What to say when the markets are listening to Stephen Jen?

The Economist's Buttonwood struggles with something I have also struggled with:

Stephen "Interest rate and growth differentials, not current account deficits" Jen has basically been right on the dollar this year; Buttonwood, Buffet, Volcker, Rogoff and Obstfeld, Roubini and Setser and other dollar bears not so much.

Or, to quote, another dollar bear -- Donald Tsang, Hong Kong's current Chief Executive link: "What's happened to gravity?"

Roubini and I did not necessarily think the dollar would fall below 1.35 against the euro, but we certainly predicted that the dollar would fall against other currencies.   There has been some convergence between the euro and other currencies, but it largely has come because the dollar has risen against the euro this year, not because the dollar has fallen against other currencies.  That is not what we predicted - though I think (and now hope) we were always clear that we saw more risks in 2006 than in 2005.

Just to be clear: the dollar has fallen a bit against the Korean won (if you include the tail end of 2004) and Brazilian real; the dollar/ euro and dollar/ yen are not the only currency pairs that matter.  But generally speaking, the dollar has risen against the other g3 currency pairs and remained broadly stable or fallen ever so slightly against key emerging economy currencies -- so on net, the dollar is getting a bit stronger.

Incidentally, it is a good thing Mr. Jen thinks current account deficits do not matter, because he sure cannot do a current account forecast.   The US current account deficit in the first half of the year was around $395 billion, so I am not quite sure how Jen predicts a $760 billion deficit for the entire year.  There is no sign the current account deficit is shrinking.  Remember, the US oil import bill is likely to be way, way up in September, October and November.  Imports are making up for lost production in the Gulf.  For a tiny fraction of what Mr. Jen earns I would be happy to check his current account forecasts against the recent data ... .

But that is sour grapes - Jen predicted a dollar rally back when the market was expecting the dollar to fall further, and he was right.  

I don't think Mr. Jen is not what Dan Drezner calls a "chartist"; his models just look at a different set of "fundamentals" than the "fundamentals" those of us who obsess about rising US trade deficits and US external debt look at.  I do worry about 2006, though, even if Mr. Jen does not.  Consider:

• At some point, the interest rate differentials that have favored the dollar will stop getting wider, and might even shrink.    Perhaps by a lot, if Calculated Risk, Bill Gross and Stephen Roach are right and there is a real risk that housing prices are close to peaking, and the US consumer is close to maxing out.  That would imply the Fed first stops rising rates and then cuts short-term rates, while European rates stay constant and Japanese policy rates perhaps rise above zero; god-only knows what happens to interest rate differentials at the long-end in that scenario ...
• No more Homeland Investment Act related inflows.   The presence of a one-off tax break for bringing profits held overseas back to the US always reduced the risks in 2005.
• The FY 06 US budget deficit is likely to be bigger than the FY 05 deficit.  
• Even if the US economy slows, I expect non-oil imports to pick up.  THe slow growth of non-oil imports in 2005 looks like an inventory correction after exceptionally fast import growth in 04.   The complication: that inventory correction took pace even as (I suspect) the underlying trend growth in non-oil imports was slowing a bit, in part because of higher spending on oil imports.  I don't expect a return to 2004; but I also don't expect continued flatness in non-oil imports.  
• At some point the stronger dollar will have an impact on US exports, and US export growth will slow.    Market moves do have an impact: at some point next year, I would not be surprised by a series of bad trade numbers even as the US economy slows - assuming, of course, that oil prices stay in their current range.  
• A true perfect storm would likely require a bit of geopolitical risk to keep oil prices in their current range even as a slowing US economy slows oil demand growth.  That is certainly not inconceivable.  Iraq temporarily stopped exporting last week.  The current market requires a lot of production from rather unstable places.
• After several years where the rising dollar value of US assets abroad (mostly the result of a rising euro and pound) limited the deterioration in the US external accounts, the US net international investment position will finally show the impact of a large current account deficit.  To avoid big valuation losses, the dollar needs to close the year at 1.35 v. tthe euro.   No one should be surprised when the BEA reports next summer that the net international investment position deteriorated substantially in 2005, but some folks may be.
• I suspect oil exporters will start to spend more of their oil windfall - or, put differently, oil prices won't keep increasing faster than oil exporters willingness to spend their existing windfall.    Since I suspect that oil exporters are a bit more willing to keep their savings in (offshore) dollars than to buy US goods, so more spending and less saving will prove euro positive and dollar negative.

Finally, let me reinforce one of the points Buttonwood makes, namely that "Official statistics capture only a fraction of what the [central] banks do with their fast-growing foreign-exchange reserves ($2 trillion higher since 2000)."  Recorded central bank inflows into the US certainly have slowed this year, but, interestingly, global reserve accumulation has not slowed.  Rather the set of countries adding to their reserves has changed.  Think China, Russia and the oil exporters of the Middle East rather than China, Japan and everyone else in Asia.

Look at the data from the first half of the year.

China's total reserve increase (adjusting for valuation) was around $137 billion (including the $15 billion transferred to a state bank, the unadjusted increase was $116 billion); recorded inflows to the US were about $49 billion.  We know what China did with around 40% of its money, but not what it did with the remaining 60%. 

And we know far more about China than about Russia and the Middle East.  Russia's current account surplus in the first half of the year was $60 billion (its reserves grew by $32 billion - probably more if you adjust for valuation - even as Russia paid down its external debt).   OPEC - which is statistically dominated by oil exporters in the Middle East - exports about three times as much oil as Russia.   So take $150 billion -- a bit less than three times Russia's current account surplus -- as a ball part estimate for OPEC's current account surplus in the first half of the year.  Consider one key country: Saudi Arabia.  SAMBA forecasts a 2005 Saudi current account surplus of $96 billion, and a $47 billion increase in official Saudi assets.   That is probably a bit low ...    

So the combined current account surplus of Russia and OPEC almost certainly topped $200 billion in the first half of the year, with Russia and Saudi Arabia leading the way. 

Where did that money go?  Don't look to the US data for any clues.

The total purchases of US bonds (as measured by the TIC data) by both private investors and central banks in Russia and the Middle East: $5 billion.  Total purchases of bonds and stocks (as measured by the TIC data):  $6.5 billion.  Data here.

In other words, we know the US TIC data is not picking up what the Russia and the Middle East are doing with their petrodollars.   The banking data may hold more clues, but it comes with bigger lags.

But in some sense, Russia and the Middle East have to be financing the US - or financing European capital inflows to the US.  The estimated 05 current account surplus of the oil exporters ($400 billion) is larger than the total current account deficit of the world IF you exclude the estimated current account deficit of the US.  The global "adding up" constraint is real.  If the US accounts for 75-80% of the world's current account deficit, in aggregate, 75-80% of the world's current account surplus has to find its way to the US.

It does seem that US 10 year yields are finally ticking up, but that probably has more to do with higher levels of inflation than with any shift in foreign demand.  Buttonwood is right that private interest in Treasuries has picked up even as demand for Treasuries from official actors (above all the Bank of Japan acting for the Japanese MOF) has fallen; I just am not sure that private interest is unrelated to the surge in offshore petrodollars ...   unfortunately, I cannot point to any clean set of data to support my suspicions.