bestr way for unbalanced export heavy trade pathing

    far faster export trade growth

    plus slower import trade growth:

    domestic protection and export incentive both

    without “intentional”
    bias toward one sector over another
    and not obviously anti fair trade the way tariffs and quotas would be

    and not aimed at any one nation either

    what’s not to like ???

    if your behind the tech fronteer
    then its the only sure route to world class technique to boot

    too good to give up without a serious outside push

    only the victims united
    can stop the scam


    in this case

    the us job class and soon the euro job class too
    forcing their gubmints to act

  • Posted by Joseph Wang

    If 1998, the banks swapped about $400 billion in bad debt for bonds in the asset management companies. These bonds were to pay down the debt over ten years at which point the AMC’s would be closed. That gives you $40 billion/year that the PRC government promised to pay the banks, and they are tapping the foreign reserves to pay off the last installments.

  • Posted by Joseph Wang

    moldbug: Money supply has outpaced GDP growth every year since the start of reforms in 1978.

    What appears to be the case is that as time passes, the money velocity is increasing.

  • Posted by moldbug


    None of the variables you mention – money supply, GDP growth, or velocity – is meaningful. Except for velocity, they all rely on aggregates which are founded on subjective impressions, ie, fudge factors. Velocity at least can be precisely defined, but it is difficult to measure, and its only conceivable relevance is in the context of MV = PQ, in which P and Q are meaningless aggregates. M is well-defined in a hard-money economy, but not under central banking. GDP growth depends on the GDP deflator, which is pure fudge.

    The example I like to use is comparing a 2007 Mustang to a 1967 Mustang. I assert that any number you can produce that purports to define how much better (or worse) the 2007 Mustang is than the 1967 Mustang is a fudge factor. In fact I can think of no better definition of “fudge factor.” But you can head over to the BLS and have them tell you all about index chaining, and who knows – they may even have this particular number. I mean, it’s compiled by a nonpartisan federal agency, so it must be true.

    My point was just that a lot of new yuan are being created in China. (I’m sure there is an explanation of the difference between yuan and renminbi somewhere on the Internet, but Google has not been helpful in connecting me with it. If anyone can explain this point I would be curious.) Many of these new yuan are no doubt ending up on the balance sheets of Chinese enterprises, making them look more profitable than they otherwise would. If China were to stop creating new yuan, many of these enterprises might seem less valuable.

    This is an old Mises point which strikes me as relatively uncontroversial. I mean, it’s not like I’m getting all Louis McFadden on the PBOC. (Which is presumably not backing any of its notes with Scotch receivables or human hair – although it is interesting to note the parallels between US:UK:1929 and China:US:2007.)

  • Posted by Joseph Wang

    The wealth available to China’s economy is growing, therefore the paper money which represents that economy needs to increase at the same rate. The question is whether “real wealth” is being created or not, and that is something that you need to go into the factories to find out. If real wealth is being created and money is ending up in the balance sheets, this is a good thing.

    There are a lot of different explanations for 1929, but here I subscribe to the ideas of the monetarists rather than the Austrians. The reason for this is that there have been a number of incidents that *could* have turned into a depression, but didn’t (1973, 1993, 1998, 1987, 2001). In these cases, the powers that be followed a policy prescriptions that was the direct opposite what what von Mises would do, and in those cases, it seems to have worked. The one case in which there was a depression (Japan-1990’s), the central bank followed a policy which monetarists would argue would cause disaster and it did.

  • Posted by moldbug


    The proposition expressed in your first sentence is certainly common. However, I have never seen anyone attempt to justify it from first principles. It corresponds to a kind of magical thinking which is cognitively very accessible, and it is politically convenient for a wide variety of actors. These factors strike me as sufficient to explain its popularity.

    Money does not “represent” an economy. It is simply a good that is exchanged for others. If your money has no intrinsic value, the monetary authority can dilute it with perfect symmetry by simply redenominating it. Asymmetrical dilution is therefore equivalent to redistribution of savings.

    Redistribution can certainly have macroeconomic effects. And it is by no means harmful by definition. If you confiscated all but the last million from the world’s 1000 richest people and distributed it evenly among the citizens of, say, Mali, the result could certainly be described as a boom, and it would be hard to argue that you had made the world worse. However, what this has to do with “growth” is by no means clear.

    As for “real” wealth, again, there is no way to define it other than a price system. And prices are momentary, not persistent. Last year’s prices for last year’s goods cannot be systematically related to this year’s prices for this year’s goods.

    If I can pack Keynesians and monetarists into one large, wriggling bag for a moment, and refer to them both as “modelers,” the way modelers think about macroeconomics is very similar to the way that pre-20C doctors thought about medicine. They had a wide number of procedures whose outcomes they could observe. However, they had no way to conduct controlled experiments on their patients, and their ideas of cause and effect were (with the benefit of hindsight) entirely fallacious. In retrospect, some of their procedures worked and others were completely pointless.

    Austrian economics is never going to be competitive on the witch-doctor market. It is not particularly useful for civil servants. As a civil servant, you want to have numbers and formulas to back up your PowerPoints. You want to convey the impression that any attempt on the part of reckless politicians to quarrel with the policies that your models dictate will only sink them deeper into a sea of science from which there is no escape. Both Keynesian and monetarist macroeconomics are perfectly designed for this role.

    But Austrian logic cannot be demonstrated or disproved by any empirical result. The Austrian edifice is large and I have no doubt that it contains its share of fallacies. But if you want to find them, you have to dust off your brain and crawl into the ducts.

  • Posted by RebelEconomist

    To answer Moldbug’s question (since this one has occurred to me before):

    The RMB or ren min bi is the generic name for China’s currency – apparently it means “people’s currency” – whereas the yuan is the largest unit of that currency. Like sterling and pounds.

  • Posted by moldbug

    Thanks, RE…

  • Posted by Joseph Wang

    Money is a measuring instrument, but what it measures exists independently of the instrument. Wealth can only be measured in the context of a market system, but it exists independent of that system. I write some words on a blank page. I’ve created wealth. How much wealth is measured by the market pricing system, but the wealth exists independent of the system.

    Austrian economics can be empirically demonstrated to be incorrect. You look at societies based on markets and societies not based on markets, the former work better.

    If you don’t use empirical facts as the basis for truth, then how do you know if you are right?

  • Posted by moldbug


    Your questions are excellent. Indeed there is a reason Mises spent the first 100 pages of Human Action on epistemology.

    There is actually no conflict at all between Misesian and Popperian epistemologies. The Popperian is simply a special case of the Misesian. Mises will, if you give him a chance, make a good case that there is such a thing as logic, and that we cannot refuse to believe in its results. Popper presents a logical explanation of why one very elegantly defined inductive method is such a predictable producer of truth. But if you don’t believe in logic, you have no reason to believe in Popper.

    Both Mises and Popper would, I believe, agree that in the words “wealth” and “measuring instrument” you have constructed concepts that cannot be defined or questioned. I cannot think of any criticism which applies to the labor theory of value from which your concept of wealth strikes me as exempt, and perhaps the two are simply identical.

    As Albert Jay Nock put it (in Memoirs of a Superfluous Man):

    In those days I noticed with amusement that some philoso-
    phers of “the social consciousness” had carried their speculations
    up into the higher realm of scholastic metaphysics. I could make
    nothing of Sir Leslie Stephen’s notion of a “social organism” but
    that society exists as an objective entity apart from the indi-
    viduals who make it up. I had long known that the Church lays
    claim to that sort of unsubstantial existence, but I was never
    metaphysician enough to get a very clear idea of how this could
    be so. I should say that if a Church or “the Church” no longer
    had any members it would no longer have any existence; and
    I should say the same of society. Albertus Magnus and his great
    pupil Thomas would sniff at Sir Leslie Stephen’s “social organ-
    ism” and the curious product, apparently ectoplasmic, which
    he calls a “social tissue,” and they would at once catch the fine
    old familiar fusty aroma of universals existing objectively.

  • Posted by arkadaş arıyorum

    If you don’t use empirical facts as the basis for truth, then how do you know if you are right?

  • Posted by alsat

    thank you this sharing post

  • Posted by iyinet

    thanks a lot…