Some would say it never really went away.
Emerging market economies did not stop accumulating reserves in 2005. Best I can tell, adjusting for valuation effects and including all the foreign assets of the Saudi monetary authority, emerging markets probably added closer to $500 billion than to $400 billion in their reserves. China, Russia and Saudi Arabia together combined for about $380 billion ($260 billion in China by my estimates, and a bit over $120b or so in Russia/ Saudi Arabia once you take into account valuation effects).
Yet there is little doubt that - setting China aside - the set of countries adding to their reserves shifted during the course of 2005.
In 2004, reserve accumulation was concentrated in Asia. Japan (early in 2004), Korea, Taiwan, India and China (along with a host of smaller countries). That fit well with the Bretton Woods 2 hypothesis, which argues that reserve accumulation is linked to the need to move surplus labor out of the country into the cities. Think China and India. Korea, Taiwan and Japan have already moved labor out of the countryside - but they worry about their competitiveness relative to the upstarts following a manufacturing-export based growth strategy.
In 2005, reserve accumulation shifted from population rich countries to resource rich countries. Think Russia and Saudi Arabia. Leaving China aside, and that is an awfully big country to set aside, the key Asian economies didn't intervene heavily in the fx market after the first quarter of 2005.
That changed in January. Big time. How fast are emerging market reserves now growing? And why haven't more central banks defected from the dollar buying-cartel -- a topic that has come up on Martin Wolf's uber-blog?
The Wall Street Journal reports Asian countries not named China added $30 b to their reserves in January. IDEA (a consultancy) says the valuation-adjusted number is closer to $20b (the euro's rise v. the dollar in January tended to inflate dollar reserve growth). Korea added $6.5b, Taiwan $4b, Singapore $3.6b (with some valuation gains no doubt), Hong Kong $3.5b, India 2.3b and so on.
The other emerging markets punched above their weight as well. Russia's reserves are up by $6b. Brazil's reserves are up by $3b. Turkey's reserves seem to be up a bit - and it repaid $1.2b in debt in January. Brazil and Turkey both bought over $20b in the foreign exchange market in 2005, but their headline reserve growth was lower because both also made significant repayments to the IMF. Still, Turkey's reserves increased by $14b in 2005, and they seem to be growing at a comparable pace in 2006. Argentina's reserves fell in January - it paid the IMF $9.5b in cash early in the month. But after paying, its reserves grew by about $1 billion.
The average monthly increase in the foreign assets of the Saudi Monetary Authority in 2005 was around $5b. Chalk up a similar amount for January.
Setting aside Argentina's big one off payment, I don't think it is a stretch to say that emerging economies outside Asia added $20b to their reserves in January. Russia, Brazil and Saudi Arabia alone generated an estimated $14b increase (not adjusted for valuation) alone. January was a good month for oil exporters and capital importers.
We don't know the number for China. Let's estimate $20b. That assumes that there was a pick up in hot money inflows into China to make up for a seasonal fall in China's trade surplus. That makes sense to me - look at the broader pick up in flows to Asia, and to the emerging world more generally.
$20b plus in Asia. $20b for Latin America and the oil exporters. $20 b for China. That sums up to $60b. And that is after adjusting for valuation. The raw number is something more like $75b (it doesn't make sense to try to do anything formal til China leaks its reserve increase)
$60 b in a month. A few years ago, $60b would have been close to these countries annual reserve increase. It is a very big number. It works out to $720 billion a year.
If all that flowed into dollars, it isn't quite enough to cover the United State's trade deficit - which was $65b in December.
Central bank financing of the US pretty clearly hasn't disappeared.
From the data we have, central bank financing of the US probably was as high as it has ever been in January - even without any intervention from Japan.
What happened? Are Michael Dooley, Peter Garber and David Folkerts-Landau even more right than they thought, with reserve accumulation from the resource-rich joining reserve accumulation by the labor-rich as the defining feature of the international monetary system?
And why haven't central banks started to defect from this system? After all, as Eichengreen (and, yes, Roubini and Setser) argued way back in 2004, every central bank would rather another central bank assume the burden of financing the US.
It certainly seems like the incentives to defect are a bit weaker than many of us initially thought. Why is that?
- The rise in oil prices let the Asian countries who seemed most likely to defect (Korea, Thailand, India) off the hook in 2005. A big part of Asia did leave Bretton Woods 2 for much of 2005 (before jumping back in now)
- The oil exporters are a bit newer to this game. It will take them a bit longer before they have gorged themselves on dollar-denominated bonds. They haven't forgotten 1998. They would rather budget for $30 oil and see $60 oil than budget for $40 oil and see $30 oil.
- So long as China can credibly commit to hold the RMB down v. the dollar no matter what other emerging economies do, it can effectively punish any labor-rich emerging economy that lets its currency rise v. the RMB-dollar. That makes coordination inside the central bank reserve-accumulating cartel a lot easier.
- China's sterilization costs were far lower than anyone anticipated, so it has been relatively easy for China to commit to keeping the RMB down. At least $100b of China's 2005 reserve increase was unsterilized, and there was no uptick in inflation. And the PBoC earns more on its dollars that it pays on its sterilization bills.
- The benefits of fast export growth are visible and current. The cost of using the central banks' balance sheet to subsidize the US government and the US consumer are hidden and far off in the future. That reduces the pressure to defect - since there are not large, visible current costs to participating in the emerging markets dollar buying cartel.
- The dollar's rise v. the euro and the yen - and higher interest rates on dollars than euros or yens - made central banks more comfortable holding dollars in 2005, reducing pressure to defect.
I am sure there are other potential reasons why defection is costly as well.
But I still think it is striking that even with interest rate differentials that favor the dollar and stable G-3 exchange rates, emerging economies had to add $700 b (at annual rate) to keep their currencies from rising against the dollar (and the big three currencies more generally). January may prove to be atypical for 2006 - a wave of money flowed into emerging market funds, oil was high, etc.
But if the US slows and dollar interest rates start to fall, that number might get even higher. Someone is going to need to lend the US about $1 trillion this year --- and next year - even if the US economy slows modestly. Put differently, if Paul Krugman is right, and the US is still at risk of a hard (ish) landing, preventing a really hard landing will take a lot of coordinated firepower from emerging market central banks even as the costs of lending to the US become a bit more obvious.
Brad DeLong has argued (Via Mark Thoma and Martin Wolf) that central banks won't defect from the dollar-buying cartel even then - in part because Asian central banks are not independent from their political masters. They will instead buy more reserves than they can sterilize and end up importing inflation.
I still think that the presence of visible costs from the current system - like imported inflation - would prompt a wave of defection. The analogy to the end of Bretton Woods 1 isn't entirely off. Bretton Woods one didn't unravel all that quickly; France had plenty of time to convert its dollars into gold before the dollar plummeted.
There is another possibility too, at least for a few emerging economies. Not all emerging market reserve accumulation stems from large current account surpluses. Some also comes as the central banks of key emerging economies try to prevent capital inflows from the US and Europe from leading their currencies to appreciate.
The resulting exchange has characteristics of a swap between foreign investors and the central banks of many emerging economies. Foreign investors get the return on emerging market local currency debt or local equities. The central bank gets the return on US treasuries or German bunds.
FDI comes into China looking for a 15% return. The central bank gets 4.5%. Some folks in China think that is a bad trade. They would prefer an equity swap rather than a FDI for US treasuries swap. Foreign investment in China could, for example, be swapped for Chinese investment in oil companies. That point of tension remains.
And some emerging economies - Brazil and Turkey most obviously - pay far higher interest rates on their local currency debt than they get on their reserves. That means sterilization has an immediate costs. Not just a hidden future cost should the dollar fall against their currencies. I suspect that too will emerge as a point of pressure.
But facts are facts. Michael Dooley has now been right for 31 months, not 28.