Monday, July 25, 2011

. Monday, July 25, 2011

The chart illustrates ownership of US government debt at the end of December 2010. US ownership is subdivided by category, foreign ownership by country. The foreign data are not broken down in to categories, but TIC indicates that 3/4 of foreign-owned US government debt is held by public authorities. The domestic data come from the June 2011 Treasury Bulletin and the foreign from the Treasury TIC. More recent data on foreign ownership exists. I could not find more recent data for US ownership. If I have mis-interpreted this data, someone please point this out.

By these figures, about 63% of US government debt is owned by central banks (foreign and domestic) and/sovereign wealth funds. Most of these entities are American friends and allies. Another 4% is owned by US state and local governments. That leaves 33%--about $4.8 trillion--in private hands. Of this, the financial institutions with the most restrictive regulations regarding asset ownership (depository institutions) own only 2% of the total ($290 billion). Mutual Funds, who may or may not have to dump downgraded debt, hold another 9% ($1.35 trillion).

What's the point? The discussion about the impact of US default revolves around the market response to default. Useful to recognize that most of the US government debt is held by public-sector agents who are much less sensitive to balance sheet pressures and regulatory constraints. These public sector agents are also substantially more sensitive to "moral suasion" and direct appeal than private financial institutions. The structure of ownership of US debt might dampen the negative impact of any default that does occur.


Kindred Winecoff said...

Thomas I don't find this the least bit reassuring. While depository institutions, pension funds, and mutual funds are < 20% of the total, that's a) still quite a lot; b) doesn't count the 40% in "other US". I'm not sure quite how that breaks down, but a decent chunk of that has to be in hedge funds and nondepository financial firms, not to mention GSEs. The relevant table (OFS-2) makes no mention of bank holding corporations or what used to be investment banks. Ie, it's not clear whether they're included in "depository institutions" or "other".

In any case, Treasuries are often used as collateral for other sort of bonds, like corporate and municipal bonds, and I remember what happened the last time repo markets froze up. I also remember what happened when hedge funds collapsed and money market funds broke the buck. I'm other words, I'm not worried about the losses per se; I'm worried about the knock-on effects.

To put it in more familiar terms, T-bills are the most central node in the network of financial assets. Attacking the central node is a bad idea.

The best way out of this may be for the Fed to promise to buy up, at par, any T-bill from any financial institution that wishes to sell. Substitute T-bills for cash, inject as much safety and liquidity into the system as necessary. QE4, many multiples the size of QE3. But can that be done quickly enough?

Thomas Oatley said...

I didn't think you would find it reassuring. But the parallel with last time seems a bit misleading. Last time there were something like $10 trillion of MBAs of uncertain value held entirely by over-leveraged financial institutions. This time we have less than $5 trillion that isn't of uncertain value held as capital by depository institutions and as assets by mutual funds, pension funds. They don't collateralize anything. The other US amounts to $1.2 trillion. If you think there really are parallels with last time, then you need to develop them much more fully than you have done. I don't see it. Could be I am blind to it.

Kindred Winecoff said...

Yes, but trillions more in other instruments -- state/municipal bonds, money market funds, GSE securities, etc. -- are also indexed to Treasuries. These are held by over-leveraged financial institutions, and are used as collateral (as are the Treasuries themselves). Hell, if a default prompts a sell off and drives up interest rates, that'll increase mortgage rates and put even more pressure on housing. Many institutional investors have to own AAA assets... if they have to sell T-bills, what are they going to buy? Cash?

Maybe I'm not convincing that the first-order effect is not the only or even primary concern. But I have to say that your straight-up dismissal of knock-on effects is even less convincing. As is your discussion of the intl dimension. Maybe the US can use "moral suasion" to get foreign authorities to be nice this time, but such an event could lead to the reorganization of the intl monetary/financial system as a way of hedging future risk.

So the *best case* scenario is that default costs the US billions going forward (the 1979 result). The worst case scenario is much worse than even 2008. That seems like it's worth being very worried about.





Add to Technorati Favorites