There should have been a dollar crisis three years ago …

So argues DeLong.

Can the dollar be talked down gradually? Consensus economic models say no: once market expectations are that the dollar will be sharply lower in five years, speculators will make the dollar sharply lower tomorrow.

Nevertheless, there is hope: consensus economic models say the dollar crashed three years ago.

He takes comfort from the fact that there wasn’t a dollar crash when there should have been one three years ago.

Actually, there sort of was one.   Not a crash certainly.  But a sustained fall in the dollar.   At least v. the currencies of other industrial countries (see this chart).   And the fall would have been much broader if a whole set of countries hadn’t upped their reserve accumulation dramatically.  Global reserve accumulation rose to $600b in 2003 – and has stayed there.  It looks to me to be a bit stronger than that in 2006.

My interpretation.  The models weren’t all wrong.   But central banks stepped in.

Prices don’t have to adjust if quanitites adjust.  And in this case, the quanity of dollars central banks in emerging economies bought to prop the dollar was the key variable that adjusted.

And that hasn’t changed.   Russia is spending $15b a month keeping the ruble from appreciating (in part by saving rather than spending a big chunk of the oil windfall).  China hasn’t told us how much it spent in April and May pushing the RMB down v. the euro, but it didn’t come cheap.  Even with local interest rates at 2.25 and US rates of over 5, private investors want to bet on the RMB not the dollar.  China spend about $20b a month propping up the dollar in 2005, and I would bet it spent $25b in April and May.

And so on.

That in a sense highlights Paulson’s challenge.  He presumably wants the central banks of the emerging world to let the dollar adjust.   Otherwise the US trade deficit is likely to keep on rising.  But he wants them to let the dollar adjust gradually.

And for foreign investors to continue to fund large US deficits even though they just might be able to earn a bit more if they invested elsewhere.  See the pretty picture posted by the Big Picture.   I think the Big Picture’s chart references total returns on equities, but if the dollar falls, the return foreigners get on their dollars – measured in their own currencies – would be equally unimpressive.  But if the US dollar is going to fall without any significant rise in US rates, foreign investors have to keep adding to their portfolio of US debt despite rather unimpressive returns.

Which likely means that the US will have to continue to rely on central bank intervention to keep the dollar from falling too quickly (and allow the nice slow orderly adjustment we all want) for a long time.    Paulson’s (implict?) message is intervene, but not too much.

Transitions, though, can be hard.   Shifting from a world where central banks intervene to support expanding exports (as US imports rise v. GDP) to a world where central banks intervene to support expanding net interest payments (as the interest the US pays on its external debt starts to rise rapidly) won’t necessary be easy.  More on that later.