Brad Setser

Follow the Money

Cross border flows, with a bit of macroeconomics

Print Print Cite Cite
Style: MLA APA Chicago Close


China v US money market funds

by Brad Setser
March 26, 2009

China’s purchases of US Treasuries in 2008 (Setser/ Pandey estimate): $245 billion

US money market funds purchases of US Treasuries in 2008, from the flow of funds: just under $400 billion

China’s purchases of US Agencies in 2008 (Setser/ Pandey estimate): $38 billion. That reflects $85 billion in purchases through July, and $47 billion in sales since then …

US money market fund purchases of US Agency bonds: $542 billion

China’s purchases of Treasuries and Agencies in 2008: $283 billion
US money market funds’ purchase of Treasuries and Agencies in 2008: $942b

I am waiting for a round of stories pondering whether money market funds will continue to buy Treasuries and Agencies at their 2008 pace!

US money market funds holdings of Treasuries and Agencies rose by close to 350% in 2008, as their combined Treasury / Agency portfolio rose from from $392b to $1334b. That pace of growth of growth won’t be sustained. The large rise came from a low base.

But money market funds did hold more Treasuries and Agencies ($1357b) at the end of 2008 than China ($1233b) did.

And we already know, more or less, that China’s purchases of US bonds are likely to fade a bit over the course of 2009. At least the purchases by China’s government, as China’s reserve growth has slowed dramatically. China’s wasn’t buying enormous quantities of Treasuries in late 2008 because the US was issuing more. It was buying enormous quantities of Treasuries because it stopped buying other US assets. At some point, the reallocation of China’s portfolio toward Treasuries will eventually run its course – and China’s Treasury purchases will start to track its reserve growth.

Treasury purchases by other central banks will also fall; the oil exporters are running down their reserves to make up for the fall in oil prices and private capital is leaving emerging economies. The central bank bid ultimately tracks global reserve growth, and global reserve growth is down. That eventually implies fewer Treasury purchases.

What we don’t know is whether American investors will continue to buy large quantities of Treasuries at something like current yields.

Obviously large the end of large Chinese purchases – or net Treasury sales by China — could change the equilibrium. China never accounted for more than 10% of the outstanding stock of Agency bonds, and its sales there had an impact. Its share of the Treasury market is currently larger.

But so too could an end to the surge in Treasury demand from US investors …

The data for China’s purchases comes from Setser and Pandey’s updated estimates for China’s portfolio. We smooth the survey adjustments over the course of the year, but these estimates are based on the US TIC and survey data. Our methodology is explained here; we have updated our estimates thought to reflect the latest survey data.
The data for money market funds come from table L121 of the Fed’s flow of funds.


  • Posted by John McLeod

    Good to know that China’s holdings of US obligations aren’t terribly critical in the larger scheme of things …

    Been wondering all day what you thought of Tim’s recent visit to your office. I can’t imagine you were away for that one 😉

  • Posted by bsetser

    Mr. Geithner’s comments today were far more significant, and I applaud his call to expand the regulatory net/ push derivative trading onto organized exchanges. I also thought his supposed gaffe on China wasn’t really a gaffe. It was clear that Geithner was talking about the portions of Zhou’s paper that discuss expanding the existing role of the SDR, i.e. within the context of the IMF (and there the US and China may well find common ground; Geithner clearly wants to increase the financing available to the IMF) not to the portions of Zhou’s paper that refer to the creation of a new supranational currency.

  • Posted by Simon

    Timmy if he’s the guy you mean …He so reminds me of that guy in Terminator Two. Was it Robert Patric as T1000.

  • Posted by Devin Finbarr

    What we don’t know is whether American investors will continue to buy large quantities of Treasuries at something like current yields.

    The appetite is infinite. When the government deficit spends, the beneficiary of the government’s spending gets dollars created out of thin air (the Fed just moves a zero on a spreadsheet). The beneficiary then must decide where to put that money. More likely than not, he puts it in a money market fund. A money market fund that holds 30-day treasury notes paying 1-2% interest has zero risk, but will pay better yield than just holding dollars. But the money market fund must buy treasuries to back this new investor. And lo, the Fed has treasuries it needs to sell, to offset the money it just created.

    The bottom line is that treasury bills are just dollars with a “not valid until date” on them. Thus questioning whether demand for treasuries will start to dry up is equivalent to asking whether demand for dollars will start to dry up. And the answer to this is clearly no, because the deflationary pressures are still stronger than the inflationary pressures.

    The receiver must It is always a dominant strategy to put the dollar bills in a money market account that holds 30-day reAt the same time, the government sells a bond.

  • Posted by Rajesh

    Money Market funds have not been traditional holders of Treasuries (because of the low yield.) However, many funds moved into Treasuries in response to the concern for safety when the Prime Reserve fund “broke the buck.” At some point, we would expect that the money funds would return to their traditional pattern of holding Certificates of Deposit and Commercial Paper. Hopefully, when that time comes, the Treasury will have less need to finance government spending (primarily due to an increase in government revenues.)

  • Posted by Off the boil


    What was the total treasury issuance in 2008 ?

    If you are saying MMs won’t be buying treasuries anymore, then how come 2009 auctions are getting covered ? There is going to be 2 trillion auctioned in 2009

  • Posted by Off the boil

    whether demand for dollars will start to dry up. — YES

    whether demand for treasuries will start to dry up — HELL, YES

  • Posted by Off the boil

    MM funds would cringe at this news :
    1 mo T-Bill went negative !

  • Posted by WStroupe

    Off the Boil,

    Yeah, and what is the official foreign demand, and the private foreign demand, for U.S. Treasuries?

    Those TIC reports for March 2009 forward might get real interesting – guess we’ll have to wait for May 15 to see.

    Devin Finbarr,

    When does the Treasury and Fed’s ‘perpetual motion machine’ cease to be perpetual? When the dollar’s debasement/spiking inflation reaches ‘critical mass’, where inflation runs away with the game? Or does it happen before that? What markers should we look for?

  • Posted by bsetser

    off the boil — counting treasuries sold off the fed’s balance sheet (i.e. the increase in “marketable treasuries” in the hands of the market), issuance was around $1.6 trillion. I would need to look up the number excluding the fed’s sale of its treasuries (and here i count the increase in treasuries loaned to market participants), but if memory serves it is close to $1.2 trillion. tis easy to find by looking at the monthly statement of the public debt (google it) for dec 07 and dec 08.

  • Posted by Indian Investor

    In the short/medium term (1-5 years from now) it’s likely that the Treasury will stay solvent. Beyond that nobody can be sure, it depends a lot on the actions taken meanwhile. US Treasury has approximately $11 trillion debt (including intra governmental holdings), it had an income of $2.7 trillion plus and expenses of $2.9 trillion plus in 2008, according to my memory. If the dollar loses its position as the world’s major forex reserve currency, you have to reason only with the underlying financials of the Treasury. This year Treasury will borrow at least 1 trillion more, making the total public debt $12 trillion+. If current trends continue, I expect that the Treasury will end up with a total debt of $15 trillion or so, and without a reserve currency status for the dollar.
    At that point in time, investors will perhaps not buy Treasuries any more, because the expectation from the Treasury will be that they should try to be cash positive month on month by having lesser expenses than tax revenues.
    Right now there’s no choice but to stimulate the economy fiscally with more borrowings.But US Taxes will need to be hiked pretty soon, addition of further unfunded liabilities needs to be reduced, and a drastic crackdown on military spending, and any other item they can save on would be needed.
    I don’t know if the relevance of my rather simple reasoning is clear. The question is not who holds how much of Treasury debt. In the medium term, the question will be whether the Treasury debt will be subscribed by anybody at all, and that’s not as far away as most people seem to imagine right now.

  • Posted by CB

    Brad, I’m wondering what you make of the very short maturities of the mmf holdings (given their avg duration/WAM is <3months). With the quant ease and effective guarantee of major banks, i doubt if many mmfs will roll over these T/agency holding. I suppose the gap between LIBOR and bills/agency just need to narrow toward 25-50bps?
    from a debt mgmt perspective, maybe the treasury needs to decrease this short term bill funding

  • Posted by Rajesh

    Off the boil,

    Money market funds are still buying a larger amount of Treasury bills than they usually do. However, this ‘flight to safety’ is likely a temporary change, since it lowers the return of the fund (4-week T-bills were bid -0.01 at latest auction). At some point, investors will feel more comfortable investing in banks and investment grade companies over a six month horizon and the mix will return to a more aggressive set of assets.

    It is not safe to assume that money market funds can absorb a large part of the Treasury short term issuance over an extended period of time.

  • Posted by bsetser

    CB — You are correct to highlight this risk. Presumably though if Money market funds start buying bank debt, the banks can pay back the fed — freeing up balance sheet space for the fed to buy yet more treasuries?

    But it is absolutely true that the surge in demand (foreign and domestic) last year was almost all directed at the front end of the curve, and there is a risk that the US will pile up too many s-term obligations.

  • Posted by Ruth Harris

    Saw you on Canada’s TV business news network last night. So there’s some overlap in your audiences! Thought you did a great job of explaining China’s comments about its US Treasury holdings.

  • Posted by The Wall Street Examiner

    “Treasury purchases by other central banks will also fall; the oil exporters are running down their reserves to make up for the fall in oil prices and private capital is leaving emerging economies. The central bank bid ultimately tracks global reserve growth, and global reserve growth is down. That eventually implies fewer Treasury purchases.”

    This is a theme that I first proposed in my reports a couple of years ago, and I believe we are at that inflection point now. I had felt that as the US economy weakened and dollar outflows fell and the domestic requirements of our foreign benefactors increased, that Treasury yields would begin to rise. That rise hasn’t happened, thanks to the buying panic you pointed out, as manifested in the activities of money market funds.

    Then last week, faced with the threat of a failed Treasury auction, the Fed stepped in and struck preemptively with its promise to buy $300B of Treasuries, add another $100 billion of GSE paper on top of the $100 billion already committed, and another $750 billion of MBS paper.

    That’s a lot of new dough, and should replace some but not all of the FCB buying that’s likely to disappear. They have already been dumping billions of their GSE paper holdings weekly. Their Treasury holdings are next, as the Fed has stepped up and promised them a ready market. So whatever buying the Fed does is likely to be offset by FCB dumping.

    Another problem is that the Treasury is slated to borrow nearly $500 billion in new money in the next 3 months alone. Sooner or later all buying panics flame out, and the panic into money market funds (mostly institutional funds) has already subsided over the past month. Therefore the panic levels of buying of short term bills will also subside, just as the supply of Treasuries continue to mushroom. When investors get zero return, they are forced to liquidate their holdings to meet current obligations. A zero interest rate environment is actually deflationary for that reason.

    The Fed’s pursuit of a ZIRP and quantitative easing will have disastrous effects. Push will come to shove in the very near future. The Fed may be the only buyer left in the market. What then?

    ‘Tis going to be ugly.

    Lee Adler

  • Posted by charlie

    I think China will continue to buy. They can effectively print money, exchange it for USD, and buy our debt. Since they run a large trade surplus, they can do this and not have to worry too much about killing their currency. It also serves the purpose of keeping their currency from appreciating too much. To me this is a no brainer.

    I doubt the domestic treasury market will hold up. Americans won’t continue to save at our current rate. It’s not in our blood. There was a big scare that caused a spike in saving, but as the threat goes away or the savings cushion becomes large enough, we’ll cut back on saving. I think it will be a gradual withdrawal.

  • Posted by Diego Montalbon, raconteur

    Vanguard CLOSED their Treasury Money Market Fund a few weeks ago and does not even let the remaining Treasury Money Market Funds be used for clearing brokerage transactions ! I don’t think they are profitable enough to justify anymore at current rates !

  • Posted by jonathan

    Just had to post this line, which I’m sure you’ve seen, from the PBOC:

    “Regulators do not have a good understanding of the cross-border activities of internationally active financial institutions. In particular, there is a lack of understanding of international capital flows.”

    That had to make you smile.

  • Posted by cmc313

    Dr. Sester:

    As the economy improves, money markets will shift some investment from the safety of treasuries into riskier assets like commercial paper, which the Fed has been supporting (A1/P1 papers only). As such, MM shifting to riskier assets will help to shrink the Fed’s balance sheet on one hand, but the Fed may also need to buy more treasuries on the other hand to keep rates down. There will always be plenty of demand for Treasuries — the question is at what interest rates. My fear is that at the first sign of sustainable economic recovery, rates will spike as investors anticipating Fed’s draining of liquidity. As such, financing cost for the US will go thru the roof.

  • Posted by Rien Huizer


    Money market funds must work with a minimum of interest rate risk and be very liquid. Momentarily that means buying treasuries (liquidity, zero default risk) and fairly short maturities as well ( in order to stay within the reach of exchange traded interest rate derivatives to optimize the fund’s repricing structure). If there is a cheaper (i e higher yielding) alternative that keeps funds in the fund, it will be used. For the time being, with all the uncertainty re bank resilence, that does not look ready for a change, and the banks need more capital to be able to absorb more deposits. So that means that all the money remains stuck near the shortest, safest end.

    I would expect (based on orthodox economics logic) that the stimulus measures in the US -absent friction created in the virtual US countries like China and OPEC- would boost inflationary expectations (be it from a deflationary level) and thus contribute to a significant interest premium over short term rates (meaning here only the component not affected by the default term structure, which is irrelevant for government securities). But there is not a lot to be seen. Some premium over the long term rate of inflation would result in 10 yr treasuries with 5 to 6 yields. Yields like that are easily available in the corporate market (from firms unlikely to default within the horizon of the current crisis, at worst until 20013), so is there perhaps a simple shortage of liquidity, even in the presence of government minificence? On of the practical problems of the otherwise beautiful school of monetarism (no input from politicians desired or required) is that no one knows what money is.

  • Posted by martin


    Nice work. Any color detailed history of the U.S.-China trade relationship? Does council have any documents or prior research on the main bullets of this unique/interesting relationship.

    I’d like to review it…Thanks.

  • Posted by Judy Yeo


    whether money market funds continue purchases can largely be answered by consideration of whether the Obama/Geithner plans are throwing good money after bad or truly solving the problem, the idea of”bad bank” /sale of distressed assets isn’t quite new and yves smith’s site has dealt with the problems of such a plan previously.

    what do you think?

  • Posted by Indian Investor

    Brad: It was clear that Geithner was talking about the portions of Zhou’s paper that discuss expanding the existing role of the SDR, i.e. within the context of the IMF

    Me: It’s difficult to see where Zhou is talking about expanding the role of the SDR in its present form. What Zhou is talking about is the “inclusion of currencies of all major economies” in the SDR, and allocation of SDRs “according to the current mix of forex reserves”, plus “wieghted by GDP”. (I’m not quoting exact words but very close) In its current form SDRs are just a different form of the dollar hegemony. The current SDRs are marked to a basket of USD, EUR and JPY, only. Given the recent demonstration that the Fed, ECB and Bank of Japan can come together to print currency in a co-ordinated way, all the more reason for Zhou not to accept this hegemony.
    Zhou clearly states in his paper that “calculation-based” allocation of SDRs isn’t a good thing. The current system of quotas reflects the ‘traditional industrial countries’ wealth’ – a euphemism for the major participants in World War II – in the proportion that it was at at the end of World War II.

  • Posted by Twofish

    Yeo: whether money market funds continue purchases can largely be answered by consideration of whether the Obama/Geithner plans are throwing good money after bad or truly solving the problem, the idea of”bad bank” /sale of distressed assets isn’t quite new and yves smith’s site has dealt with the problems of such a plan previously. what do you think?

    I think that Yves fundamentally misunderstands what Citigroup is trying to do with the good/bad bank. The reason for this separation is so that you can treat the bad bank differently from the good bank. If you want to let the bad bank fail without killing the good bank, you can.

    I think the big disagreement is whether or not the banks are solvent or not. If we are not looking at a second great depression, most banks *are* pretty clearly solvent, and so policy solutions based on the premise that they are not will do more harm than good.

    Part of the reason that commercial banks are solvent now is even in the crazy period, most of their operations were still pretty tightly and effectively regulated so the situation with most commercial banks is not quite as bad as with the really crazy stuff.

  • Posted by HKInvestor

    Will the US investors continue to buy treasuries? Yes, of course! Where else will all the savings go? I don’t think there are much debate that the savings rate has gone up and will go higher. These savings will need to be channeled somewhere. Why not treasuries since people have been burnt on equity and debt already? Where else can be put their money with safety in the return of principal?