Bessent’s yen bet and the machinery behind US debt

United States · Japan · Market funding

Speaking in Dallas on 8 September 2026, Scott Bessent described his new position in the markets in casino terms: “I have asymmetric information. I am the house now.” The Treasury secretary was discussing intervention in the yen, as reported by the Financial Times and quoted by Fortune. He was not promising to defend a particular yield on US government bonds. ^[1]

Yet there is a substantive connection between the currency market and Treasury financing. Japan has confirmed a coordinated intervention with Washington and announced plans to use a Federal Reserve facility that provides dollars against Treasuries. Separately, Bessent has described government purchases of older bonds as a way for their sellers to borrow again and buy new issues. ^[3][5][9]

The common thread is the capacity of investors to hold their positions. A public buyer can make an inventory easier to sell. A funding facility can make an outright sale unnecessary. A credible intervention threat can make a currency bet less attractive. None of those mechanisms eliminates the underlying need for financing, or gives the policymaker control over every response.

What Bessent knows about Japan’s intentions

At Southern Methodist University, Bessent linked his confidence to his knowledge of what the Bank of Japan and Japanese policymakers might do. That is an advantage he claims to possess. The confidential information to which he refers has not been made public. ^[2]

The warning can matter without a date or an amount attached to the next intervention. An investor shorting the yen must weigh interest-rate differences and commercial flows against the possibility of official purchases large enough to disrupt the trade. If that risk looks less tolerable, the investor may buy back yen and reduce the position. This is an interpretation of the signal, not evidence that particular orders followed his remarks.

Knowing a government’s intentions is valuable. It does not reveal every private position, the next demand for cash, or tomorrow’s economic release. Nor are policy decisions necessarily fixed: the market’s response can change the authorities’ own plans.

The remarks concern the exercise of public office. The documents examined here provide no basis for turning them into an allegation of personal insider dealing. The economic question is more useful: how far can a policymaker raise the cost of betting against an intervention without generating strains elsewhere?

Tokyo still decides whether Japan intervenes

Japan’s Ministry of Finance confirmed in a statement published on 3 August 2026 that it had bought yen on 31 July, US Eastern Time, in coordination with the US Treasury. It left open the possibility of further joint action. The stated purpose was to counter excessive volatility and disorderly currency movements, not to defend a published exchange-rate target. ^[3]

In Japan, the finance minister authorises currency intervention and the Bank of Japan executes it on the minister’s instructions. Monetary-policy decisions are a separate matter. Bessent’s access to officials in Tokyo does not give him authority to set Japan’s policy rate. ^[4]

The currency sold also matters. A Reuters report on 31 July, citing the Financial Times, described the US leg as the New York Fed selling euros to buy yen on behalf of Treasury. That is not the same transaction as selling US government bonds for dollars and then buying yen. Japan’s statement does not itself supply all the execution details. ^[18][3]

The New York Fed can execute foreign-exchange transactions for the Federal Reserve’s portfolio under FOMC direction, or as Treasury’s fiscal agent for the Exchange Stabilization Fund. The same dealing desk can act for different principals. Its presence does not identify whose balance sheet is being used. ^[7]

The last sentence of Japan’s statement points to Treasuries

The statement’s final sentence says Japan intends to use the FIMA Repo Facility, the repo facility for foreign and international monetary authorities, in the future. The intention is public. A specific Japanese draw used to fund a particular currency intervention has not been established in the material reviewed here. ^[3]

FIMA allows approved foreign monetary authorities to raise dollars temporarily against US Treasury securities, with an agreement to repurchase the securities at maturity. A repo is economically a collateralised loan. It offers an alternative to selling the bonds outright into the market. The Fed explicitly presents the facility as a way to limit the transmission of international dollar-funding pressures to US markets. ^[5][6]

The potential connection is straightforward. An authority can mobilise the value of its Treasuries without finding a permanent buyer immediately. It could then use the dollars to purchase yen. That is a possible funding route, not a reconstruction of how the 31 July operation was financed.

There is still a repayment date. The Fed’s published terms provide for an overnight or seven-calendar-day repo. Pricing is designed for a backstop and generally sits above private-market rates when markets are functioning well. Access requires Federal Reserve approval. This is not a permanent financing commitment that the Treasury secretary can grant on his own. ^[5]

An authority using the proceeds for currency intervention would therefore still have to obtain dollars to repay the repo, or secure alternative financing. The need to sell securities may be deferred or spread over time; it is not removed by definition. Strong collateral does not cancel the borrower’s obligation.

Currency intervention and FIMA answer different questions: what should the authority buy, and how can it finance the purchase? A better answer to the second can protect Treasury-market liquidity without guaranteeing success on the first.

Older bonds, new borrowing

At a separate Breitbart event in Washington on 8 September, Bessent described Treasury buybacks in similarly direct terms. After older bonds are bought back, their sellers “re-leverage and buy new bonds”. He also said the expansion was intended to calm overheated expectations. Breitbart organised the event and is a partisan outlet: its account documents his explanation, not the policy’s effectiveness. ^[9]

Older issues are known as off-the-run securities once a more recent issue has become the benchmark. They can be harder to trade quickly and in size without conceding much on price. A predictable buyer gives a dealer an outlet for inventory it would otherwise have to finance while awaiting another buyer. That is the Treasury’s stated liquidity-support objective. ^[10][11]

On 19 August 2026, Treasury announced that the per-operation ceiling for nominal securities with 10–20 or 20–30 years remaining to maturity would rise from $2 billion to at least $4 billion, effective 9 September through the next quarterly refunding on 4 November. These are operation limits, not guaranteed purchase volumes. ^[8][10]

The official framework remains narrower than an unlimited market rescue. Treasury’s FAQ says buybacks are not currently intended to address episodes of acute stress. Bessent is also seeking to influence expectations. A bout of unease is not automatically acute dysfunction. The two statements nevertheless raise a useful question: what deterioration in trading conditions justifies a larger operation? ^[9][10]

A simple example helps unpack the borrowing. Imagine a dealer holding a bond worth 100, funded with 95 of debt and 5 of equity. The dealer sells it for 100, with no change in price and no costs, then repays 95. It retains 5. Using that equity and a fresh loan of 95, it can purchase another position worth 100.

The inventory has turned over. The dealer has not acquired 100 of new equity, and its final borrowing has not increased. Bessent describes a possible reuse of financing capacity, not proof that aggregate leverage has risen. Leverage could increase if positions expanded or lenders required a smaller equity contribution, but that would have to be demonstrated.

Nor do all participants perform the same job. A dealer holding inventory to facilitate client trading and a fund borrowing to exploit a price discrepancy are not interchangeable. Treating every improvement in intermediation as the same speculative bet obscures where the risks sit.

How Treasury finances its buybacks

Treasuries bought back by the issuer are retired at settlement. Treasury pays for them using resources that can include proceeds from debt issuance and money in its general fund. A buyback can therefore change the composition of outstanding debt without reducing total indebtedness by the same amount. The New York Fed executes the operation as fiscal agent; it is not thereby undertaking a central-bank asset-purchase programme. ^[10][12]

Both sides of the financing need to be examined. If an old long-term bond is replaced with a new bond of similar duration, better tradability may be the main benefit. If it is financed with much shorter-dated issuance, private investors absorb less long-term interest-rate exposure while the government has to refinance sooner. These are alternative debt-management scenarios, not a description of the financing assigned to every actual buyback.

Duration measures a bond’s sensitivity to changing yields. Replacing long debt with short debt can reduce that exposure for investors while increasing the frequency with which the issuer confronts new market rates. An improvement today can have a financing cost tomorrow. Looking only at the bonds purchased cannot settle the question.

There is, however, evidence that the programme can deliver a genuine service. In a working paper published in May 2025, Jing Zhou finds modest liquidity improvements and effects on dealer inventories. The improvement is concentrated around the announcement of eligible securities. Actual purchases show no robust additional narrowing of bid-ask spreads: the baseline estimate even finds modest widening, with identification limitations that warrant caution. Its data run through 1 May 2025, so the findings do not automatically validate the September 2026 expansion. ^[11]

That qualification does not make buybacks pointless. It prevents an unwarranted leap from improving particular transactions to setting the government’s long-term borrowing rate.

A stronger yen can force sales elsewhere

The potential spillover involves investors FIMA does not finance: private borrowers with yen liabilities.

A currency carry trade borrows in one currency and invests in another, often to earn a higher yield. Without an exchange-rate hedge, appreciation of the funding currency makes repayment more expensive in the investment currency. BIS researchers documented how the unwinding of leveraged positions amplified the market turbulence of August 2024. Markets subsequently stabilised quickly. That episode identifies a vulnerability, not an inevitable path for 2026. ^[14]

Consider a deliberately hypothetical one-year trade. An investor borrows ¥160 million at 0.5%, converts it at ¥160 per dollar, and invests the resulting $1 million at 4%. At maturity the investment is worth $1.04 million and the yen repayment is ¥160.8 million.

With the exchange rate unchanged, the dollar investment converts into ¥166.4 million, leaving a ¥5.6 million surplus after repayment. At ¥150 per dollar, however, it converts into just ¥156 million. The investor is ¥4.8 million short, even though the asset delivered its promised 4%. None of these assumptions represents a current market quote.

The calculation leaves out collateral calls during the year. In real markets, investors may have to cut their positions before the planned maturity. If many do so together, their asset sales and yen purchases can reinforce the moves they are trying to escape. The BIS analysis of August 2024 describes this amplification channel; it does not justify attributing every decline in equities or cryptoassets to the yen. ^[14]

The critical separation is between an authority’s reserve portfolio and a leveraged private investor’s assets. FIMA can reduce the need to sell the former. It does not offer general protection to the latter. Stabilising Japan’s currency could remove one source of disorder while setting off portfolio adjustments elsewhere.

The vulnerabilities are not uniform. The Fed’s May 2026 Financial Stability Report places hedge-fund leverage near the highest levels in its observed history, with the latest comprehensive observations covering the third quarter of 2025. It also describes relatively low broker-dealer leverage and adequate intermediation capacity under normal conditions. That is not a measurement of positions on 9 September. ^[13]

Public financial power has more than one owner

In the SMU discussion, Bessent also defended using the US balance sheet as a foreign-policy instrument, including to strengthen alliances in the Western Hemisphere, Reuters reported. He is explicitly assigning public financial capacity a purpose beyond the return on a trade. ^[15]

The legal framework is not that of a hedge fund. The Exchange Stabilization Fund, or ESF, can buy and sell currencies and finance foreign governments. Its operations require the Treasury secretary’s authorisation, under legislation providing for presidential approval. FIMA belongs to the Federal Reserve’s framework; debt buybacks belong to federal debt management. The instruments can complement one another without becoming one pool of money controlled by a single person. ^[16][5][12]

The political thread predates the Dallas metaphor. In his official 6 March 2025 speech, Bessent linked public-sector deleveraging with private-sector releveraging while arguing for financial deregulation. Read alongside his September account of purchases of new Treasuries, the implication is revealing: stronger private financing capacity can also help carry public debt. The comparison does not establish that the government has actually deleveraged. ^[17][9]

How to assess the strategy over time

The strongest case for the strategy is coherent. Coordinated authorities can prevent rushed sales, release dealer capacity and reduce the compensation investors demand for illiquidity. The public sector then saves on a market malfunction without claiming that every high yield is unjustified. The published objectives of FIMA and the buyback programme support that interpretation. ^[6][10]

The less favourable possibility is increasing dependence on official support. If investors rely on a predictable public exit to maintain more fragile positions, withdrawing the support could become harder. That is a behavioural hypothesis to investigate, not an outcome established by the quotation or the buyback announcement.

Evaluation has to follow the mechanism. Useful buybacks should improve trading conditions for the relevant securities, not merely raise their prices. More resilient funding should survive less accommodating credit terms. An actual use of FIMA must be distinguished from an intention to use it and assessed against its repayment schedule. A well-covered auction is one observation, not evidence that Bessent’s remarks caused the result. ^[5][10][11]

That is where the casino comparison breaks down. A public authority can change trading conditions and knows some of its own decisions before investors do. It still has to fund its actions, respect other institutions’ powers and live with private-sector responses. Bessent’s strategy will be judged by the resilience of its funding: can positions survive a change in official support or credit conditions?

Further reading: how the US Treasury market works.

Sources and documents

1. Fortune · 9 September 2026. Bessent’s yen remarks. Quotation sourced to the Financial Times.
2. Bloomberg HT · 9 September 2026. Bessent yen aleyhindeki pozisyonlara meydan okudu. Turkish-language report: Bessent claims insight into Japanese policymakers’ intentions. The private discussions themselves are not public.
3. Ministry of Finance, Japan · 3 August 2026. Statement by Ms. KATAYAMA Satsuki, Minister of Finance, Japan. Confirmation of the 31 July operation, US Eastern Time; final sentence on FIMA. The statement confirms coordination and an intention to use FIMA in future, not a particular draw, amount or use of proceeds.
4. Bank of Japan · Page accessed 9 September 2026. What is foreign exchange intervention? Who decides and conducts foreign exchange intervention?. Finance minister’s decision, Bank of Japan execution and the FEFSA account. Institutional explanation, not a transaction record for a 2026 intervention.
5. Federal Reserve · 21 February 2024. FIMA Repo Facility FAQs. Questions on eligibility, terms, pricing and legal authority. Date shown is the last update. Reviewed terms provide for overnight or seven-calendar-day transactions and Fed-approved access, not guaranteed permanent funding.
6. Federal Reserve · Page accessed 9 September 2026. FIMA Repo Facility. Overview and purpose: an alternative to outright Treasury sales. Describes a facility available to approved authorities. Its general purpose does not establish a particular Japanese use.
7. Federal Reserve Bank of New York · Page accessed 9 September 2026. Foreign Exchange Operations. Section on operations for SOMA and the Exchange Stabilization Fund. One operating desk can serve distinct mandates: the FOMC’s and Treasury’s. Execution does not identify which institution owns the policy decision.
8. U.S. Department of the Treasury · 19 August 2026. Treasury Announces Increased Sizes of Nominal Long-End Liquidity Support Buybacks Beginning September 9. Operation ceilings for the 10–20-year and 20–30-year remaining-maturity sectors. The $2 billion ceiling rises to at least $4 billion per operation, effective from 9 September through 4 November 2026. These are not executed purchases.
9. Breitbart News · 8 September 2026. Scott Bessent: U.S. Bond Market ‘Best Performing in the World'. Passages on repurchases, renewed borrowing and expectations. Account by the partisan event organiser, covering an appearance separate from SMU. Used for Bessent’s reported words, not as verification of his performance claims.
10. TreasuryDirect · Page accessed 9 September 2026. FAQs about Treasury Securities Buybacks. Authority, Purpose, Structure, Acceptance of Offers and Settlement sections. Purposes, exclusions, retirement at settlement and discretion to purchase below the ceiling. Treasury states that the programme is not currently intended for acute stress.
11. IMF Working Papers · Jing Zhou · 9 May 2025. Testing the Liquidity Support Effects of the U.S. Treasury Buyback Program, WP/25/88. Sample on printed page 9; listing versus purchasing effects on pages 20–21, Table 8. The author’s research, not an official IMF position. Data end on 1 May 2025. Purchases show no robust additional improvement in liquidity spreads, with identification limitations. It does not test the 2026 expansion. PDF · 43 pages.
12. Federal Reserve Bank of New York · Page accessed 9 September 2026. Treasury Debt Auctions and Buybacks as Fiscal Agent. Fiscal-agent role and the two buyback objectives. The New York Fed executes buybacks at Treasury’s direction, not as monetary-policy asset purchases for its own portfolio.
13. Federal Reserve · Page accessed 9 September 2026. Financial Stability Report, May 2026 · 3. Leverage in the Financial Sector. May 2026 report; broker-dealer and hedge-fund sections. Comprehensive hedge-fund observations run through the third quarter of 2025, not September 2026. For hedge funds, it uses gross notional exposure relative to NAV, not the accounting measure used in our example.
14. BIS Bulletin · Aquilina, Lombardi, Schrimpf & Sushko · 27 August 2024. The market turbulence and carry trade unwind of August 2024, No. 90. Key takeaways and analysis of margin calls and position unwinds. Historical amplification followed by rapid stabilisation. No 2024 exposure estimate is recast as September 2026 data, and the illustrative examples use no market quotes from the paper. PDF · 8 pages.
15. Reuters · via Investing.com · 8 September 2026. US Treasury chief Bessent backs using US financial power as foreign policy tool. SMU remarks on foreign policy, distinct from each instrument’s legal authority.
16. U.S. Department of the Treasury · Page accessed 9 September 2026. Exchange Stabilization Fund. Legal framework, instruments and transaction authorisation. Transactions require the secretary’s authorisation under legislation providing for presidential approval. The ESF is distinct from FIMA and debt-management funding.
17. U.S. Department of the Treasury · 6 March 2025. Treasury Secretary Scott Bessent Remarks at the Economic Club of New York. Passage linking public-sector deleveraging with private-sector releveraging. Official transcript used to compare stated policy ideas, not to establish that public-sector deleveraging occurred.
18. Reuters · reprise Investing.com · 31 July 2026. US Treasury undertakes intervention in yen market, FT reports. Report attributing euro sales and yen purchases to the Financial Times. One reporting chain, not two independent confirmations. The detail concerns the US leg and establishes neither Treasury sales nor FIMA usage.