Thread by @Brad_Setser

Foreign demand for U.S. Treasuries fell off in the fourth quarter of 2018.

Total foreign purchases for the year are about a quarter point of GDP.

The U.S. external deficit isn't being funded by inflows into Treasuries. So how is it being funded (1/x)

The source of the fall off in demand for Treasuries in q4 2018?

"Official" investors, e.g. foreign central banks

(2/x)

Foreign demand for Treasuries has historically largely come from official investors abroad (reserve managers) not private investors, and with EM reserves under pressure, the fall off in official demand isn't really a surprise (also payback for a strong q3)

(3/x)

Implication of course is the 6.5 pp of so of Treasuries that the market needed to absorb in 2018 (counting the fall in the Fed's holdings were almost entirely absorbed by domestic investors ...

(4/x)

In the past few years, foreign investors (think private pension funds and insurers in Europe, insurers and banks in Japan, Korea and Taiwan) were buying a lot of corporate bonds.

Not in 2018. Bond inflows were down

(5/x)

There is a complicated story here.

Flatter yield curve meant fx hedged purchases weren't as attractive for many private investors abroad.

And the tax reform mean US firms weren't holding funds abroad that they needed to invest in US assets.

But that's for another time

(6/x)

What then helped finance the current account deficit in 2018?

FDI.

As in 2015 ...

(7/x)

But it is a funny kind of net inflow.

Actual foreign direct investment in the U.S. is down (doesn't bother me much - the high level of 2015 was a function of some inversions)

Inflow is thus a function of the absence of U.S. direct investment abroad

(8/x)

It should go without saying that this is tax related.

Before the tax reform, a lot of US FDI abroad was in a sense fake, as it reflected cash that US firms were holding abroad so as to be able to defer their corporate tax bill.

(9/x)

What's interesting is that the net FDI inflows in 2015 were tax related (inversions) ... and so too is the net FDI inflow of 2018.

And if U.S. firms structurally hold less "cash" abroad, offshore demand for U.S. financial assets will also be a bit lower going forward

(10/x)

Flow effect could be significant -- before the tax reform "reinvested" earnings abroad (largely from holdcos in tax centers not from firms building up "real" assets) were running at 1.5 pp of GDP a year.

and of course in '18 have a one off stock adjustment too

(11/11)