Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


The dollar is no longer a high carry currency …

by Brad Setser
March 21, 2007

That at least is what the market seems to think.  The carry trade has bounced back

The Australian dollar — a high carry currency — is setting new highs (v. the US dollar at least).  The kiwi has recovered.   Iceland's krona has been fairly resilient in the face of a rating downgrade.  The Malaysian ringit and Thai baht are heading back toward their 1998 levels.  The Indian rupee — a new carry trade destination currency — has been doing well.   And, judging from the strong ongoing growth in Brazil's reserves, inflows in Brazil have continued.  Brazil's reserves are up $5.5b so far in March.  China's low interest rates (and capital controls) keep the RMB from being a destination currency, but its stock market has bounced back — hot money doesn't seem to be shying away from China, or perhaps Chinese money isn't heading out of China even with higher interest rates in the rest of the world.

The US dollar, by contrast, is once again rather weak.  Especially against the Euro.  1.34 is back to the dollar's lows of 2004 and early 2005.  The Fed may be on hold — and, if Dr. Roubini is right, about to cut.   But US rates are still higher than eurozone rates, let alone Japanese rates.  

So why — as Teis Knuthsen of Danske bank notes — has the dollar diverged from the high carry currencies?

After all, Australia, New Zealand, Iceland, India and Turkey all have current account deficits, and all but India have large current account deficits.  How can the Aussie dollar be strong and the US dollar weak when both offer a yield pickup over the euro and yen?

Two thoughts:

One, most of the currencies of other current account deficit countries offer a bit more carry than the US dollar.  So if you are going to finance a country with a big deficit, why not finance the country that offers the most yield?  That is the logic behind my Deutsche Bank friends' recent recommendation to go long Turkey v. South Africa.   The same logic may apply globally. And if you want carry and don't want to finance a current account deficit, there is always Brazil.

Two, on an absolute scale, the deficits of even Australia and Turkey — let alone tiny countries like New Zealand and Iceland — are fairly small.  Japanese households looking for yield could easily finance the entire current account deficit of New Zealand.  Swiss private bankers (or Swiss households) could do the same for Iceland.   All any of these small countries need to do is offer a bit more attractive risk/ reward combination than the US and get a small fraction of the global flow of funds.

The US external deficit, by contract, is quite big.  As Alberto Musalem Borrero notes, the carry trade is far too small — even taking the biggest estimates — to meet the United States roughly $850-900b financing need.  And that is just what the US needs to cover its current account deficit — it needs to borrow even more if it wants to increase its holdings of foreign assets.

The US deficit is bigger than China's (not small) surplus, Japan's (not small) surplus and the oil exporters' (not small) surplus.   You need to sum up all three big surpluses to come close to meeting the United States need for financing.

And there are a host of signs that the financing of that deficit has gotten a bit more difficult recently.   The dollar's weakness against the euro is one.   But so to is the acceleration in global reserve growth.  Stephen Jen — quite accurately in my view — has noted that global reserve growth is now close to $80b a month.  That is about $1 trillion a year.  Throw in $100b or so in non-reserve Chinese foreign asset growth in state banks and state companies and $100b or so in non-reserve growth in the oil investment funds, and total official asset growth (counting the Chinese banks as part of the official sector) is potentially close to $1.2 trillion.

That is more than enough to finance the US deficit.   As Stephen Jen notes, inflows from emerging economies are the fundamental source of all the liquidity sloshing through the markets. 

The ‘real’ sources (i.e., the world’s savings-investment surplus) of global liquidity remain robust:  the Asian countries and the oil exporters continue to generate some US$800 billion worth of combined C/A surplus.  Global long-term interest rates, as a result, have remained near a generational low.

Those looking for a reason why real rates in all the advanced economies that are on the receiving end of these inflows — largely official inflows — need look no further.   Mohammed El-Erian of Harvard now believes emerging economies are a source of stability (hat tip: Global Liquidity Blog). They have been willingness to finance the US when others won't.

But the official sector also doesn't just want to add dollars to its portfolio.

Alberto Musalem Borrero implicitly contrasts the carry trade with official inflows (after all, those official flows play a bigger role in financing existing large deficits at current prices).  But I am increasingly interested in the intersection between the carry trade and global reserve growth.  Think of a hedge fund borrowing yen to buy Brazilian real or Indian rupee, and thus fueling Indian and Brazilian reserve growth.  Both India and Brazil then invest some of their growing reserves in the US.   That is one way of financing the US deficit —

But it is a bit less "efficient" than a hedge fund borrowing yen to buy dollars.  Brazil keeps most of its reserves in dollars, so it does its part to sustain the current unbalanced world.   India — like Russia — only keeps about 1/2 its reserves in dollars.  And with Indian and Russian reserves both growing fast, global reserve growth consequently has to exceed the US current account deficit if official investors in the emerging world are going to finance the US deficit … or so it seems.

It is an interesting world. 

And it could be an even more interesting world if the Fed actually starts to cut — and global demand for dollars falls further.  A slowing US economy will cut into the trade deficit, but the overall US current account deficit may not fall that fast.  I think the income deficit is set to rise sharply.  And that means that even a slowing US that doesn't offer much of a yield pickup over the rest of the world would need to attract a very large share of the world's spare savings.  


  • Posted by Emmanuel

    Dr. Setser, please elaborate on “the income deficit is set to rise sharply” either here or in a future post. I too believed this would be the case, but those dastardly Q4 2006 figures had me scratching my head. While you’re probably right, US BOP accounting has been funky as of late with humongous statistical discrepancies that could hide the GDPs of several developing countries.

  • Posted by DF

    “And it could be an even more interesting world if the Fed actually starts to cut — and global demand for dollars falls further. A slowing US economy will cut into the trade deficit, but the overall US current account deficit may not fall that fast. I think the income deficit is set to rise sharply. And that means that even a slowing US that doesn’t offer much of a yield pickup over the rest of the world would need to attract a very large share of the world’s spare savings.”

    Could that mean also that unless the FEd is ready to engage in very heavy open market operations, buying huge chunk of US debt, it may prove to be unable to set the US rates.

    I mean if there s a recession AND the global reserve growth in dollars stops … Is there a risk that the USA will have to raise rates to attract foreign investors ?

    When the fed lowered the rates to 1% post 2001, was it not with chinese support ?
    Could they do it again without either :
    – strong support from the rest other central banks
    – strong open market activities buying US federal debt and monetising it
    – Sudden return to a normal interest rate curve with a low short term rate unable to reduce the long terme one
    – a sudden deflation so that real interest rates rise despite a fall in nominal terms ?

    I mean if you want “to attract a very large share of the world’s spare savings” there should be a time when this starts to have a cost. Or I m wrong ? Where ?

  • Posted by DF

    Emmanuel, the fiscal revenues of the federal government are increasingly dependant on profits, profits vary faster than wages. A recession would hit federal tax revenues hard.
    Besides the fall in housing prices is bound to hurt states and US schools tax revenues and this will turn into another problem.
    Finally of course higher unemployement will create all sorts of federal expanses.

  • Posted by bsetser

    Emmanuel — tis for a future post. setting aside the q4 improvement in net income from fdi, which looks a bit stange/ may be revised (reinvested earnings by foreigners were very low), the key to the lack of a bigger decline in 06 has been that the interests rate on us lending (mostly short-term) rose a lot faster than the interest rate on US borrowing. the gap is now something like 75 bp (I would need to consult my spreadsheet, if not more). I expect that gap will fall as short-rates on are on hold/ average US borrowing rates should rise above 4.4-4.5% (sure treasury yields are low at the 10 yr mark, but the US doesn’t finance its deficit exclusively by selling 10 yr treasuries, and most other things pay a higher rate). combine that you get a big deterioration.

  • Posted by Guest

    Perhaps hedge funds and others indirectly facilitate diversification of central bank reserve positions used to recycle the U.S. current account deficit – i.e. ultimately allowing central banks of surplus countries to diffuse some of their concentration risk in the U.S. dollar, as others take up the difference at the margin. This reminds one of arguments favored by Greenspan and Bernanke, in defending the role of hedge funds as conduits for the transfer of risk around the system – the invisible hand at work.

  • Posted by Guest

    ( # 2 )

    ” The ‘real’ sources (i.e., the world’s savings-investment surplus) of global liquidity remain robust: the Asian countries and the oil exporters continue to generate some US$800 billion worth of combined C/A surplus. Global long-term interest rates, as a result, have remained near a generational low. ”

    I’m puzzled why some (including Bernanke) claim the existence of a global savings/investment imbalance. Clearly there is an imbalance between the U.S. and the rest of the world. How that translates to a ‘global’ imbalance is far from clear. Why should the surplus side of the equation bias interest rates lower rather than the deficit side bias them higher? We know interest rates are low. But why should this asymmetric interpretation of surpluses over deficits be a reasonable explanation?

  • Posted by bsetser

    Guest — low real interest rates are central to Bernanke’s argument. If the deficit side was driving the equilbrium (i.e. high real rates in deficit countries were pulling scarce funds out of the rest of the world, pushing up rates globally), there wouldn’t be a glut of savings so much as a surplus of spending/ investment …

    Other guest — I don’t quite see the HF channel. There is a channel where Central banks bid up the euro and bid down eurozone rates to the point where private capital flows from Europe to the US, but that channel works with real money. And I see another channel whereby Hedge funds effectively add to central banks dollar (and even more so dollar/ euro) concentration risk, but in effect borrowing dollars/ euros and sending them in emerging economies who then send them right back to the US/ Europe (think Brazil, with tons of inflows and no external financing need, so rapid reserve growth right now … ). Could you spell your argument out a bit more; I don’t yet see it.

  • Posted by Guest

    I was thinking of your Brazil example as an example of my general point, which was not that individual central banks like Brazil avoid U.S. dollar concentration, but that the activity of hedge funds can prod such reserve growth across a greater number of central banks. To the degree that more central banks are involved in aggregate U.S. dollar reserve accumulation, the average burden of reserve accumulation is less (i.e. the distribution of aggregate reserve accumulation among central banks becomes less concentrated), although the standard deviation of such average reserve holdings is surely still huge. I was attempting to extrapolate such a general point from your example, but perhaps I misunderstand.

  • Posted by Macro Man

    At the risk of belaboring a point, the dollar IS a carry currency…but the biggest net sellers of dollars against the euro, sterling, AUD, etc are already super-long. I refer, of course, to the FX reserve guys, whose activities then colour the view of the rest of the market.

    Of course, it isn’t that simple…markets are after all discounting mechanisms. And from that perspective, the dollar’s expected rate premium over the euro by the end of the year has narrowed 35 bps over the last six perhaps it’s no wonder that EUR/USD is higher.
    More broadly, markets have been confronted with hawkish surprises in EMU, UK, Oz, NZ over the past couple of months…but more or less only dovish surprises in the US since the sub 50 ISM.

    On a longer term, basis, the failure of the dollar to behave as a carry currency is more than an academic problem. I suspect that a lot of currency managers have tried to be long dollars over the last few years because their models tell them that it is a carry currency. Industry statistics (various performance indices) suggest that these funds have, in aggregate, been losing money for quite some time…with the failure of the dollar to perform as expected probably a reasonable explanation.

  • Posted by bsetser

    The increase in Brazilian reserve growth hasn’t led to a decreate in indian or russian or chinese or saudi reserve growth (tho saudi reserve growth is lower with oil at $60 than $70) … so i would characterize the current situation as one where the $ exposure of all the big players is increasing.

    Macroman — interesting points, as usual. I wasn’t trying to argue that the US isn’t a high carry currency, only to note that it hasn’t been performing like the other high carry, big current account currencies — perhaps b/c, as you note, there are a lot of big players holding more dollar exposure than they want out there (and they ironically are in the strange position of seeing the dollar holdings grow rapidly b/c of their countries fx policies even as they want to reduce the $ share of their portfolio).

  • Posted by Guest

    Peter Schiif’s Economic Commentary:

    Don’t Uncork the Champaign Just Yet

    By omitting a few key words from their most recent statement, the Fed led Wall Street to the premature conclusion that the next move in interest rates will be down. With the economy clearly headed for recession, there is no doubt that the Fed would like nothing more than to do just that. However, given that it wants to pretend otherwise, and considering the damage it would do to the already shaky U.S. dollar, an actually rate cut seems highly suspect. Rather than offering a true assessment of the current economy, the official statement that follows Fed meetings has become a political farce used primarily to placate markets. For the bond market and the dollar, the Fed pretends that inflation is still under control, and that the Fed remains poised to snuff out any inflationary sparks should they appear. For Wall Street, the housing markets, and the economy in general, the Fed pretends that the economic expansion will continue, but shows mild concern that growth might falter. If the Fed were to admit that the economy was in trouble, the stock market would sell off, led lower by a collapse in the dollar and a potential spike in long-term interest rates. With its parsed language, the Fed preserves the pretense that all is well while simultaneously allowing for the possibility of future easing. So by validating the goldilocks scenario, but holding the door open to future rate cuts, they can have their cake and eat it too. One of the biggest bones the Fed threw to the markets in its last statement was its failure to directly mention the problems developing in the mortgage market. This omission suggests that the Fed is not overly concerned with the subprime crisis, or the possibility of that weakness spreading into the broader mortgage market or the economy in general. In other words, a problem isn’t a problem until the Fed says it is. This ignores the fact that the Fed is reluctant to actually identify a problem, no matter how severe; for fear that such recognition alone might spark an even greater panic. So with the apparent blessing of the Fed, Wall Street can now borrow a page from the Las Vegas promotional playbook and claim that the “what happens in sub-prime stays in sub-prime.” Unfortunately, like an out of work showgirl with a folder full of embarrassing photos, the problems with subprime will soon show up on everyone’s doorstep. Think of the Fed as a juggler trying to keep five balls in the air simultaneously. Those balls are the stock market, the bond market, the dollar, the housing market, and the economy. If the Fed tells the truth, all the balls will come crashing down. So it says what it needs to say to keep them all in play. However, my guess is the first ball to fall will be the dollar, which sold off immediately following the release of the Fed’s statement. Compounding the problem is a recent report that China may no longer be willing to expand its foreign exchange reserves. This means the dollar ball is about to get a lot heavier. Once the dollar breaks down the bond market ball will be that much more difficult to keep aloft. Once it falls, the rest will soon follow. The bottom line is that waiting for the next rate cut is going to be a lot like waiting for Godot. The Fed wants everyone to think one is coming, but will likely never deliver the goods. If I am wrong and the Fed actually does cut, expect the easing cycle to be extremely short-lived, as an embarrassed Fed will be forced by the bond and currency markets to quickly reverse course. Wall Street mistakenly believes that the Fed’s job is to keep the expansion going. In reality, the Fed’s job is to take the punch bowl away from spendthrift American consumers and the leveraged speculators lending them money. If the Fed were to actually do its job, they would accelerate the onset of the inevitable recession. Perpetuating a phony expansion only compounds the problems that a recession would help solve. However, by repeatedly spiking the punch bowl rather than removing it, the Fed merely guarantees a much bigger hang-over when it inevitably runs dry.

  • Posted by Alex

    I actually read your big paper that you wrote with Roubini and to me the big question is what happens when these countries decide they have enough dollars.

    Or lets make the story really interesting and ask what happens when the system for recycling dollars collapses. What I’m talking about of course is the real estate bubble. A good deal of the real estate bubble has been perpetuated by foreign investors( or central banks) recycling their dollars via mortgage backed securities. What happens when that market stops working. I.e. there is a spike in defaults or huge concerns in credit worthiness become apparent.

    If there is an MBS collapse will everyone run for the door. Prompting a huge crash in the MBS market and a massive liquidity crunch followed by a simultaneous global economic collapse in conjunction with a dollar collapse.

    Or will we have an orderly S&L style bailout of Fannie Mae and Freddie Mac aka Fanron?

    At this point keeping these two issues apart and treating them separately is probably quite unrealistic because the MBS markets have been part of the giant dollar recycling engine.

    I think the danger we have here is that all these major players have large quantities of dollars and if the Chinese and other people decide that they want “real assets” instead of treasury bonds and MBS’ we could have serious problems. Begining with protectionism and ending with hyperinflation.

    With China’s decision to stop accumulating dollars coupled with a collapse in the MBS markets we potentially have a very ugly situation on our hands.

  • Posted by OldVet

    I don’t see the Apocolypse just yet, but the MBS question is valid. If private foreign investors decide for any reason to lighten their purchases, do you think they’d buy Treasuries instead? Or equities?

  • Posted by Alex

    “If private foreign investors decide for any reason to lighten their purchases, do you think they’d buy Treasuries instead? Or equities?”

    I think the equities market is more likely to get a bump. After all with all this talk of a massive portfolio its difficult for me to see the Chinese investing all their proceeds into even more bonds. Of course that could have a very destabilizing affect if they make good on that promise. We could be talking about another stock bubble ala 2000.

    However I think that if the MBS market goes to hell its hard to see how the economy could not collapse as the housing market has been the driving economic engine here not just in the US but throughout all of Europe.

    OFHEO actually came out with a report back in 2001 about what it called system risk or the collapse of Fannie Mae and Freddie Mac. The fellow who wrote it was dismissed a week after its release.

    The predicted consequences were ugly with Fannie/Freddie taking out most of the investment banks in the process as Fannie/Freddie fail since they are the counterparties in many large derivative contracts with Fannie and Freddie. Also the authors believed that if both GSE’s failed that it would create a massive financial crisis in the mortgage market.

    St Louis Fed govenor William Poole has also said a few words on the subject. But approached it from the angle of the GSE’s being imperfectly hedged so there were a spike in short term rates they would be in big trouble:

    Of course you have to remember these are the more obvious consequences. Who knows what would happen if the more exotic consequences were factored in and a large dollar change and interest rate spike plays itself out in the derivative markets? The consequences are potentially enormous.

    Something wicked comes this way.

  • Posted by Guest

    “If there is an MBS collapse…”

    most of the problems are in subprime and ARMs, which the GSEs largely sat out (thanks to the repubs, FNM/FRE ironically have been strengthened — financially — relative to the private sphere); i’d be more worried about prepayments than a collapse…

    even if there was a wholesale collapse, china et al would just end up owning large swaths of US residential real estate! there’s more than one way around CFIUS 😛

    the mechanism i can see for an MBS collapse is that ARM resets over the next couple years take away enough spending such that the savings rate actually rises, while collapsing profit margins from top line sales undermine corporate credits and business/professional employment — in the context of dollar decline, higher inflation and real interest rates.

    iow, if US consumption falters, the US becomes a less (levered) place to recycle dollar liquidity and the world will move on to finance the next consumption binge someplace else (presumably w/ higher productivity, more sustainable growth/development and better assets).

  • Posted by Alex

    “most of the problems are in subprime and ARMs, which the GSEs largely sat out (thanks to the repubs, FNM/FRE ironically have been strengthened — financially — relative to the private sphere); i’d be more worried about prepayments than a collapse…”

    Obviously you and I probably not talking about the same GSE’s . Fannie and Freddie both haven’t finished cleanup their accounting problems which run into the billion of dollars. In fact Fannie just filed a 6 billion dollar earnings restatement in December:

    In fact its was the first time Fannie had done an annual report in three years! In fact Fannie hasn’t filed earnings for 2005 and 2006!

    Although Freddie Mac has filed earnings thet have some great news to report too.

    “Freddie Mac said last month that it will no longer buy those that it deems to be the most vulnerable to foreclosure. The change will take effect in September.”

    So much for Freddie not having subprime exposure the new rules do not even take effect until September!

    “most of the problems are in subprime and ARMs, which the GSEs largely sat out ”

    Obviously we must be talking about different GSE’s.

    All thus of course is bad but what’s worse is that the GSEs’ have more outstanding debt than the US government.

    Its going to be a catastrophe that will make the S&L scandal look like a budget item.

  • Posted by Guest

    As America falters, policymakers must look ahead: As a US slowdown looms, it is better to respond to current circumstances than focus on past mistakes, writes Lawrence Summers.