Brad Setser Profile picture
CFR senior fellow. Views are my own. Retweets are not endorsements. Writes on sovereign debt and capital flows.
79 subscribers
May 9 8 tweets 3 min read
A bit more on China's April trade data.

So did the tariffs bite? In some sense, no: exports grew 8% y/y, and the trade surplus continued to soar (driven by a rising surplus with places other than the US)

1/ Image But in some sense yes -- exports to the US were $33b, and they should have been $45-50b (recently exports to the US and EU have basically tracked)

2/ Image
May 8 9 tweets 2 min read
The Caymans islands aren't complaining about being flooded by US exports of financial services ...

Ireland is not complaining about being flooded by US exports of research and development services. Nor is Switzerland.

1/ TSMC (Taiwan) wouldn't have a business without importing chip designs from Singapore (NVIDIA I suspect) and Ireland (Apple) -- US service exports were are indirect ...

2/
May 7 13 tweets 4 min read
Excellent piece on the dollar and the euro by @helene_rey

This distinction is absolutely critical

1/many Image Dr. Rey also -- and correctly -- pushes back on the notion (she says "misunderstanding") that the dollar's international role requires that the US run current account deficit (reserve inflows can fund capital outflows)

2/

ft.com/content/5bc026…
May 6 9 tweets 3 min read
Almost impossible to overstate just how crazy the US trade data -- and the pharmaceutical trade data -- has become.

1/ Image I would not believe my spreadsheets if it wasn't 100% clear exactly what was going on -- big US pharma (and some European pharma companies) rushing high value goods into the US to beat the tariffs ...

2/ Image
May 4 20 tweets 7 min read
A bit of background on Taiwan ahead of what may be an eventful week. Certainly there will be a lot attention on the action to Taiwan's central bank after Friday's TWD move.

Key context: the TWD is incredibly weak

1/ Image That weakness is obvious from the 15% of GDP current account surplus or any examination of purchasing power parity. In real effective terms, the Taiwan dollar is down 25% from its pre-Asian financial crisis level (i.e. the mid 90s)

2/ Image
May 2 4 tweets 2 min read
A January 2025 article that those just now waking up to the big overnight Taiwan dollar move may want to read.

Tons of material on the lifers hedging need ... and how the long period of TWD weakness masked an ongoing asset-liability mismatch

ft.com/content/972c54… Lot of interesting detail on Taiwan (Lifers open position, CBC intervention patterns, etc). But at the end of the day Taiwan is simple: its massive external surplus can be sustained only so long as someone big is willing to add to their external assets Image
May 1 8 tweets 3 min read
An enormously important question, and one where my obsession with pharmaceutical trade may help provide the answer as the import surge was heavily driven by stockpiling of high value patent protected meds

1/ The GDP data release suggests (backed by EU data and Jan/ Feb US import data) that pharma imports rose by about 0.5 pp of US GDP in q1 (a huge sum!) -- massive jump over an already rising baseline

2/ Image
Apr 30 14 tweets 5 min read
A few words on Japan's balance of payments, which confuse many (including, at times, my friends at the IMF) --

Japan now runs a large current account surplus (more than China) of 5% of GPD even with a 1% of GDP trade deficit

1/ Image The reason, of course, is that Japan now runs a massive surplus in its investment income account (something the IMF with its obsession on fiscal debt tends to overlook). A weak yen of course inflates foreign earnings, and rising global rates

2/ Image
Apr 30 4 tweets 2 min read
A shout out to my friends in the pharmaceutical industry, whose skill at avoiding taxes extends to avoiding import taxes (tariffs).

Pharma alone increased q1 imports by a half point of GPD

1/ Image The IT industry also led the tariff front running charge.

and there was surprisingly little tariff front running in either autos or semiconductors (chips) despite the threatened USMCA tariffs on fentanyl and the chips 232

2/
2/ Image
Apr 30 9 tweets 3 min read
A very DC joke back in 2012 was that the IMF should be renamed the European Monetary Fund. It certainly should now be called the Argentine monetary fund.

Lending to Argentina now far exceeds the funds lending to the rest of the world after COVID

1/ Image Measures of balance sheet concentration should be viewed with a grain of salt as the IMF has a lot of unused financial firepower these days -- but $50b plus is a lot of exposure to a single country even for the Fund.

2/ Image
Apr 29 7 tweets 2 min read
The IMF has had a hard time demonstrating that the focus of a renewed discussion of global imbalances should be on China rather than Europe (EA/ EU) and Japan, largely b/c China's reported current account surplus is low and Europe and Japan get a lot of investment income.

1/
Services trade is also bit difficult b/c tax avoiding US firms add to the Irish and thus the EU services surplus (Microsoft exports services from Ireland); remember that Ireland is like the fourth or fifth largest services exporter globally ...

2/
Apr 29 21 tweets 6 min read
A lot of the analysis of China's trade exposure to the US has been unsatisfying (at least to me) as looks exclusively or primarily at the bilateral trade flows.

A broader perspective is needed.

1/ Image Looking at the bilateral flows first raises a basic issue -- whether to use the US import data ($440b, somewhat over 2 pp of China's GDP) or the Chinese export data ($530b, somewhat under 3 pp of China's GDP).

2/
Apr 28 8 tweets 3 min read
China hasn't been adding in scale for a long time, but if the euroclear custodial account is included, Chinese holdings of all Treasuries rose a bit recently -- but their holdings are increasingly short-term.

1/ Image A better chart than the BoA chart -- straight from the US balance of payments but with the "Belgium/ euroclear" adjustment. LT bonds includes Agencies. No significant net purchases and sales in the US data since 2016

2/ Image
Apr 27 5 tweets 2 min read
Amid all the debate about trade and the dollar's global role, one facet of the United States international economic position hasn't gotten enough attention -- the traditional US surplus in investment income has disappeared

1/ Image Net interest payments to the rest of the world have gone up because rates have gone up, nothing fancy --

2/ Image
Apr 27 5 tweets 2 min read
The 2022 BEA data (last available) put in China sales of US multinationals at $490 billion, with in country "income" of $37 billion or so. Table 3

bea.gov/news/2024/acti…

1/3 x.com/VKMacro/status… The bilateral current account data (which has some problems because a firm like Apple will report its profits from China in "Ireland" given its corporate structure) reports an even more modest total for FDI income accruing back to the US (less than $10b in 2024)

2/ Image
Apr 22 8 tweets 3 min read
Stability in the central parity rate (the "fix") for the yuan led to a reduction in visible pressure on the yuan in February and March. Settlement was ever so slightly positive in March.

No data for April obviously

1/x Image The yuan did depreciate a bit in April (the fix dipped below 7.2) but not by all that much. The small move was likely a bit of a signal to the US, but so far the yuan hasn't central to China's response to this round of tariffs (@marcmakingsense got this right)

2/ Image
Apr 22 5 tweets 2 min read
China's most reliable proxy for the intervention of the state banking system (together with the PBOC) -- fx settlement -- is out of March. And it suggests that yuan stability "worked"; there were almost no net sales -

1/ Image The settlement series has been volatile -- large inflows with the carry unwind last fall, large sales in January in anticipation that the yuan might move with Trump -- but it has been very stable the last two months

2/ Image
Apr 20 10 tweets 3 min read
On one hand, I am pleased to see a few more folks recognize that China doesn't just hold Treasuries ...

On the other hand, the numbers here are WAY off. The TIC data shows current holdings of $200 billion, down from ~ $270b at the end of 2023

1/ Image The right numbers for China.

~ $700 in LT Treasuries in US custodians
~ $300b in LT Treasuries in non US custodians (mostly Euroclear), obviously an estimate
~ $250b in bills and deposits
~ $200b in Agencies in US custodians
~ $300b in equities ...

2/ Image
Apr 17 14 tweets 5 min read
Some medical test results need to be interpreted by a trained professional.

China's high frequency cross border data probably should come with the same warning.

The PBOC & state bank data for March is out -- and has a few surprises

1/ Image The big surprise is that in March, amid the first round of tariffs, the state banks continued to add to their net foreign assets ... net foreign assets topped $1.1 trillion, gross foreign assets are close to $1.3 trillion

2/ Image
Apr 17 9 tweets 3 min read
Why good balance of payments data matters for policy

the IMF now wants to focus more on global imbalances (a shift from last year, when the external sector report said there was no problem ...)

The reported current account data suggests Europe is the problem not China

1/ Image Europe's surplus tho is inflated by tax avoidance in subtle ways (being the home to Ireland and Luxembourg) while China's surplus (just over $400b in the reported data) is -- in my view -- wildly understated in the official Chinese data.

2/

imf.org/en/News/Articl…
Apr 17 4 tweets 2 min read
China's reported growth accelerated at the end of 2024 and in the first quarter of 2025, partially on a bit of stimulus and partially on strong investment in manufacturing and net exports ...

1/ As the charts of @PkZweifel show, investment in manufacturing is still increasing -- and driving part of the expansion. net exports contributed a crazy 2.5 pp to q4 growth (y/y) in China's official data

2/ Image