Why Japan Is Quietly Selling US Treasuries to Defend the Yen
Tokyo's Golden Week holidays were not as quiet as the calendar suggested. When the yen
broke through the politically sensitive ¥160-per-dollar line on April 30, Japan's
Ministry of Finance, operating, as it always does, through the Bank of Japan as its
market agent, reportedly stepped into the foreign exchange market twice in a single
week. Estimated firepower: around $54.7 billion in yen-buying. The fingerprint showed up
almost immediately in Federal Reserve custody data, which recorded an $8.7 billion drop
in the Treasuries it holds for foreign official accounts during the same period.
That correlation matters because Japan is the largest foreign holder of US government
debt, with roughly $1.2 trillion in Treasury securities according to the latest TIC
data. To buy yen at scale, the MOF needs dollars, and the fastest way to get dollars is
to draw down or stop rolling over the Treasury bills sitting in its FX reserves at the
New York Fed. JPMorgan's Yuxuan Tang notes that Tokyo deliberately sticks to T-bills
rather than long-dated Treasuries, and intervenes during US trading hours, to keep
market disruption to a minimum. The flows are surgical. The cumulative effect is not.
The deeper issue is straightforward: the yen's weakness is not random, and intervention
does not address its cause. The Federal Funds rate sits at 3.50–3.75%. The BOJ's policy
rate is 0.75%. That 300 basis-point gap is what fuels the carry trade, investors
borrowing yen, selling them for dollars, and pocketing the spread, and as long as the
gap persists, capital keeps flowing out of Japan. CNBC quoted one analyst this week with
a particularly memorable line: intervention without changing domestic monetary policy is
like tapping the brake while keeping your right foot firmly on the accelerator. The yen
rallied roughly 3% on April 30, then weakened over the next three sessions. Goldman
Sachs estimates Tokyo has perhaps thirty interventions of this size left before its
reserves come under real pressure.
So what is Tokyo actually trying to achieve? Three things, with diminishing realism. The
first is to introduce two-way risk into a market where speculators have been positioned
one way for too long. If hedge funds know the MOF will punch back at ¥160, they trim
shorts. The second is to buy time for Governor Ueda to keep tightening. Three of nine
BOJ board members already supported a rate hike at the March meeting, according to the
minutes, and the bank explicitly cited room for further moves if the Iran-related energy
shock persists. Prime Minister Takaichi prefers loose policy, but a weaker yen means
more imported inflation, which gives Ueda political cover. The third, and most
underappreciated, is coordination with Washington. Treasury Secretary Scott Bessent, who
has called himself America's "top bond salesman", has every reason to want yen weakness
contained. If the carry trade unwinds disorderly, Japanese institutional investors stop
rolling their US Treasury holdings and bring capital home, where 10-year JGBs now yield
2.5%, the highest in 29 years. Bessent's three-day visit to Tokyo this past week, with
separate meetings of Takaichi, Finance Minister Katayama and Ueda, was widely read as
Washington publicly endorsing further BOJ tightening.
The signal for global markets is uncomfortable. The world's largest creditor is becoming
a marginal seller of US debt at exactly the moment when fiscal deficits and the war in
Iran are pushing 30-year US yields toward 5%. Even if Tokyo never sells anything close
to a meaningful share of its $1.2 trillion stockpile, the optionality alone is repricing
duration risk worldwide.
Sources
- Bloomberg, "Fed Data Suggest Japan Sold US Debt Amid Intervention" (May 2026)
- CNBC, "Japan may have fired its yen bazooka twice" (May 2026)
- Japan Times (May 2026)
- US Treasury TIC data
Will Kevin Warsh Trumpify the Federal Reserve?
Jerome Powell's term as Chair of the Federal Reserve ends on May 15. His designated
successor, Kevin Warsh, Donald Trump's nominee, cleared by the Senate Banking Committee
on a 13–11 party-line vote on April 29, with full Senate confirmation expected the week
of May 11, has openly promised "regime change" at the central bank. Whether that phrase
ultimately means institutional reform or institutional capture is now the central
question in monetary policy.
The case that Warsh will deliver something close to capture rests on circumstantial but
accumulating evidence. Trump nominated him explicitly because he wanted a chair who
would consult the White House on rates, breaking with a century of formal Fed
independence. The administration has demanded faster, deeper rate cuts; pursued an
unprecedented criminal investigation into Powell over Fed building renovations (closed
by the DOJ on April 25 but referred to the Fed's inspector general); and is currently
before the Supreme Court attempting to fire Governor Lisa Cook over unproven
mortgage-fraud allegations. Warsh declined in his confirmation hearing to weigh in on
the Cook case, saying the Fed should "stay in its lane". He also signalled openness to a
Fed-Treasury swap-line arrangement, superficially technical, but a structural opening
that, taken to its extreme, would shift control of part of the Fed's balance sheet
toward Treasury.
The case against fast capture is more interesting and rests on three institutional
realities.
Powell, first, is not actually leaving. In a move with no precedent since Marriner Eccles
in 1948, he will step down as chair but remain on the Board of Governors, potentially
until January 2028. Eccles' decision then preceded the 1951 Fed-Treasury Accord, which
re-established Fed independence after WWII fiscal dominance. The historical parallel is,
by Powell's own staff, deliberate. By staying, he denies Trump a board seat to fill with
a loyalist, and his presence creates what some analysts are calling a "two Popes"
problem: a former chair on the same board, fully credentialed and fully audible.
Second, the FOMC is already pushing back. Four officials dissented from the April 30
statement, the most since October 1992, with three of them rejecting language that
hinted at future cuts. Brookings' David Wessel described it as regional presidents
Hammack (Cleveland), Kashkari (Minneapolis), and Logan (Dallas) firing a shot across
Warsh's bow. Worth remembering: the chair sets the agenda, but monetary policy is
decided by a 12-member committee of seven governors and five regional presidents. One
vote out of twelve is leadership; it is not control.
Third, the data themselves resist what the White House wants. Headline inflation is
running above 3% on the back of energy prices from the Iran war. Even if Warsh wished to
deliver cuts on day one, the macro context makes it economically indefensible. Nomura's
David Seif put it dryly: it will take Warsh longer to build the consensus he is trying
to build.
The economist's view is that Fed independence is not a switch but a credibility stock,
accumulated over decades and depleted at the margin. Warsh's reform agenda — fewer FOMC
meetings, scrapping the dot plot, narrower forward guidance, a smaller balance sheet, a
reframed inflation target, is defensible on technocratic grounds, and several of the
proposals have respectable academic support. The danger is in the bundling, not any
single piece. If Warsh is seen to deliver cuts that the data do not justify,
market-implied long-term inflation expectations rise, the term premium widens, and the
rate cuts the White House wanted end up offset by higher long-term yields. Erdoğan's
Turkey is the canonical recent example. The 30-year US Treasury already trades around
4.93%.
Full Trumpification requires not just Warsh, but a board majority and a quiet committee.
Powell's stay, the four-vote dissent, and 3% inflation all push the other way. The
September 2026 FOMC will be the first real test.
Sources
- Washington Times, "Federal Reserve faces unorthodox leadership change" (April
2026)
- CNBC, "Warsh's take on Fed independence" (May 2026)
- CNN Business, PBS NewsHour, Marketplace
Switzerland's Inflation Surprise — The SNB's Problem Just Flipped
For the better part of two years, the Swiss National Bank's defining macro problem ran in
the wrong direction. Not too much inflation, but too little. Until last week,
Switzerland was the only major economy where central bankers were having serious debates
about whether to return to negative rates. April changed the framing. Swiss consumer
prices rose 0.6% year-on-year, up from 0.3% in March, the fastest reading in 16 months.
Every forecaster Bloomberg surveyed expected an acceleration; few had positioned for one
of this size.
For the SNB, which cut rates six consecutive times between March 2024 and June 2025 to
bring the policy rate from 1.75% down to 0%, this is a meaningful adjustment of the
conversation. Inflation at 0.6% is still safely inside the 0–2% target band and would
barely register in Frankfurt or Washington. But it does end, at least for now, the
deflation discussion that had dominated SNB communications since November 2025, when
consumer prices flatlined at 0.0% YoY.
The driver is almost entirely external. Switzerland imports roughly 23% of its consumer
basket, and what April's print really shows is the Iran war's energy shock arriving in
Swiss CPI through the import channel. The same forces that pushed eurozone inflation to
3.0% are reaching Switzerland with a lag and a much smaller amplitude, thanks to two
structural buffers. The franc has appreciated about 2% against the euro since the
conflict started on February 28, on safe-haven flows, which mechanically dampens
imported energy inflation. And the Swiss electricity mix is overwhelmingly hydro and
nuclear, leaving the country far less exposed to gas and oil shocks than Germany or
Italy. Both buffers work. Neither is infinite. April is the data telling the SNB that
even Switzerland's insulation has limits when oil hovers near $100 a barrel and the
Strait of Hormuz remains contested.
So what does the SNB do now? On rates, almost certainly nothing — and that itself is the
policy stance. Swiss policymakers have been unusually explicit since late 2025 that they
prefer foreign-exchange intervention to further rate cuts. The "willingness to
intervene" language in the March 2026 monetary policy assessment was deliberately
strengthened. The SNB is, in effect, running a two-instrument approach in the spirit of
the Tinbergen rule: the policy rate manages the domestic credit cycle, balance-sheet
operations on the FX side manage imported prices. By holding at zero and standing ready
to sell francs if appreciation goes too far, the bank tries to anchor both inflation
directions without crossing the politically toxic threshold of negative rates.
April's print pushes the negative-rate path further out of view. A Reuters poll in March
showed all but one of 29 forecasters expecting the SNB to hold at 0% through 2026, and
markets are now pricing in a small probability of a hike by December. That repricing has
happened in days, not months.
Two angles are worth keeping in mind for anyone watching Swiss monetary policy. First,
the Phillips curve still has signal in small open economies, but the slope is dominated
by import prices. Swiss core domestic inflation remains around 0.4%; almost the entire
move is in traded goods. That is why the SNB treats the franc as its primary lever in
both directions. Second, the SNB's loss function is asymmetric. The bank weighs the
disutility of negative rates more heavily than the disutility of small deflation, which
is a defensible Brainard-style case for cautious instrument use, with a Swiss twist:
pension funds and banks are politically powerful, and negative rates impose visible
costs on both. The April reading quietly removes the most awkward scenario from the
agenda for the SNB's June meeting.
The remaining question is whether 0.6% is a one-off energy spike or the start of a slow
climb back toward the 1% midpoint of the target band. The June monetary policy
assessment is the next read.
Sources
- Bloomberg, "Swiss Inflation Hits 16-Month High" (May 5, 2026)
- Federal Statistical Office (BFS) April 2026 CPI release
- SNB Monetary Policy Assessment, March 19, 2026
- Reuters/ING/Capital Economics commentary, March–May 2026