Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Martin Wolf explains the fall in the renminbi-dollar

by Brad Setser
December 22, 2004

Martin Wolf did the math: it turns out the renminbi-dollar zone has a current account deficit (projected) of $260 billion with the world this year. The United States’ current account deficit of $650 billion must be offset by a roughly $400 billion surplus in Japan and emerging Asia. A fall in the renminbi-dollar helps in two ways — it at least slows (one hopes, the evidence is still rather lacking) the expansion of the US trade and current account deficit, and it increases emerging Asia’s surplus with the world. Bigger surpluses in Asia = a smaller overall deficit between the renminbi-dollar zone and the rest of the world. So long as Asia is willing to save all of its surplus (actually, all of its current account surplus and then some — China also lends its capital account surplus to the US) in reserves and then lends those reserves to the US, the US can keep on running large deficits.

No surprise: I agree with Martin Wolf’s proposed solution to the US current account deficit. The global adjustment that will lead to a fall in the US deficit will require emerging Asia — basically Asia minus Japan — to shift from a significant current account surplus with the world to a current account deficit to the world. That process will support global demand — and US growth — during the shift, since US demand growth has to slow.

And Wolf is right in another way: if emerging Asia’s deficit is financed by ongoing inflows from foreign direct investment, it does not need to give rise to the same financial imbalances (too much short-term debt relative to reserves) that gave rise to the 1997 Asian crisis. The pattern of global capital flows before Asia’s 97 meltdown was not all wrong. Aging Japan should be running a current account surplus and using that surplus to finance emerging Asia: Japan’s investments in emerging Asia today will let let aging Japan finance a future trade deficit off the profits from its overseas investments. If you invest abroad, in principle, you don’t need to import labor to sustain your living standards as your population ages (and shrinks). Conversely, emerging Asia should be running a current account deficit during the “boom” phase of rapid development, as it did before the 1997 crisis. I have long argued that booming China should be running a current account deficit of $50 billion (financed by FDI inflows) right now, not a current account surplus of $50 billion. China currently saves its $50 billion or so current account surplus in reserves, saves $50 billion plus in FDI in reserves and saves $50 billion or so in “hot money” inflows in reserves as well — accumulating so many dollar denominated reserve assets make no long term sense, as Wolf rightly notes. And all the evidence suggests the pace of China’s reserve accumulation is growing, not falling.

Four other notes:

1) It would be nice to look at global capital flows to the “renminbi-dollar” area, not just the current account numbers. That, alas, is a bit more difficult. We know China is attracting large capital inflows, but don’t know how much of those inflows are coming from outside the “dollar” zone. We know that large amounts of capital — probably around $200 billion — are flowing out of the US (both for FDI abroad and for portfolio investment), but we don’t know what fraction of those flows are going to countries (like China) inside the dollar zone. If the renminbi-dollar zone has a net capital outflow (driven by outflows from the US) as well as a current account deficit (driven the the US current account deficit), its total financing need from the rest of the world would be quite large.

2) It is pretty clear that the “emerging Asia” portion of the dollar zone is attracting net capital inflows from the rest of the world. Emerging Asia then uses those flows to build up its reserves, and thus to lend to the US. China is attracting at least $100 billion in net inflows this year (the final number will likely be substantially higher), though not all from outside the renminbi-dollar zone. Add that $100 billion to the $400 billion Asian current account surplus, and the Asian portion of the dollar zone is financing $500 billion of the $650 billion US current account deficit. My numbers are a rough guess, but it seems likely that in addition to running current account surpluses to offset the United States’s deficit, Asia is attracting some of the external financing the renminbi-dollar zone needs to avoid falling against the rest of the world. For the renminbi-dollar zone to work internally, emerging Asia has to turn capital inflows into Asia into capital inflows into the US — i.e. act as a financial intermediary.

3) If my numbers are right — or close to right — the US needs roughly $150 billion in capital inflows from the non-dollar zone to cover the remainder of its current account deficit. The US also needs a bit more from the world to finance the United States’ citizens own desire to purchase of “non-dollar zone” external assets. If $50 billion of the United States’ roughly $200 billion of FDI and portfolio outflows is going to emerging Asia and Japan, then $150 billion is heading out of the dollar zone. That would imply the US needs to attract around $300 billion in net financing from outside the dollar zone — $150 billion to finance that portion of the current account deficit not financed by Asia, and $150 billion to finance “capital flight” by Americans …

4) There is a huge misalignment inside the renminbi-dollar zone. One pole — China — entered into a de facto currency union with the US at an massively undervalued (in my view, and given Wold’s analysis, clearly in his view as well) parity with the dollar. That generates many of the same set of problems that were created when East Germany entered the German monetary union at an overvalued parity. East Germany has been sustained by ongoing financial transfers from former West Germany. The US is being sustained by ongoing financial transfers from China (and others in Asia). The financial transfers are taking the form of ongoing low interest loans rather than outright grants, but no matter — the net effect is the same, it lets the US avoid adjustment. If the financial transfers were to disappear — and if the monetary union were to be preserved without any change in the renminbi-dollar nominal rate — US prices and wages would need to fall to make the US more competitive (deflation), and Chinese wages and prices would need to rise to make China less competitive (inflation).

Of course, the analogy between East Germany-West Germany and China-the US is not perfect: the former West Germany is supporting an overvalued parity in a smaller (both in terms of population and economic weight) and poorer region — East Germany; China is supporting an overvalued parity in a larger (in terms of economic weight, though not population) and richer region — the US. The tensions in such an arrangement — the financing burden it places on China, the deflationary impact on labor in the US — are why the renminbi-dollar zone cannot survive in its current form without some major internal changes …


  • Posted by anne

    December 22, 2004

    Brad DeLong:

    I don’t see any possibility of a severe crisis for the U.S. as long as our foreign debt is denominated in dollars or consists of equities. The dollar falls steeply, interest rates rise, the U.S. has a slowdown and (likely) a recession as eight million foreign-funded jobs in construction, investment, and consumer services vanish and the workers have to find new jobs in export and import-competing industries–but the big problems all all abroad. Foreigners and their central banks take huge capital losses on their dollar-denominated assets and find the U.S. market for their exports drying up. It’s our currency, but it’s their problem.

  • Posted by anne

    Brad DeLong:

    Let me put it this way: suppose foreign investors lose confidence in the dollar, and suppose that the Fed’s reaction is, “Our monetary policy is to maintain internal balance: we are going to peg the dollar price of the 10-year Treasury bond at what we regard at an appropriate level.” What happens then?

    The dollar falls. The dollar falls until foreign investors think, “It’s undervalued. It’s so undervalued that 10-year Treasuries have got to be a good investment.”

    Are there any negative consequences to that fall in the dollar?

    Yes–for foreign central banks that find that their dollar interest receipts on their reserve portfolios no longer cover the renminbi payments they owe on the debt they issued to buy their reserves. Yes–for foreign producers who find U.S. demand for their exports dropping like a stone. Yes–for U.S. workers who ship, distribute, and sell foreign-made products.

    But the big domestic costs would come only should the Federal Reserve allow domestic interest rates to spike and keep them high. And why should the Fed allow that? Should the Fed raise interest rates to keep the value of the dollar from sinking too low? Should the Fed try to engineer a deeper recession to keep a one-time jump in the price level caused by higher dollar import prices from setting off an inflationary spiral? It’s not clear to me it should. It’s pretty clear to me it would not.

    Thus, as I said, I see a problem for the U.S. economy–and a probable recession–but not a real crisis. Exorbitant Privilege carries the day…

  • Posted by anne

    Brad DeLong

    Interest rates are presumably staying low because lots of people think the break is still a ways away, and think they’ll see it coming and be able to sell before the crunch.

    If there were just one Asian central bank, it probably wouldn’t ever dump the dollar. But there are at least four with huge positions. And then there are the European investors who hold dollar-denominated assets, who will one day decide that they would rather that Asian central banks were the ones bearing the risk of a dollar collapse…

  • Posted by anne

    Brad DeLong

    Let’s distinguish two cases:

    (1) Foreigners decide to dump non-mature Treasuries. There is immediate downward pressure on the dollar, the yen, and the euro prices of Treasuries. The Fed decides to support the dollar price of Treasuries: it buys them for cash. The U.S. money supply goes up. The yen and euro prices of Treasuries collapse–hence the exchange rate collapses. But the U.S. still roughly maintains internal balance (or does the rising money stock ignite a wave of inflation?) And it seems to me the big problems are outside.

    (2) Foreigners decide not to rollover mature Treasuries. The supply of dollars spikes on the foreign exchange markets, as foreigners take their dollars at maturity and run. The dollar collapses. The Treasury turns around and needs to find domestic buyers for its extraordinary new issues. The Fed steps in and buys a bunch of Treasuries to keep their prices from falling too much. The money stock rises, but rough internal balance is maintained… or is it?

    The problem with me trying to think through both these stories is that I think them through assuming that financial markets are in rational-expectations equilibrium. Yet when I look around me I cannot believe that: the dollar is priced too high. The long-term Treasury bond is priced too high.

  • Posted by brad

    I need to read delong responding to roubini …

    But my immediate sense is that the fed will have fight against a lot of negative wealth effects, as the “spike” in inflation reduces the real return on a host of financial assets that are priced for perfection right now. Could the fed overcome this, and push all the costs abroad? who knows …

    Am also not sure we can be indifferent to the impact on foreigners — in a standard adjustment scenario, the trade deficit goes to one -zero % of GDP, enough to stabilize the US debt to GDP ratio, but the current account deficit remains due to a spike in interest payments — which implies continued net inflows. otherwise rather than the 3% current account deficit in wolf’s scenario, you have to go immediately to balance. that implies more adjustment in the US and in the world, and the bigger adjustment (read recession) abroad, the harder it is for the US to adjust by raising exports rather than cutting imports …

  • Posted by anne

    Nouriel Roubini:

    Brad says that the risk in my scenario of a severe rollover crisis is small if our foreign debt is in our currency and is mostly equities. But, increasingly our foreign liabilities are not equities but rather debt and, increasingly ,public debt (see for the BEA latest report on the US Net International Investment Position). In the 1990s our current account deficit was driven by a real investment boom and the capital inflow that was financing it was mostly foreign equities (FDI, M&A, greenfield investments). But since 2001, our current account deficit has worsned in spite of a fall in investment of 4% of GDP. Why? Our fiscal deficit with our public savings of 2.5% of GDP in 2000 turning into a fiscal deficit of 4% of GDP. So, for the last four years foreigners are financing our budget deficit and most of the increase in the net foreign liabilities of the US is debt, not equities and public debt especially.
    By the end of 2003, foreign central banks held 1,472 billion of reserves (mostly US Treasuries), other foreigners held 542 billlion of US Treasuries and other foreigners held $1,852 of corporate bonds (a good chunk of which are GSEs, a semi-public for of debt). Thus, out of $ 9,633 billion of foreign liabilities over 2,000 billion are US Treasuries and almost another 2 trillion is corporate bonds. If you add other foreign debt of the US (liabilities of the banking system), only about 3 trillion of the US foreign liabilities are equity (FDI and equity portfolio). So, over two thirds of our foreign liabilities is now debt.
    Thus, as i already agreed we do still borrow in our own currency (but for how long if we keep on debasing our currency?) while most of our foreign liabilities are now debt, not equity.
    Also, in Brad’s mild scenario the fall in the US $ should lead to a sharp increase in US interest rates; thus, both traded and non-traded sectors will be hurt by high rates, more so the non-traded but also the traded one. Thus, the ensuing recession will hit both traded and non-traded sector. In other terms, what would US growth be if long rates were now 6 or 7% rather than 4% once foreign central banks stop intervening to prop the value of the dollar?

    Wow! That’s certainly a quick response from Brad…

    the Fed directly controls only short term interest rates and any action to stabilize long-term interest rates would be more than unorthodox, it would be an attempt to manipulate long term interest rates that, while not unheard of (Operation Twist in the US in te 1950s or Japanse purchases of long term bonds in the recent Japanese deflation) it would be highly unusual and not consistent with Greenspan philosophy (but Ben Bernanke may think otherwise as he considered such unorthodoxy in fighting deflation).

    But let us assume the Fed does intervene to stabilize the long rate: domestic and foreign residents are dumping US Treasuries (both short and long) not because they want to hold dollar cash assets; it is because they are trying to flee a plunging dollar. Since the US does not have enough reserve to prevent the $ from collapsing, then the question is whether bond market intervention is a substitute to forex intervention to stop the free fall of the dollar.
    My answer is not. First, intervening in the bond market is first of all an act of true desperation; it undermines confidence. Second, that action increases by massive amounts the monetary base in the US; it could more than double it or triple it overnite. Third, in a situation in which investors are trying to flee US assets in a rollover crisis such increase in liquidity puts massive further pressure on the US dollar and since the US does not have the forex reserves to stop the dollar free fall, the dollar falls further and the flight from the bond market is further exacerbated leading to even further liquidity intervention to sustain a falling bond market. Then you get both a currency crack and a bond market crack…Again, the probability of such severe crisis scenarios are small and a nasty shock to the bond market is anyhow the pain that the economic idiots in Washington and the White House need to feel to reverse their tax cuts and give up on social security privatization. Thus, bond market intervention is neither helpful nor desirable.

    I am not saying a severe crisis will occur with certaintly. I am saying that continuing reckless fiscal policies will make it highly likely and force a policy adjustment. That is what we need.

  • Posted by anne

    Well, the leaning is to Roubini and Setser. We will not readily pass our debt-deficit problem abroad.

  • Posted by anne

    Nouriel Roubini:

    The two “cases” described by Brad De Long, where the real effect of the dollar crash are dampened do not appear as realistic.

    In Case 1, the Fed needs to intervene to support long Treasuries; apart from my previous critique of this, the ensuing collapse of the dollar driven by massive liquidity injection leads to sharply higher inflation and the need for the Fed to tighten short rates. Also, markets may test the willingness of this highly unorthodox Fed manuever to defend a particular long rate (that is causing a truly massive liquidity injections and sharply falling dollar that are both highly inflationary). And in this game of chicken the Fed gives up the defense of the long rate peg sooner rather than later.

    In case 2, you got a debt rollover crisis on all maturing debt, be it short or long that is coming to maturity. The De Long solution is to fully monetize the whole stock of public debt that is maturing and the one that would otherwise finance the budget deficit. Then, we are talking of liquidity injection (increase in monetary base) of over $1,000 billion in 2005 and much more if the crisis occurs in 20056 or later. Then, base money is liteally exploding (tripling, quadrupling or more), the dollar is then in real free fall and inflation goes through the roof.

    Note that in all these scenarios you get not just a dollar crash (as you get a currency run and a bond market crash but also a stock market crash as in 1987.

    In fact, as very intelligently pointed out by Billmon in a reply to my blog posting:

    “It seems to me the events of the summer and fall of 1987 provide at least a partial precedent for the kind of rollover crisis Dr. Roubini is describing. The short-term failure of the Louvre agreement to stablize the dollar, plus an abrupt perk-up in U.S. leading inflation indicators led to a fairly massive exodus of Japanese institutional investors from Treasury debt, albeit longer-dated maturies, not T-Bills.(If you look at the Treasury Dept’s chart referenced in the post, you can see the abrupt downward spike in average maturity that this produced.)

    The end result, of course, was a rip-roaring bond bear market, a stock market crash, an emergency injection of liquidity by the Fed, and – depending on whose memoirs you believe – something close to a global financial crisis in the winter of 1988.

    On the other hand, 1987 was in the rosy dawn of our new world order of massive U.S. financial imbalances – domestic savings rates were higher, debt loads lower. And, as Dr. Roubini points out, the Treasury had not yet transformed itself into the modern-day version of the Weimer Republic’s Reichsbank. So in the end, the Fed was able to engineer a soft landing, kind of, sort of.

    Alas, now we’re two decades older, and a hell of a lot more leveraged. Presumably, a good old fashioned run on the T-Bill market would be infinitely more spectacular than the ’87 crisis.(If nothing else, the effect on the monetary aggregates would be truly volcanic.) But if you were alive and sentient back then, and you remember what the world looked like on the evening of October 19, 1987, then you’ve got some idea what’s in store.”

    So, we get a triple whammy ( a dollar crash, a bond market rout and a 1987 style stock market crash…Of course, every other risky asset collapses in this scenario as pointed out by my co-author Brad Setser: Housing collapses, corporate spreads go through the roof, emerging market debt collapse and every other risky asset under the sun….

    Then, we will have to sell our Treasures rather than our Treasuries as discussed in another recent blog posting of mine (

    Sounds too gloomy? In 1987 our fundamentals were much sounder than today both in flow and stock terms…so, this time around “the harder they will fall”…

  • Posted by anne

    Simply to be annoyed. That we are cutting agricultural development assistance to Africa bothers me deeply since agricultural development in Africa is tied so closely with long term conflict, and because we have an obligation for such assistance and do too darn little. Grumble.

  • Posted by tj

    Testing out indenting on an old post.

    Worker shortage in China: Are higher prices ahead?

    To much of the world, China’s reputation for manufacturing can be summed up in one word: cheap.

    But on the factory floors here and in the newly built neon-lit offices, managers say times are changing – and costs rising.Skilled workers and technicians are taking advantage of acute shortages to demand double-digit salary increases. …
    For consumers around the world, rising costs in China could signal an end to the years of deflationary cycles of cheaper and cheaper products
    . …
    The days when you could open a factory near Hong Kong and easily fill it with dirt-cheap workers are over, human resource managers say, especially with a labor shortage of 2 million people in Southeast China
    . …
    At the upper end of the pay scale, competition for technical and white-collar talent in Beijing, Shanghai and other industrial areas is so fierce that salaries for certain categories of employees are approaching U.S. and European levels.

    (Emphasis added)

    Kind of looks like inflation is heating up in China.


  • Posted by Guest