Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


Stephen Jen might want to rejigger his model

by Brad Setser
January 30, 2005

Apparently, Mr. Jen thinks the dollar — despite the United States’ 6.2% of GDP current account deficit — is significantly undervalued against all major currencies, and fairly valued against most Asian currencies.

Call me old-fashioned, but in my book, a large current account deficit is a sign that a country’s currency is overvalued, not undervalued. Another sign of overvaluation: heavy central bank intervention to support the currency. The US just happens to have outsourced currency intervention to a group of largely Asian central banks, who spent close to $200 billion in the fourth quarter to keep the dollar from falling against their currencies.

However, Jen argues that the dollar today is like the dollar after in 1988, after the Louvre agreement (Plaza = G-7 signal the dollar had to fall, Louvre = G-7 signal the dollar had fallen far enough), of the dollar in 1995, after it has slumped v. the yen. It is poised for a large rally.

Hey, why worry if a dollar rally would push the US current account deficit toward 7% of even 8% of GDP — or somewhere between 70 and 80% of US export revenues?

I recommend that Mr. Jen add the current account deficit to his model of fair currency valuation. Adding the current account deficit would highlight the difference between 2005 and 1995, and 2005 and 1998. In 1988, the current account deficit was 2.3% of US GDP, down from 3.4% in 87 and on track to fall toward 1.4% in 1990. In 1995, the US current account deficit was only 1.5% of GDP. The US net external debt position was a lot better back in 1988 and 1995 than it is now too …

Of course, Morgan Stanley has Stephen Roach as as well Stephen Jen. And a market requires buyers as well as sellers. But I do wonder if Morgan Stanley is willing to put its money behind Mr. Jen’s model. After all, doesn’t the model say the dollar is 20% overvalued v. the euro, and should trade at 1.09?

One caveat. While I — like Bill Gates (thanks Jesse) — think the dollar is overvalued today on a broad trade-weighted basis and a further fall in the dollar is likely, that does not necessarily mean that the dollar is overvalued v. all other currencies. Clearly the dollar already has fallen by much more against the euro than against the renminbi. Still, how can Jen discuss the fair valuation of the dollar without at least mentioning the current account deficit? I will be travelling for the next few days, so do not expect any more posts before Tuesday.


  • Posted by Andrew Boucher

    Can just look at the current account deficit in the case of a depreciating currrency (as with the dollar over the past few years)? What about the J-curve? Maybe the U.S. is just at the bottom of the J and now, as exports/imports finally adjust, the current account deficit will diminish. If I were a betting man, I’d wager that the euro is more over-valued than undervalued at this level. Asian currencies are, however, still undervalued.

  • Posted by brad

    Andrew, current US import growth is not explained by the J-curve. In the classic J curve phase of current account adjustment, import volumes are flat or falling, but higher import prices still increase the overall import bill. Right now, non-oil US import growth is largely explained by rising import volumes, not rising import prices.

  • Posted by Andrew Boucher

    Brad: OK no problem with that. But hen it’s not the current account deficit but import/export volumes which need to be added to Mr. Jen’s model.

  • Posted by anne

    January 28, 2005

    Brad Setser:

    “I suspect the december trade number will be closer to 57-58 than 55, i.e. i think you may have been overly conservative. but if you extend out q4 for the next four quarters, the deficit is $735 billion. Get rid of the + 20 billion on income implicit in that forecast, and it goes to $755 billion …

    “That is $90-100 b worse than seems likely this year (have not looked at a spreadsheet to get a precise 04 estimate). I don’t see how the highly paid folks on wall street are coming up with their numbers here — ok, that is a bit unfair, Goldman has laid out their assumptions and they are predicing a turnaround in the trade numbers in h2 based on the lagged impact of the $’s fall. But with no signs non-oil import growth is deaccelerating, and pretty good evidence that US export growth (along with global export growth) is slowing, that seems a bit optimistic …”

  • Posted by jm

    From Andy Xie’s commentary at

    “While virtually everyone at the MacroVision conference was bullish about the yen, one Japanese participant from a big insurance company stood up and begged to differ. ‘The spread between Treasury and JGB is 300 bps. That’s good enough for us to buy’, he said. No one reacted to his comment.”

    “I checked the data afterwards. Sure enough, the average spread between 10Y Treasury and JGB was 299 bps in the past three years. While hedge funds borrow dollars at 2.25% to buy yen with zero yield, Japanese insurance companies are buying US Treasuries with 300 bps yield pickup. With 300 bps pickup, they are better off buying Treasuries against JGB if dollar-yen is higher than 80 ten years from today. Of course, if you listen to some hedge funds, dollar-yen should reach 60 or lower and soon. With dollar-yen at 60, Japan’s per capita income would be 60% higher than the US’s. Have you been to Japan lately?”

  • Posted by jm

    Though only indirectly related, the info at is very interesting.

  • Posted by anne

    Janaurd 31, 2005

    US tax amnesty could rake in $100bn
    By FT.COM

    A tax amnesty for multinationals is expected to bring approximately $100bn of foreign exchange earnings into the US over the next few months after a stronger-than-expected take-up by companies last week.

    The flow of money, triggered by tax breaks in the controversial American Jobs Creation Act, is so large that many currency strategists expect it to give noticeable support to the dollar.

    Last week, four pharmaceuticals companies Johnson & Johnson, Eli Lilly, Schering-Plough and Bristol-Myers Squibb committed themselves to repatriating $37.4bn. Pfizer said it hoped to bring back a further $37.6bn held offshore.

  • Posted by Robo

    I suppose the issue with the tax amnesty is whether the funds that are going to be repatriated are currently held in USD or other currencies. The risk is that most are already in USD.

  • Posted by brad

    FT on monday estimated the total in fx at 100 b.

    Not small. But not enough to finance the US for 2005 either.

    I am back, more tomorrow.

  • Posted by anne

    At Forum, Leaders Confront Annual Enigma of China

    DAVOS, Switzerland – In almost every panel discussion at the annual meeting of the World Economic Forum here, there comes a moment when somebody mentions China.

    A hush typically ensues, as panelists draw their breath, gather their thoughts and struggle to put the bewildering vastness of the topic into a few words.

    ‘China is going to be the change agent for the next 20 years,’ said Bill Gates, the chairman of Microsoft, when asked about the country’s future by the television interview host Charlie Rose.

    China’s staggering potential, coupled with the steep language barrier and cultural discomfort of many Chinese who come to this conference, has made it Davos’s annual enigma.

  • Posted by amitabh

    I am sympathetic to Steve All these current-account based models of dollar valuation (inevitably reduced form)do not consider that in addition to the goods markets, there is an asset market equilibrium too. The U.S. current account deficit is being financed mainly by Japan and Europe. If in PPP terms, the dollar is fair or undervalued to the yen and euro and real interest rates are higher in the U.S., why should investment not flow into the U.S.? BTW, external debt service to exports is well below 15%. So no cause for alarm there. Steve Roach has been crying wolf for years. AA

  • Posted by brad

    AA —

    a) it is easy to keep your external debt service to exports ratio down when real interest rates are negative. the US external debt to exports ratio is about 300% — quite high. net interest currently is negative or zero (higher earnings on US assets than payments on liabilities), but project out a couple of years, when the US external debt to GDP ratio is 40% rather than 30% and interest rates are at more normal levels, and you get 2% net interest payments (20% of exports), a current account of over 7% of GDP (70% of exports) and gross interest/ profits on FDI payments of closer to 40% of exports — all scary numbers. Jen’s model seems to downplay classic external sustainability variables entirely; that may have worked in the past, i personally doubt it will work going forward.

    b) the US current account deficit is most certainly not financed mostly by “japan and europe” — It is heavily financed by Japan, yes, but emerging asian reserve accumulation has been extraordinary and I suspect Emerging Asian central banks alone are in pole position, followed by Japan, and then europe. But europe’s current account surplus is not big enough for it to be providing large scale financing for the US; if may be intermediating between oil exporters who want euros and the US financing need, but if that is the case, i would like to see the evidence.

  • Posted by amitabh arora


    Of course no one sets out to “finance a deficit”. Except for pegged currencies, capital flows are autonomous. Look at the purchase of U.S. assets from Europe – in excess of $200 billion. More than Japan, more than emerging Asia. This was more than Europe’s multilateral current account surplus (as you point out). How dis this happen? Eurpeans bought U.S. assets and the rest of the World bought euro assets. But the decision by europeans to invest in the U.S. is independent of capital inflows into europe. if you were bearish on U.S. assets, why would private european investors (or OPEC for that matter) invest in the U.S. A lot of the flows into the U.S. are private, not official.

  • Posted by brad

    AA — thx.

    any chance you could direct me to the best source of data on European (private) purchases of US assets. would like to know more about the distribution across asset classes. Sort of want to know what European investors are thinking re: potential risk/ reward given the currency risk, and when they were doing the purchasing — i.e. in the spring/ summer with the $/ euro at 1.20-1.25, or later in the year?

    I sort of doubt that this is more than from emerging Asia tho. China alone added $200b to its reserves, and the other emerging Asian economies added a fair amount too — the overall emerging asia reserve total is, I suspect, in the range of $350b +. Unless 1/2 went into non-dollar assets, that would provide at least 200b in financing from the central banks alone.

  • Posted by amitabh arora

    Brad, the data is from TIC. There are few official purchases from Europe – there are no large central bank reserves there. As for reserve accumulation, i had seen estimate indicating the share of euros was increasing.