Yes, China probably bought close to $400 billion of Treasuries too. My top secret model says China bought exactly $374.571 billion of Treasuries in 2008, a record. China certainly bought far more Treasuries in 2008 than in 2007. My model, which accounts for flows through London, suggests that China added $120.3 billion to its Treasury portfolio in 2007.
But the big surge in demand for Treasuries in 2008 didn’t come from China. Other investors increased their holdings of marketable Treasuries by $1310 billion. That is a huge increase from the (estimated) $127 billion increase in their holdings of marketable Treasuries in 2007.
It stands to reason that investors should be debating whether this surge in non-Chinese demand can continue, not whether China will keep on buying Treasuries.
Relatively speaking, the big change in 2008 was the emergence of non-Chinese demand for Treasuries. And not all of that demand came from central banks either.
My best guess is that central banks bought about $650 billion of Treasuries in 2008, up from about $290 billion in 2007. $650 billion is a record by the way. It is also far more than the TIC data indicates, as I am adjusting the TIC data upwards to reflect the fact that the TIC flow data tends to understate official purchases.* It is also a bit more than the roughly $500 billion increase in central banks custodial holdings at the New York Fed. I am not trying to understate the impact of central banks on the market.
But given the scale of new issuance by the Treasury (The Treasury issued $1257 billion in marketable Treasuries) and the scale of the fall in the Fed’s holdings of Treasuries (down $427 billion, counting the securities the Fed has lent out to the market), private investors added over a trillion dollars to their Treasury holdings in 2008 — more than the world’s central banks.
That is a huge change from 2007. In 2007, best I can tell, private investors were net sellers of Treasuries, as central bank purchases exceeded the net issuance of marketable Treasuries ($194 billion) and the Treasuries the Fed sold into the market ($54 billion).
Moreover, we more or less know that foreign central bank demand for Treasuries will fall over the course of 2009. Commodity prices are down. Capital inflows to emerging economies have turned into capital outflows. Reserve growth stopped in the fourth quarter. Unless something changes, that implies fewer central bank purchases of Treasuries over time. Right now central bank demand for Treasuries is being driven by a reallocation of their existing portfolio towards safety and liquidity, not reserve growth. That process likely has further to go, but it won’t last forever.
To put it most simply, record central bank purchases of Treasuries reflected record central bank reserve growth. Record Chinese purchases were also a reflection of record Chinese reserve growth. And reserve growth has slowed dramatically.
The falloff in Chinese reserve growth reflects a rise in private capital outflow from China, not a fall in China’s current account surplus. Those private outflows have to go somewhere — and most likely make there way into the US market in one way or another. But China’s 2009 current account surplus will be dwarfed by the United States’ 2009 fiscal deficit. Ultimately, the majority of the 2009 fiscal deficit and the broader US borrowing need will have to be financed by private investors, not by China.
Strange as it seems, more Treasury issuance doesn’t necessarily imply greater reliance on China to finance the deficit.
This has been a hard case to make, so let me go through a lot of detail.
In 2007, my best estimate is that China accounted for $120.3b of the $247.2b increase in the outstanding stock of marketable Treasuries not held by the Fed. China absorbed 49% of the net increase.
In 2008, my best estimate** is that China bought $374.6 billion of the $1684.8 billion increase in the outstanding stock of marketable Treasuries not held by the Fed. China absorbed 22.2% of net issuance. Foreign central banks share of the total outstanding stock of marketable Treasuries not in the hands of the Fed also went down, falling from 47.2% at the end of 2007 to 44.5% at the end of 2008.***
Let’s drill down even further.
In the first quarter of 2008, China added an estimated $38 billion to its Treasury holdings (it was buying a lot of Agencies then) while the stock of Treasuries not held by the Fed increased by $430 billion (The Fed reduced its Treasury holdings to help manage the fallout from Bear). In the second quarter, China bought $60.5 billion of Treasuries and the stock of marketable Treasuries not held by the Fed rose by $87.6 billion. In the third quarter, China added $113.5 billion to its Treasury portfolio but the US government increased the stock of marketable Treasuries in circulation by a stunning $681.6 billion. And in the fourth quarter, China increased its Treasury purchases to an estimated (especially estimated, as I had to infer the December data) $162.5 billion while the stock of outstanding marketable Treasuries increased by $485.8 billion.
That implies that China bought a (stunning) $276 billion in Treasuries in the second half of 2008.
And it also implies that non-Chinese investors bought an (even more stunning) $892 billion of Treasuries in the second half of 2008.
Obviously, it would be a big deal if China stopped buying and started selling.**** But it would be much bigger deal if private investors lost their appetite for Treasuries.
* My methodology is simple. I use the pattern of past revisions to the US data (the data is revised after the annual survey) to reattribute some foreign private purchases to central banks. In practice, this means that purchases by private investors in the UK are reassigned to central banks, and to China in particular. My estimates are meant to anticipate the revisions that will come with the survey data.
** The TIC data for December isn’t out. I used China’s average purchases over the last three months of data to estimate China’s December purchases (roughly $50 billion)
*** Here i am using my estimate of official holdings of Treasuries. That estimate is higher than the data the Treasury reports on its web site, as I try to account for purchases through the UK.
**** The context matters. If China was selling because hot money outflows were cutting into China’s reserves, the private investors taking money out of China would need to invest in something — so the net effect on the Treasury market might be modest.