Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Richard McGregor on China’s (huge) reserves

by Brad Setser
September 25, 2006

The one trillion dollar mark is a perfect hook.  Richard McGregor's excellent story in today's FT is the first of no doubt many stories on China’s phenomenal stockpile of reserves.

McGregor highlights my argument that the composition of China’s reserves is a second order issue, at least from China’s point of view.  Shifting reserves from dollars to euros won’t prevent China’s central bank (the PBoC) from taking losses if the RMB appreciates against both the euro and the dollar.  And, as my regular readers know, the RMB sure looks undervalued against the euro.   The size of China’s capital losses will be primarily a function of the size of China’s reserves – not the composition of China’s reserves. 

But that is not the reason to read McGregor.  I don’t have any influence over Chinese policy.   But a lot of the other folks that McGregor quotes do.  He provides a great window into the internal Chinese debate, and that is really the only debate that matters.

What struck me? 

The PBoC’s worries that it will be blamed for taking big losses on its foreign exchange portfolio, even though those losses were basically “baked in” — to use Dr. Swagel's phrase — at the moment China bought its current holdings of dollar bonds at an inflated price.

The PBoC’s concerns are understandable.   The last Chinese reserve manager who took large losses (by betting on the euro a bit too quickly) took those losses hard.  And the Bank of Korea has come under pressure for the capital losses (more here) it has taken as the won has risen against the dollar.

At the same time, so long as the state council instructs the PBoC to resist pressure for the RMB to appreciate against the dollar, losses are unavoidable. 

China Foreign Exchange’s Mr Zhong is right:

“We cannot blame the US Treasury [for future losses]” he says. “No one forced us to buy dollars.”

I think McGregor’s reporting on China’s internal debate– which really is worth reading closely — raises two important issues. 

First, China’s large financial stake in the US economy could be a source of future tension, not a source of shared gains that lubricates the overall US-Chinese relationship.

The argument that China’s huge holdings of US debt (which now dwarf US FDI in China) give it a stake in America’s success is a bit off.   It is true, but it misses a key point.  The interests of a creditor are not always the same as the interest of a debtor.  China (at least its central bank) would rather see the US take policy actions that minimize the scale of China’s capital losses by minimizing the needed adjustment in the RMB/ dollar.    The US, on the other hand, has traditionally taken the view that the dollar is our currency but your problem — it has never promised to direct its policy to maintaining the dollar’s external value.  And it certainly has never promised China that if China keeps financing the US, the US will look after China’s financial interests. 

That could be a point of future tension, particularly if the PBoC starts taking capital losses and feeling the heat.

Second, I think DeLong may be underestimating the costs of China’s current policy.  He argues that China is basically taking a one-off 3% of GDP capital loss (on its 10% of GDP annual reserve accumulation) in exchange for a permanent increase its exports and its income.  I would argue that sustaining China’s current level of exports requires ongoing financial flows from China to the US.    If Chinese reserve growth slows and China stops financing the US, the US wouldn’t be able to afford its current imports from China.   Consequently, China has to take an annual capital loss to sustain its current (inflated) level of exports. 

China doesn’t just have a $1 trillion in reserves.  It has put itself on a course that requires adding at least another $1 trillion to its reserves over the next four years.  I actually suspect the pace of Chinese reserve accumulation might pick up to around $300b a year, which would imply Chinese reserve would hit $2 trillion in 2009.   

Basically, I suspect that China will need to keep financing the US (on subsidized terms that imply losses for the PBoC) even after the pace of Chinese export growth slows – just to sustain its already high level of exports. 

That raises another point where I think I differ from DeLong.  DeLong argues that China’s policy of using the balance sheet of its central bank to subsidize its exports has generated a permanent increase in China’s income.  I would argue that it more likely has pulled China’s future export growth (and investment growth) forward – very, very rapid export growth now will likely be offset by far slower export growth in the future.   And very, very rapid investment growth now likely will be offset by far slower investment growth in the future.  That means a higher level of income now, but not necessarily a higher level of income in 10 or 20 years.    

Given the size of the PBoC’s balance sheet — $1 trillion is a bit under 40% of China’s GDP – and the impact China has had on the world economy, these questions obviously matter, for both China and the world.


  • Posted by camille roy

    What would be the consequences of China choosing to use say 25% of its reserves to fund health care, unemployment insurance, and a social security system?

    Would this allow China to revalue the currency by lessening the political consequences of a slump in export led growth?


  • Posted by Guest

    there’s a plan to do that with SOE dividends…

    “The World Bank notes that a 50 per cent pay-out would have allowed the government to increase health and education spending by 85 per cent in 2004. Better welfare provision might give Chinese savers an incentive to switch into riskier assets like equities. For the SOEs it would lead to more disciplined investment, thereby boosting returns.”

  • Posted by Butch

    China will soon join Japan in the “Dollar Hegemony Victims Anonymous” club.

    The whole process reads like a script:

    1. China bases its economy on exporting to the U.S.

    2. The U.S. (The “Mouth of the World”)accepts Chinese real goods and offers “Gladly Pay You Tuesday” bonds denominated in U.S. “Gladly Guarantee You Nothing” dollars.

    3. China, also practicing terminal fractional reserve lending and central banking (piled on top of their reserves of “hard currency”, ha, ho, ha, ho, ho, don’t make me laugh!!! “Hard currency”!!! Man, if THAT isn’t an Orwellian oxymoron, I don’t know what is!), over-expands, mal-invests in plant, equipment, real estate projects–you name it, only to create it’s OWN boom/bust cycle.

    4. Sooner or later (but probably sooner) China suffers a Great Depression/Japanese Bust of its own. Hundreds of Billions of Yaun/Rmb loans are defaulted, tens of millions of people are thrown out of work, real estate crashes–you know the drill.

    5. Meanwhile, the U.S., if it also doesn’t fall into a deflationary abyss of its own, skips merrily down the road, finds another sucker, er “emerging market” to begin the cycle all over again. However, the game is getting more dangerous for the U.S. as:

    a. The U.S. consumer is up to his/her eyeballs in debt, much of it collateralized by falling home prices.
    b. The “Rest of World” that currently holds a net $4 trillion in U.S. debt just might be getting nervous.

    Now, having been screaming for years that “This can’t go on!”, only to see the whole, bizarre Ponzi scheme find yet another way to perpetuate itself (including the $400 trillion in derivatives bets which I suspect have a nefarious background), I am cautious about stating unequivocally that “This is it.”. However, I will state for the record that at some point the U.S. will reach debt saturation or the “R.O.W.” will demand higher interest rates, or stop buying U.S. treasuries/agencies, or God-forbid, actually SELL said securities!!!

  • Posted by bsetser

    China cannot directly use its reserves to finance a safety net.

    It could borrow funds domestically to finance a safety net (i..e increase the government’s fiscal deficit) or finance a safety net out of dividends (reducing business savings). the net effect should be lower national savings — which, assuming everything else is unchanged — would lead to a smaller current account surplus/ less need to continue to add to China’s reserves.

    Olivier Blanchard’s paper is great on this topic — i recommend it.

    Butch. I thought so too. At least worried about the risk. About two years ago. I am still waiting for the saturation point though.

  • Posted by Emmanuel

    The FT article is thought-provoking –

    (a) If the Chinese government actually buys the Stiglitzian theory that accepting American pressure to adjust will lead to China becoming like Japan in the 1990s, there will be strong resistance indeed. If this is the case, join me in the Reserve Pumpkin Patch to watch for the Great Revaluation Pumpkin;

    (b) I was going to be Mr. Wiseguy about currency composition mattering less than total accumulation by suggesting that the PBoC buy more JGBs, but then it would be a tradeoff between accepting minuscule yields while waiting for the $ shoe to drop;

    (c) Er, what sorts of “risks” are they talking about? China isn’t going to endure a BOP problem or a speculative attack anytime soon. Also, a local banking crisis we’ve learned from RGE cannot be remedied with greenbacks. From the FT article:

    The sheer size of the reserves has been the trigger for the new debate about how else the money might be deployed. According to some local analysts, Mr Zhou’s crisp comment – that China has “enough” reserves – is better read as one of a number of signals coming out of the central government that Beijing has settled on an amount it needs to set aside as reserves in the traditional sense, as a national insurance fund against financial risk.

    Xia Bin, an economist at the Development Research Council, a think-tank under the State Council, the cabinet, has suggested China needs about $700bn in foreign reserves to this end.

    (d) Why does DeLong think that this export-driven reserve accumulation pattern adds to China’s GDP permanently? A global recession is unlikely to guarantee growth rates of even 5% as he assumes, and has a good chance of throwing China into reverse gear.

    As an aside, have you noticed the 10-year Treasury bill is now yielding just 4.55%? It’s still party time at China’s expense. To paraphrase Mr. Spock, live large and prosper. There’s this bare wall at my place that would be nicely filled by a 100″ plasma screen…

  • Posted by algernon

    China’s subsidization of it’s exports also over-expands its production capacity beyond what a ‘free market’ would support.

    Not only are the Chinese stealing from their future, they are putting themselves at greater risk of recession when the consumers of those exports decide they can’t borrow anymore, even at low interest rates.

  • Posted by bsetser

    yep, I noticed that the conundrum is back and bill gross is a happy man … lots of folks who were holding out for the peak Treasury rates apparently piled in in august, and well, since then the data suggests a us slump.

    emmanuel — good points.

  • Posted by Joseph Wang

    Actually, I don’t think that the Financial Times does a good job of covering the Chinese debate. The trouble is not McGregor’s fault but rather the fact that you can’t really do a good summary of the issues in a newspaper article.

    Anytime an article reduces the argument of a major policy question to a debate between only two sides, it’s probably very much oversimplified what is actually going on.

  • Posted by Joseph Wang

    Re: health and education. It’s not a matter of money since there is more than enough money to flow around, but rather that putting together a health and education system for a 1 billion people doesn’t happen overnight.

    It takes about two to five years, and right now with respect to health there is a lot of local experimentation on what to do.

    With respect to education, the central government is picking up the tab -> less pressure on local government finances -> fewer land sales -> fewer riots. This was after a few years of local experimentation to see what systems work and what ones don’t.

  • Posted by Gcs

    brad your faster ex grow now
    slower ex grow later i think makes sense in fact it must make sense
    because surely there’s some steady state share of world trade china will eventually settle in to like japan has
    (in fact its an over shooter and
    is settling back some in share )

    but overall internal demand growth
    should accelerate and comp the export slow down

    and any way
    what’s so wrong with a pull forward if its not comped by a lower level
    just a slower rate in the future

    only if the some how crash the system
    and cause a freak outprotectionism wave to hit the north world economy
    which is a varient
    on the recent stiglitz scenario

    when joe talks about
    the 1914 to 1939
    international market slump
    he’s saying
    nothing sez this thing has no reverse gear here


    china is after
    the fastest possible tech transfer makes perfect sense
    and fastest possible build up of human capital

    which needs a serious trade potmetial
    to aquire from the trans nats that “own “the hi tech

  • Posted by Guest

    What macroeconomic model you have in mind? You write:

    “I would argue that it more likely has pulled China’s future export growth (and investment growth) forward – very, very rapid export growth now will likely be offset by far slower export growth in the future. And very, very rapid investment growth now likely will be offset by far slower investment growth in the future. That means a higher level of income now, but not necessarily a higher level of income in 10 or 20 years”.

    You seem to believe (not only in this blog) that
    Y = C + I + (X-M)
    implies that as (X-M) falls, Y falls alike, missing the correlation that runs from a fall in net export to a boost in domestic expenditure.

  • Posted by quiz

    Pretty interesting dilemna. The only place the mercantilists can safely spend their dollars is the US.

    I think the Chinese have already discounted the value of their dollar holdings by 25 – 30 %. They figure it’s worth it to get the juice for their export only economy. What happens when the US cannot absorb the dollars?

    They might try a few more high profile takeovers. Communist government buys America. Sheesh.

    Is the stable disequilibria about to destabilize ?

    Thanks for sharing your extraordinary knowledge.

  • Posted by DOR

    My favorite comparison these days is between the relative historic values of current US T-bills and 1913 Chinese railway bonds.

    You get what you pay for.

    * * *

    Madam Wu has a very important bit of personal history. In a nutshell, she was responsible for selecting among a range of policy options, got it wrong, and watched the horror unfold. Perhaps overly dramatic, but definitely worth considering.

    She was plucked from near-obscurity in the oil industry to become Vice Mayor of Beijing in 1988. Her job was the municipal economy, which (if we believe the numbers) had grown an average of 12% for five years, and in 1988 was accelerating.

    One of her means of attacking inflation (over 20% in late 1988) was to slash student subsidies. That was one of the triggers of the Tiananmen Square protests and massacre six months later (but, by no means the only one).

    * * *

    What jumped out at me in the McGregor articles was:

    * ” Mr Paulson did discuss the issue of China’s currency valuation directly with Mr Hu [Jintao] . . .”

    * ” The Chinese account of the meeting made no mention of the currency . . .”

    * “At the very least, the US statement [about Wu Yi having ‘full decision-making authority across all aspects of the Chinese economy’], which was not replicated in the local Chinese media, may give some clues to shifting responsibilities at the top of the Politburo.”

    It reminds me of “The two sides recognize that there is one China and that Taiwan is part of China” [PRC version] and “Chinese on both sides of the Taiwan Strait believe there is one China and that Taiwan is a part of China.” [US version].

    * * *

    And, “In doing so, she [Wu Yi] impressed not just her own side, but US President George W. Bush as well, . . .”

    I’m guessing it doesn’t take much to impress Dubious.

  • Posted by Gheorghius

    Dilemma non existent, quiz, unless you’re yourself a mercantilist…

    The dollar reserves: China’s Gov. will take a loss but (slightly reduced in value) can spend them anyway anyhow anywhere no problem you know… think…: “no problem if you have the green, problem if you don’t have the green…”.

    In exchange they got from exports … not so much demand-led growth (as a mercantilist would believe and you and Brad seem to believe), but technology transfer and other positive externalities (typical of industrialisation). And these know-how gains, Brad, they’re not going to lose them. So China’s is a worthy development strategy (ok, I would have re-valued 8% 18 months ago); and Brad’s argument makes perfect sense… unless you’re a mercantilist!

    And remember: dY = dC + dY + d(X-M),
    NOT: dy = C + I +d(X-M)

    A re-evaluation of the Renmimbi will reduce X-M but boost domestic C and I. Don’t worry, be happy.


  • Posted by Gheorghius

    Oh yes, one more. When you discuss the forthcoming “capital losses” on Chinese dollar-Euro-UkPound-etc. reserves, don’t forget to count the interest earned for years on these reserves…

    Would the reserves have been employed more efficiently (i.e. yielded more income) elsewhere? Investing in Wall Street, the Delhi Stock Exchange, in Chinese infrastructure or health? Maybe. (Maybe not: there are absorbtion constraints in China). But it is typical of fixed income assets to yield low returns in exchange for low risk… which adapted to the Chinese situation, means that many macro, financial, and development risks have been contained or “assured” with their reserve-ballooning strategy. Given the rapid growth of income they had, it does seem to be a balanced, prudent approach.

  • Posted by HK

    Brad–I think China has already been locked in excessive dollar holding; if it started a 7-8% annual reminbi appreciation in 2003, it would have largely eliminated currency undervaluation by now, and it would have only about $400 billion (instead of $1 trillion)reserves. But now, the fear of criticism on huge potential foreign exchange loss (probably in the order of 10% of GDP) prohibits anyone (including PBoC governor Zhou) from strongly arguing for currency appreciation. The end result is further delay of adjustment and reserves accumulation ($2 trillion by 2010). Nothing but a catastrophe would stop this process.

    I suspect that the US would act well before China might stop it, either though market forces (household could not borrow more) or through government policy (protectionism or selective defalt). Anyway, the US economy will suffer, and the Chinese economy would suffer more.

  • Posted by Guest

    once you start talking about selective default as an option, i think you’ve already lost… yes, it’s ‘less bad’, but when retaliation might then take the form of selective nationalization that could quickly escalate into economic warfare or worse… and then you have to wonder what sides other creditors might take… japan and UK would go along, but saudi or russia? i’m not so sure… point being, at what point are there ‘no good options’?

  • Posted by a

    Yep the PBOC is just exhibiting the usual behaviour of a rogue trader, who keeps snowballing a position in order not to take a loss. Unfortunately, its position is so big that when it takes the loss, it probably won’t be pretty – for China or for the U.S. But hey that’s trading.

  • Posted by bsetser

    HK — right you are.

    Gheorghius — dI will likely slow independently of any change in the RMB, and it could well slow sharply. Investment booms often turn into investment busts, and right now, i fear the boom has become so big that there is a big risk of a bust.

    dC is not directly a function of d (X-M), tho i would hope that without the crux of export growth, China would take policy steps to encourage C.

    Lardy has noted in the past that investment slumps have led to consumption slumps, though — should that happen, growth in China would really, really slow.

    All this is to a degree independent of the RMB/ $ — tho i woudl argue that that:

    a) by holding the RMB down and keeping interest rates too low, China has fueled its investment boom …
    b) should China’s economy slow, Chinese trade surplus could easily expand even if export growth slows. And the politics in China will favor continuing to snowball their $ position …

    at the end of the day, China might end up both with no higher growth and a capital loss.

    contrary argument is that China gets the tech transfer from greedy MNCs in exchange for the capital loss.

  • Posted by Gcs

    brad do u have any reason to believe the party would NOT off set export
    demand deccelleration
    or investment growth itself for that matter
    why domestic demand acceleration??

    my concern would be
    a timid budget defiict resopnes

    too small to fully off set
    spontaneous market driven
    slow downs
    in both
    domestic investment
    and export demand

    the japanese bug a boo
    lacks one key parallel

    japan circa 1990 had high employment and state of the art industrial base

    not so china

    plenty of room for 10-12+ exspansion
    of gdp by d means alone

  • Posted by Gcs

    brad i doubt you can scare the party into faster reval
    by telling em their exports
    are growing to fast

    and domestic investment is headed for a wall

    nothing will induce them voluntarily
    to reduce their trade and industrial base exspansion

    if anything they act as if
    they’re getting away with murder here
    time is running out for them
    regardless of how they moderate
    it will end no sooner
    or later
    so grab max now
    “better get as much
    technical know how
    and training
    as we can
    fast as we can …”

    only facts
    ie north land actions
    on asian imports
    will get the message thru
    its past talk time now

    its walk the walk time

  • Posted by Guest

    I disagree on this:

    dC is not directly a function of d (X-M), tho i would hope that without the crux of export growth, China would take policy steps to encourage C.

    As a Current Account moves from black to red, capital flow in and boost botth C and I. This is a market mechanism, independent of the Gov.


  • Posted by HZ

    Faster growth now, slower growth later is better than even growth — assuming total growth to the saturation point is the same. There is time value to many things.

  • Posted by chengooi

    The worry is not “DeLong may be underestimating the costs of China’s current policy”. The real worry is that DeLong’s portrayal of the collective thinking of the Politburo and State Council is right, even if he doesn’t agree with their analysis.

    It is highly possible that the majority of the folks in the Politburo do not engage in such a sophisticated cost-benefit analysis as explained by Brad.

    Or they discount the future benefits very heavily as their legitimacy is rested on keeping the Chinese folks happy for as long as they could (3% extra of GDP definitely help a lot).

  • Posted by bsetser

    chengooi — good points.
    HZ — true, but volatility also may be costly, particularly for the politburo/ state council. fast growth = v. good. but sudden decelerations are often not good for anyone’s political health.

  • Posted by Phil

    “a) by holding the RMB down and keeping interest rates too low, China has fueled its investment boom”
    Think the investment boom started well before they were keeping the currency down, Brad. Didn’t it start in earnest just after the 1997-98 Asian Financial Crisis, when China was by its own admission keeping its currency artifically high against the dollar as its neighbors depreciated?
    Perhaps we are talking chickens and eggs here but don’t think it’s fair to say China’s currency policy spawned the investment/trade boom.

  • Posted by bsetser

    Phil — investment/ GDP has always been high in China. But there was a noticeable acceleration that started in the course of 2002 and really got rolling in 2003 — that acceleration was tied to a very noticeable surge in bank lending that seemed linked to the reversal of hot money flows (from out of China in 98-99 to not out of China or into China in 01-02 to into China in 03) that fueled a noticeable increase in reserve growth. So i do think that it meaningful to talk of a recent investment surge (bubble) tied to something that changed in 02/03 and that has pushed investment as a share of GDP up by almost 10%, to a number close to 50%.

  • Posted by Phil

    Point taken. But on the ground the more important factor to me was the changing pattern of Asian trade, which kicked into gear post crisis. And Chinese bank lending had very little to do with the choice of foreign companies to do end-stage manufacturing in China, which started to snowball in those years.

  • Posted by Gheorghius

    Dear Brad,

    I’m a bit shocked by your statement that:

    dy = C + I +d(X-M)

    or as you put it,

    “dC is not directly a function of d(X-M), tho i would hope that without the crux of export growth, China would take policy steps to encourage C”.

    Really you deny the existence of a market adjustment mechanism that runs from (more) imports to capital inflows to (more) private C and I?

    This implies that the less we import (the higher net exports), the faster we grow. This is pure mercantilism, my dear. Alas, not rarely, these old & wrong ideas creep in surreptitiously in the thinking of contemporary economists! Well, no surprise that the WTO cannot agree on anything: let them import our stuff, but let’s not take theirs!

    For an empirical “back of the envelope” confutation of this view, think of the US in 2005, Thailand in 1994, and the many many other countries who grow at 4-6% while experiencing a -6%GDP trade deficit: your view implies that they would grow 10% if the trade deficit was suddently lowered at, say, -2%. But in Thailand 1996-7 the trade balance swung from -6% to (if I remember well) something like +5%, and the rate of growth did not increase: it collapsed. Think also of the low growing surplus countries (Japan for one).

    I have tried myself a few years ago a regression with growth (left hand variable) and the trade balance (right), and with all possible lags the result was that growth is totally independent from the trade BALANCE.

    For a theoretical argument, see

    “Do Imports Hinder or Help Economic Growth?”
    by Owen F. Humpage
    March 15, 2000

    I look forward to read your comments.

    Best regards,