Happy New Year to all readers. I will not write much today (to reflect to on-going holiday spirit!). But, there was an article in Bloomberg media (December 30, 2016) – Beware the Foreign Exodus From Treasuries – stirring up fear about the recent sales of foreign-held US government debt. I guess it was a slow news day or something because there is very little in the story that is relevant to assessing whether the US government can run an appropriate fiscal policy stance. The fact is that the foreign sales of US government debt are largely irrelevant for the US government’s capacity to maintain its net spending program. The sales are in US dollars and only the US government itself issues those dollars. To think that a foreign purchaser of a US Treasury debt liability are ‘providing dollars’ to the US government is to completely misunderstand the nature of the transaction. This blog considers the current data and explains how to think correctly about these matters. The question that financial commentators really should be asking is why should the US government extend that corporate welfare (risk-free bonds with income flow) to domestic bond-buyers and foreign governments/private investors. There is no financial reason (in terms of facilitating fiscal policy) for the bond issuance. It is just a form of welfare spending which helps the top-end-of-town.
The gist of the Bloomberg article is as follows:
1. “The biggest foreign buyers of U.S. government bonds are quickly retreating after years of absorbing record amounts of the securities”.
2. “This is an important dynamic to understand when looking at the potential fate of the $13.6 trillion Treasury market in 2017.”
3. “It’s hard to see how these bonds can significantly rally, even in the face of bad news, given the exodus of foreign central banks from the U.S. government-debt market”.
4. “it’s not just the big governments that are selling. Private foreign investors have been net sellers recently, too”.
5. “This foreign withdrawal will most likely continue for the foreseeable future”.
6. “there’s a solid argument to be made that the latest selloff has gone too far and will correct soon enough”.
7. “But this backdrop of foreign selling is important. Yields on 10-year Treasuries may not quickly shoot up, but they most likely won’t plunge back to the lows of 1.36 percent seen in July.”
The point to understand is that while the trends described in the article are actually happening the importance of those trends is close to zero when assessing the capacity of the US government to run its fiscal policy program, which, might include a large increase in public infrastructure spending under the new administration.
As we see below, the foreign purchases of US government cannot be construed as funding the US government. If that government chooses it can keep spending with zero foreign government debt purchases.
Here is an update on some research I did in this regard a few years ago which appeared in these blogs:
There are a few data sets that you can pull together to break the total US public debt outstanding down into various categories:
- The US Treasury’s Bulletin Chapter on Ownership of Federal Securities, which provides a detailed breakdown of who holds the outstanding US government debt.
- US Treasury Foreign breakdown.
- US Federal Reserve Bank’s Consolidated Balance Sheet data.
- US Federak Reserve Bank’s Financial Accounts of the United States.
So what has been going on?
The following pie-chart shows the proportions of total US Public Debt held by various categories as at the March-quarter 2011.
The government sector held about 42 per cent of its own debt. These holdings were either in the intergovernmental agencies or the US Federal Reserve Bank.
The US Federal Reserve held 8.9 per cent of total US public debt. Its total holdings were around $US 1,274.3 billion.
The three largest foreign US debt holders at the March-quarter 2011 were China (8 per cent), Japan (6.4 per cent) and Britain (2.3 per cent). The total foreign held share was equal to 31.4 per cent.
Now, to the most recent data – September-quarter 2016.
As at the end of September 2016, there were $US19,597.4 billion Federal Securities outstanding. These were broken down into:
1. Privately held – $US11,709.9 billion (59.8 per cent of total).
2. Federal Reserve and Intragovernmental holdings (SOMA and Intragovernmental Holdings) – $US7,863.5 billion (40.2 per cent of total).
3. Foreign and international – $US6,154.9 billion (31.4 per cent of total) – of which $US3,901.7 billion held by Foreign Official institutions.
A comparison of movements over the last 5 or so years in the composition of US Treasury debt holdings is provided by the following two pie charts
The following pie-chart shows the proportions of total US Public Debt held by various categories as at the September-quarter 2016.
The government sector held about 40.4 per cent of its own debt in the September-quarter 2016. This is slightly down on the proportions held in the March-quarter 2011.
These holdings were either in the intergovernmental agencies (27.6 per cent) or the US Federal Reserve Bank (12.6 per cent). So the central bank has increased its holdings over the period in question and now holds $US2,463.5 billion and increase of 93.4 per cent in overall levels.
The Chinese holdings were around 5.9 per cent of the total and hardly consistent with the rhetoric that China was bailing the US government out of bankruptcy. These holdings have fallen in recent years (see below).
The three largest foreign US debt holders at the September-quarter 2016, were China (5.9 per cent); Japan (5.8 per cent) and Ireland (1.4 per cent). The total foreign held share was equal to 31.4 per cent.
The impact of the British austerity is noticeable. In 2011, Britain was the third largest holder of US public debt (at 2.3 per cent of total US public debt). By March 2013, this had dropped to 0.95 per cent. They shed $US166 billion worth of US treasury securities in that 24 month period. As at the September-quarter 2016, the British share had consolidated at 1.1 per cent of the total.
The following Table shows the proportions of the total change in US federal government debt for the calendar years 2009 (so the change is between 2008-09 and so on) to 2015-16 (up to September 2016) accounted for by the main holders – Federal Government Accounts, Total non-government (private), which is, in turn, decomposed further into US Federal Reserve holdings. There are other components not included.
In 2011, the increases in the US Federal Reserve’s holdings accounted for 54.7 per cent of the total increase in US federal government debt. That fell away to 4.3 per cent in 2012 and then rose again to 84.9 per cent in 2013.
The proportion has been decreasing since then and in 2016, the proportion was only 2 per cent as the US Federal Reserve ended its asset-buying programs.
The recent history thus provides a very important lesson. The US central bank can initiate very dramatic shifts in the mix of debt holders whenever it chooses.
This means that the central bank can always purchase any debt that the private sector chooses not to purchase via the primary auctions.
There is never a reason for US government bond yields to rise above a level that the government considers to be acceptable.
To repeat, the US government (consolidated Treasury and central bank) can always assume the role of its own largest lender and borrow as much as it likes from itself (subject to Congressional ceilings etc).
That is how nonsensical these voluntary conventions are. They sound robust but in the end the currency-issuer is the currency-issuer and can always work around the so-called constraints.
The next graph shows the evolution from March 2001 to September 2016 of the US public debt by private, public and foreign holdings (%). The foreign holdings are a subset of the private series.
There are some interesting points to note. At a time when the US public debt ratio has risen beyond what the mainstream claim is the danger point (80 per cent) – the point after which we are repeatedly told a nation will struggle to ‘fund’ its outstanding public debt – the private demand for US public debt has risen.
Private markets know that US Treasury debt is risk free, despite all the scare mongering that the likes of the Peter Peterson Foundation and its ilk might regularly entertain.
The other point, in relation to the rising foreign share is that you cannot conclude that the foreigners (China, Japan etc) are ‘funding’ the US government. The US government is the only government that issues US currency so it is impossible for the Chinese to ‘fund’ US government spending.
To understand the trend shown in the graph more fully we need to appreciate that the rising proportion of foreign-held US public debt is a direct result of the trade patterns between the countries involved (and cross trade positions).
For example, China will automatically accumulate US-dollar denominated claims as a result of it running a current account surplus against the US. These claims are held within the US banking system somewhere and can manifest as US-dollar deposits or interest-bearing bonds. The difference is really immaterial to US government spending and in an accounting sense just involves adjustments in the banking system.
The accumulation of these US-dollar denominated assets is the ‘reward’ that the Chinese (or other foreigners) get for shipping real goods and services to the US (principally) in exchange for less real goods and services from the US. Given real living standards are based on access to real goods and services, you can work out who is on top (from a macroeconomic perspective).
Note that a worker in Detroit or some other degenerate industrial region in the US who is suffering from unemployment as a result of cheaper imports coming from nations with lower labour standards (pay and conditions) than the US is unlikely to agree with me. In his/her case I wouldn’t agree with me either.
But from the perspective of macroeconomics there is no issue.
The point is that the US Federal Reserve Bank (as in the case of other central banks around the world) have been taking up significant proportions of the outstanding government debt in the last 10 years. In some years, this proportion has been very high.
We didn’t observe any major inflationary spikes associated with these shifts in monetary operations underpinning the government debt-issuance (and deficit) outcomes.
Please read my blog – Direct central bank purchases of government debt – for more discussion on this point.
Here are some other observations that impact on how one interprets the Bloomberg article.
The following graph shows the total US Federal debt held by foreigners since 1970 (in $US billions). There was a dramatic escalation after 2000.
The same data expressed as a proportion of total US Federal debt is shown in the next graph since 1970. You can see that the downturn since 2015 has not been very dramatic in historical times. The US has witnessed variable shares of total debt held by foreigners over this long historical period.
The next two graphs provide some insights into how scale affects our assessment of graphical evidence.
The following graph comes from the Bloomberg article (under the heading Fire Sale) and shows the holdings of US Treasury debt (in US billions) since July 2010 up to October 2016.
The vertical scale chosen gives the ‘falling over the cliff’ impression.
The next graph shows the same data (in $US billions) from March 2000 to October 2016 without any adjustment to the vertical scale (other than it is linear).
These Chinese sell-offs have occurred in the past – and no-one really blinked an eyelid.
The next graph shows the monthly percentage change in Chinese holdings of US Federal Debt from March 2000 to October 2016. I adjusted the data to exclude the impacts of the series breaks which have occurred in the month of June (in several of the years shown).
Historically, the time series has been quite volatile and the current monthly reductions do no seem out of the ordinary.
As a proponent of Modern Monetary Theory (MMT), I do not consider there to be a public debt problem in the US so the analysis presented here is out of interest rather than establishing anything substantive.
However, what it shows is that even within the voluntarily-constrained system that regulates the relationship between the US treasury and the central bank, the latter can still effectively buy as much US government debt as it likes.
So if yields rise on US Treasury debt in the coming year it will be because the US government has chosen to allow higher interest payments on the corporate welfare it extends the non-government sector in the form of voluntary debt-issuance.
The question that financial commentators really should be asking is why should the US government extend that corporate welfare to domestic bond-buyers and foreign governments/private investors.
There is no financial reason (in terms of facilitating fiscal policy) for the bond issuance. It is just a form of welfare spending which helps the top-end-of-town.
That is enough for today!
(c) Copyright 2017 William Mitchell. All Rights Reserved.