- Blog Post
- Blog posts represent the views of CFR fellows and staff and not those of CFR, which takes no institutional positions.
That, by the way, would be a bit of a blow for Nouriel Roubini and myself -- not just Dooley, Garber, Folkerts-Landau. Nouriel and I argue the Bretton Woods 2 of central bank financing of the US current account deficit will come to an end sooner rather than later. But we also argue that the end of central bank dollar reserve accumulation will have a significant impact on financial markets.
We don’t know China’s January reserve accumulation (yet), but reserve accumulation by ten other emerging economies (The Asian NICs, Thailand, Malaysia, India, Russia, Brazil, Mexico) is way down in January. After an average increase of over $35 billion a month in new reserve accumulation during the course of q4, the reserve of ten of the largest emerging economies increased by less than $2 billion in January. No doubt, the fall in the euro, which reduced the dollar value of their euro reserves, masked a slightly higher pace of new reserve accumulation. Still, there was a signficant fall in the pace of their reserve accumulation.
My provocative headline (for some) aside, this is a long, geeky and data intensive post -- my apologies. This is important stuff, but probably not of general interest.
Let’s assume that China’s new reserve accumulation in January was $20 billion -- A bit slower than in q4, but still a $240b annual pace. Shifts in the euro/dollar subtracted $8 billion from china’s reserves -- so total reserves went up by $12b. Suppose the other ten countries also lost $8b because of shifts in the euro/dollar; their underlying pace of reserve accumulation then would have been around $10b.
Then reserve accumulation by eleven of the biggest emerging economies -- countries whose central banks hold a bit under 50% of all central bank reserves and over 60% of non-Japanese reserves -- totalled about $30 billion in January (with $16b in offsetting valuation losses). That is well off the average pace of monthly pace of $68 billion of q4, but that pace includes valuation gains from the rising dollar value of these countries euro reserves. Net out valuation gains, and reserve accumulation by these 11 central banks was probably about $60 billion a month in the fourth quarter -- still well above my $30 billion estimate for January.
Yet, as we all know, long-term Treasury bonds rallied and Treasury yields fell in January. Not very good supporting evidence for thesis that Asian central bank demand has helped to keep Treasury yields low.
What is going on?
Let me put forward a couple of ideas.
1) It is pretty clear that there is a big gap between the known increase in central bank reserves, and the recorded increase in central bank holdings of US debt. The creditor-side data does not match the debtor-side data.
The following central bank data comes from 11 large emerging economy central banks.
October increase in reserves: $54.2 billion
Recorded official purchases of US securities: $14.9 b November increase in reserves: $81.6 b
Recorded official purchases: $27.9 b December increase in reserves: $69.2 b
Recorded official purchases: $10.3 b Total reserve increase in q4: $205 b
Recorded official purchases: $53 b Maybe Asian central banks, Russia and two big Latin countries were buying lots of euros and yen in q4. But I sort of doubt it. Some euros, sure. But not enough to explain the majority of the gap.
Large central bank purchases of euros would require large offsetting outflows from private European investors to make the global current account balance add up. Europe is not running a big current account deficit; the US is.
I think it is more likely that there was a fair amount of disguised central bank buying of dollar assets.
And it is also quite likely that central banks built up their bank accounts in q4 -- and probably more so in December than in October and November. If that is true, then central bank purchases of US debt securities may not fall off in January even though the pace of reserve accumulation clearly fell off. Look how long it took Japan to put all the dollars it bought at the end of 2003 and in early 2004 to work.
2) Reserve accumulation by these 11 banks tracks overall purchases of US debt surprisingly well in the fourth quarter.
October increase in reserves: $54.2 billion
Foreign purchase of US debt securities: $61 b (technical note: I included recorded central bank purchases of equities in the total as well; data from the TIC)
November increase in reserves: $81.6 b
Purchase of debt securities by foreigners: $87.7 b December increase in reserves: $69.2 b
Purchase of debt securities by foreigners: $76.4 b Total reserve increase in q4: $205 b
Purchase of debt securities by foreigners: $225 b That suggests to me that central banks did play a big role in the market in q4. Emerging market central banks like to hide their footprints, unlike the Bank of Japan, but they are still out there.
Obviously, there are some private purchases as well, and not all of the central bank’s increase in overall reserves went into the purchase of US securities. Some went into euros, some went into bank accounts. On the other hand, there are also more than 11 central banks in the world.
So score one for Bretton Woods 2.
3) $30b in monthly reserve accumulation is not $60b, but it is still a lot for 11 emerging market central banks.
Don’t be fooled by the fall in the stock of global reserves that is simply the product of the shift in the euro/ dollar. The underlying pace of reserve accumulation probably did not fall off quite as much as it seems at first glance in January. $30b a month works out to $360b a year. That is not chump change, particularly for a set of 11 emerging economies. There are some other central banks out there that are intervening as well, so global reserve accumulation would be higher. But at this pace, emerging economies alone would not make up for the absence of the Japanese from the market. Central banks added $600 b or so to their reserves in 2004 (excluding valuation gains); maintaining that pace in 2005 implies central banks need to add around $50b a month to their reserves.
Since the US Treasury pays interest in February, the global pace of reserve accumulation (assuming coupons payments are reinvested, not spent) may even pick up a bit next month ..
Still, there is little doubt that central bank intervention slowed in January, but both the dollar and the Treasury market rallied. Isn’t that evidence against the argument that Bretton Woods 2/ Asian central bank demand is playing a key role in the Treasury market?
Not necessarily. The pace of reserve accumulation in the fourth quarter was truly extraordinary -- $205b from 11 emerging economy central banks in a SINGLE Quarter ($840b annual pace), and no doubt oil exporters added to their public and private reserves in a big way too. That may have provided a bit of financing overhang, in much the same way that Japan’s heavy intervention in q1 of 2004 did at the beginning of last year.
Moreover, reserve accumulation does not need to provide the financing the US needs every month for the basic thesis that reserve accumulation is providing a key source of financing for the US to hold. It may matter more in some months than in others. In January, it seems quite likely that European (and maybe Canadian) private investors picked up the pace of their purchase of US assets -- presumably because they figured the dollar had fallen about as much as it was going too. The repatriation of corporate profits previously held in say euros to exploit the US tax holiday no doubt generated some flows as well. Pension funds are being pushed to hold more long-term bonds, creating some "real money" demand for longer-term Treasuries.
It is too early to declare the end of Bretton Woods 2.
On the other hand, if the January fall off in the pace of reserve accumulation is sustained for the full year, or even for several quarters, and such a fall off has no impact on US debt markets, that would be a blow to the Bretton Woods 2 thesis.