Buying dollars to stem their currencies rise?
Selling euro to keep the dollar share of their portfolio constant as the value of their existing holdings of euro rises?
Or selling dollars to try to reduce their exposure to future falls in the dollar?
David Woo of Barclay’s capital notes that keeping a constant euro/ dollar portfolio share in the face of a 1% rise in the euro against the dollar requires central banks to sell $10b of their existing euro bonds … (assuming that central banks hold roughly $1 trillion in euros). Woo in Bloomberg:
“For every 1 percent the euro gains versus the dollar, central banks have to sell 10 million euros to keep the ratio of currencies in their reserves constant, he [Woo] said.”
That is right, but I would argue that central banks in Asia typically don’t sell their existing euros – they just keep the new reserves that they are buying in dollars. Remember, Asian reserve portfolios are still growing, as central banks in Asia intervene to smooth markets (and to keep their currencies from appreciating against the dollar).We should have a better sense of how much intervention took place soon, as many Asian central banks will report their November reserves soon. Bank of Korea usually reports first -- we now know its reserves grew by $4.8b in November.
If the Bank of Korea held 80% of its reserves in dollars and 20% in euros/ pounds, the 3.8% move in the euro explains about $1.3b of the increase – meaning that Bank bought around $3.5b in the market. If only 70% of Korea’s reserves are in dollars, valuation gains would explain $2.6b of the increase, implying intervention of around $2.1b.
I don’t think it is likely that only 60% of Korea’s reserves are in dollars, as the $3.5 b in implied valuation gains would imply only $1.3b in intervention – and the Bank of Korea reportedly bought around $1b in mid-November, and it presumably bought a bit more later in the month (it certainly intervened verbally). The Korean Times reports:
“Currency dealers said Korean authorities continued to buy dollars.”
Thailand also clearly has been intervening. India's foreign currency reserves rose by $5.7b in the first three weeks of November and even if (as is likely) 50% of India's reserves are in euros and pounds, valuation gains only explain $2.4b of that -- implying $3.3b of intervention in three weeks, or an average of $1 billion a week.
Brazil isn't in Asia, but its reserves are up $4.9b in November, far more than can be explained by valuation changes. And so on.
As Stephen King of HSBC notes (Stephen Jen has made a similar point), it is hard to reconcile stories of central bank diversification with the current low level of US bond yields. 4.43 on the ten year suggests there are lots of buyers, and not so many sellers. King:
One [explanation for dollar weakness] doing the rounds is the idea that the dollar is softening because emerging market central banks, awash with cash as a result of persistently large balance of payments current account surpluses, are diversifying out of dollars into euros, yen, sterling and anything else that isn't a dollar.
But, for me, this argument doesn't really stack up: emerging market central banks tend to buy Treasuries with their dollars so diversification out of dollars should also have meant the sale of Treasuries. Yet Treasuries have strongly rallied.
Given strong central bank demand for agencies this year, I would amend Stephen’s King’s point so that it reads central banks tend to buy “dollar bonds” not just Treasuries.
And I rather suspect that the rapid growth in central bank dollar deposits this year has contributed to continued strong demand for credit risk. Hedge funds looking to gear up to bet on credit spreads can borrow the dollars central banks have on deposit in the world’s banks. And pension funds who don’t find Treasury yields attractive (in part because of central bank demand) are forced to bid on spread products.
Justin Lahart’s column today pointed out that low junk bond spreads are a bit inconsistent with low Treasury yields and an inverted curve. But his point could be framed another way – there has been strong demand for bonds of all kinds, so all kinds of bonds have been bid up. I suspect the rapid growth in official assets has something to do with that – as the official sector buys bonds rather than equities (setting a few oil investment funds in the smaller gulf countries aside). And by my estimates, central banks are on track to add $500b or so to their dollar portfolios this year – and that total leaves out demand for dollars from Chinese state banks and oil investment funds.
That isn’t to say that central banks didn’t diversify their portfolios away from the dollar earlier in the year. Some – Russia most notably – clearly did. I suspect the dollar’s share in Korea’s portfolio also has been falling, though the November data suggests that its dollar share hasn’t fallen quite as much as I might have thought.
However, it is one thing to diversify when the market is calm and there is private demand for dollars; it would be quite another for Asian (and other) central banks to move away from dollars when the dollar is under pressure -- as of yet, that doesn’t seem to be happening … thought the data is also, as of yet, incomplete.
If central banks do decide to diversify in a destabilizing way, I suspect it will come in a two-step process.
Central banks would first add dollars to their portfolio to try to keep their currencies from moving …
And then, as they see their stock of dollars continue to rise, they might start to get scared, and start to reevaluate a policy regime that forces them to buy dollars when others don’t want dollars.
Update: An article from the FT on Asian central bank policy, though not one that I found particularly good or informative. It notes that Korea and Thailand have intervened heavily recently, while Japan has not. That's true, but it is in large part because the won and baht has appreciated significantly v. the dollar over the course of 2006 while the yen has not. The yen was quite weak before its recent (small) appreciation; the baht and won were not weak -- so both Thailand and Korea have more reason to be concerned about additional appreciation. And the article doesn't emphasize a key point -- China is continuously intervening to keep its currency from appreciating, and since the RMB has moved by less against the dollar in November than the dollar has fallen v the euro and many others, the RMB has depreciated in real terms. Concerns about competition with China is a big reason why those who have already let their currency appreciate (Thailand, Korea) are resisting furhter appreciation.