There will be a lot written about financial deglobalization when folks pour over the 2018 data. But it is a mistake to fit last year's financial deglobalization into a Trump trade driven narrative.
It is basically a function of the shift in U.S. tax policy.
(thread)
1/x
The fall in U.S. outward FDI is entirely a function of a fall in U.S. direct investment in the world's tax havens; there was not real change in the pattern of investment in other economies.
(under the old law profits reinvested abroad could defer paying US tax)
2/x
The fall in U.S. FDI "reinvested" abroad in low havens had a host of other effects - firms building up assets in low tax jurisdictions were buying U.S. debt, inflating gross flows in both ways.
(there is actually a good fit in the BoP data here,using flows over last 4qs)
3/x
E.g. a lot of US FDI abroad was in practice the rising "cash" of a Techco (Ireland or Bermuda) sub, and a lot of foreign demand for US debt was coming from the same Techcos (or Pharmacos) offshore subs
4/x
I think I have found this in the BoP - the fall in cumulative FDI in a set of tax havens was mirrored by a fall in the cumulative purchases of U.S. debt of a slightly different set of tax havens
(cumulative flows = proxy for the stock" of offshore claims)
5/x
The match here isn't "pure." The debt holdings line for example includes Russia (which moved its reserves out of the US). But other Europe is the breakdown in the US data alas. & I couldn't include the Caribbean's holdings of U.S. debt b/c that was picking up something else ...
but I don't think it is totally spurious. here is the same plot for the set of EA countries that includes Ireland.
Both US FDI in Ireland & Irish holdings of US debt have gone into reverse (the fall in FDI tho is just a fall in the cash held by the Irish subs of US firms)
6/x
and since so much of this involved or touched a euro area country, it has similar implications for the euro area's balance of payments. FDI into the EA fell (US firms were "reinvesting" less in tax havens) and European demand for US debt fell ...
7/7
p.s. will do a blog on this too, but likely not til after labor day ...
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Bloomberg reports that China's regulators have warned China's state banks about the risk of holding too many Treasuries --
The Chinese regulators must know something that the Treasury doesn't, as the Treasury data doesn't suggest that China has been buying any Treasuries
1/
The official US data on foreign holdings doesn't show any basis for Chinese concern -- China's Treasuries in US custodianship (in theory state accounts as well as state bank accounts) are heading down not up
2/
That is of course inconsistent with the warning that the regulators provided to the state banks! They seem to be warning about nothing ...
The Treasury has indicated that it will look at the activities of China's state banks in its next assessment of China's currency policies--
It is hard to see how this doesn't become a bit of an issue ... unless of course summitry gets in the way of analysis 1/
It is quite clear that state bank purchases (and in 23/ early 24 sales) of fx have replaced PBOC purchases and sales and the core technique China uses to manage the band around the daily fx -- i.e. settlement looks like an intervention variable
2/
My latest blog looks both at how fx settlement (a measure that includes the state banks) has displaced the PBOC's own reported reserves as the best metric for Chinese intervention & lat some of SAFE's balance sheet mysteries
The blog is detailed and technical -- and thus probably best read by those with a real interest in central bank balance sheets, the balance of payments and how to assess backdoor foreign currency intervention
2/
Drawing on historical data, I propose that the gap between fx settlement and the foreign assets on the PBOC's balance sheet (fx reserves + other f. assets) is a good indicator of hidden intervention --
Obviously overshadowed by the news about a Fed nomination, but the Treasury released its delated October 2025 FX report today and it is worth reading -- not the least b/c of a clear warning to SAFE.
1/
This seems clear
"An economy that fails to publish intervention data or whose data are incomplete will not be given any benefit of the doubt in Treasury’s assessment of intervention practices."
This report only covers the period between July 24 and June 25, so it misses the bulk of the 2025 surge in fx settlement (December = $100b plus). But this chart suggests the use of more sophisticated analytical techniques than those used in past reports --
A bit of background. Taiwan's lifers hold $700 billion in foreign currency assets abroad (more counting their holdings of local ETFs that invest heavily in foreign bonds) v ~ $200 billion in domestic fx policies -- so fx gap (pre hedging) of $500 billion
2/
Taiwan's regulator (perhaps the most complicit regulator on earth) not allows the lifers NOT to mark their fx holdings to the fx market -- so the lifers are incentivized not to hedge (and they are rapidly reducing their hedge ratio)
Japan is an interesting case in a lot of ways. It has a ton of domestic debt (and significant domestic financial assets) which generates heated concerns about its solvency/ ability to manage higher rates. But it is also a massive global creditor --
1/
Japan's net holdings of bonds (net of foreign holdings of JGBs) is close to 50% of its GDP (a creditor position as big v GDP as the US net det position). That includes $1 trillion in bonds held in Japan's $1.175 trillion in reserves, + over $2 trillion in other holdings
2/
That translates into big holdings of US debt -- the MoF's Treasuries all show up in the US TIC data, but the corporate bonds held by the lifers, postbank and the GPIF are only partially captured in the US data b/c of third party management/ the use of EU custodians