High Politics, the Yen Carry Trade, and the Limits of US-Japan Financial Coordination
Years ago, while studying international relations with Professor Ralph Pettman, I wrote papers that repeatedly returned to a distinction between high politics and low politics.
Professor Pettman tended to think about international relations as a hierarchy.
At the top was high politics: diplomacy, national security, war, alliances and the direct collision of sovereign national interests.
Below that was low politics: economics, trade and finance.
And below that, in what I loosely thought of as lower politics, were culture and broader social interaction.
Recently, I found myself taking those old papers off the shelf.
Not because I wanted to revisit an academic debate, but because the framework suddenly seemed remarkably useful for understanding the Federal Reserve, the Japanese yen, U.S. Treasury yields and the global financial system in 2026.
The world that followed the Cold War encouraged investors to think that economics could operate relatively independently of geopolitics.
Supply chains were optimized for efficiency. Capital moved internationally in search of returns. Central banks were analyzed primarily through inflation, unemployment and growth.
That world is changing.
Semiconductors are now national-security assets. Energy infrastructure is strategic infrastructure. Critical minerals are instruments of state power. Supply chains are increasingly organized around alliances rather than cost alone. Currencies, trade policy and even sovereign bond markets increasingly intersect with national strategy.
In the current era of supply-chain realignment and economic security, high politics is increasingly constraining low politics.
That is the framework through which I want to examine the Federal Reserve's September decision.
My base case is straightforward:
The Fed is likely to hold rates in September.
But that is only the first part of the story.
The more important question for investors may be what happens after the Fed pauses.
Because the Federal Reserve controls the policy rate.
It does not control the entire yield curve.
And it certainly does not control Tokyo, the yen, global capital flows or every leveraged position financed through the world's most important funding currencies.
1. Why Another Fed Hike Is Becoming Harder
The Federal Reserve left the federal funds target range unchanged at 3.50%–3.75% on July 29.
That decision was not unanimous.
Beth Hammack, Neel Kashkari and Lorie Logan preferred a 25-basis-point hike, producing a 9–3 vote. The Fed simultaneously described inflation as elevated while noting that productivity growth and capital investment remained strong.
So another hike cannot be dismissed.
But the data released since that meeting have shifted the balance.
The July employment report was particularly important.
U.S. nonfarm payrolls fell by 23,000, versus expectations for roughly 80,000 new jobs. May and June employment estimates were revised downward by a combined 103,000 jobs.
The unemployment rate actually declined from 4.2% to 4.1%, but that headline masks what happened underneath it: another 264,000 people left the labor force, pushing the labor-force participation rate down to 61.4%, near a five-and-a-half-year low.
Average hourly earnings were only 3.2% higher than a year earlier. So I would not characterize the U.S. labor market simply by saying unemployment is rising.
The more accurate conclusion is that the unemployment rate remains deceptively stable while underlying labor demand is weakening.
A central bank preparing to tighten policy further should want convincing evidence that employment demand and wage pressures remain strong enough to tolerate additional restraint.
2. AI Productivity May Be Changing the Business Cycle
The second piece of the puzzle is productivity.
U.S. nonfarm business productivity increased at a 1.4% annualized rate in the second quarter and was 2.2% higher than a year earlier.
At the same time, unit labor costs increased only 1.3% annualized and 1.4% year over year.
Very roughly, Unit Labor Cost = Labor Compensation ÷ Productivity
When productivity increases, businesses can generate more output from each hour of labor.That means wage growth does not necessarily have to translate one-for-one into inflation.
And this is where AI becomes economically important rather than simply technologically exciting.
I would not claim that current productivity growth has already been proven to be caused by artificial intelligence. The evidence is not mature enough to support that conclusion.
But the coincidence is increasingly difficult to ignore:
AI-related capital investment rising → productivity improving → marginal labor demand weakening → wage pressure moderating.
The Federal Reserve itself described both productivity growth and capital investment as strong in its July statement.
This raises a possibility that is far more interesting than the usual soft-landing debate.
The United States may be able to experience weaker employment growth without immediately falling into recession.
3. Employment Can Slow While GDP Keeps Growing
Real U.S. GDP expanded at an annualized 1.5% in the second quarter, down from 2.1% in the first quarter.
That is not an overheating economy. But it is not a recession either.
Consumer spending, investment and exports all contributed positively to second-quarter growth. This matters because the composition of American growth is changing.
Massive investment is flowing into data centers, computing infrastructure, electrical equipment, power generation, grid infrastructure and advanced manufacturing.
The traditional business-cycle relationship was relatively intuitive:
Higher Employment → Higher income → Higher consumption → Higher GDP
and in reverse:
Lower Employment → Lower income → Lower consumption → Recession
But an economy undergoing rapid capital deepening can behave differently.
If:
capital investment accelerates
productivity increases
output per worker increases
then GDP can continue growing even with slower labor-force expansion.
That does not mean employment no longer matters.
It means employment and output may become less tightly coupled than they were in previous cycles.
For the Fed, that creates a very unusual policy environment.
The economy may be too resilient to cut rates aggressively while the labor market is simultaneously becoming too weak to justify further tightening.
4. Inflation Is Still High to Cut
This is why I am not arguing for a September rate cut.
July CPI rose only 0.1% month over month, while headline inflation eased to 3.4% year over year from 3.5%.
Core CPI rose 0.2% for the month and slowed to 2.5% year over year.
Producer prices were also relatively benign.
The July PPI was unchanged month over month, while its year-over-year increase slowed from 5.5% to 4.7%.
But the Fed's preferred inflation gauge remains much more uncomfortable.
June headline PCE inflation was 3.7% year over year, while core PCE stood at 3.3%.
That is simply too far above the Fed's 2% target to make a September rate cut compelling.
So the question is not whether inflation has disappeared.
The more useful question is “Is inflation broad enough, persistent enough and sufficiently demand-driven to require another rate increase?”
The recent data make that argument increasingly difficult.
5. Warsh's Interest in Trimmed Inflation Is Important
This distinction becomes particularly relevant under Fed Chair Kevin Warsh.
Warsh has shown interest in “trimmed” measures of inflation, which attempt to remove unusually extreme price movements and identify the broader underlying inflation trend.
The Dallas Fed's Trimmed Mean PCE is particularly useful here.
For the 12 months ending in May, the measure increased only 2.4%, compared with headline PCE inflation of 4.1% and core PCE inflation of 3.4% at the time.
That does not mean trimmed inflation should automatically replace headline or core inflation.
Dallas Fed researchers themselves have cautioned that unusually skewed price distributions can sometimes cause trimmed measures to understate meaningful inflation pressure.
But the divergence tells us something.
Some of the current U.S. inflation problem may reflect large sector-specific shocks rather than generalized domestic overheating.
That distinction matters enormously when the largest shocks originate outside traditional monetary economics.
Which brings us to oil.
6. Oil Is Sending a Strange Signal
The geopolitical situation in the Middle East remains extraordinarily serious.
Yet on August 18, despite renewed U.S.-Iran tensions and the collapse of peace efforts, Brent crude traded around $91–92 per barrel, while WTI was around $85–86.
Those are high prices. But consider the geopolitical environment. The Strait of Hormuz has been severely disrupted. The U.S.-Iran conflict has persisted for months. Maritime security remains unstable.
Under those circumstances, oil is not dramatically high under such an environment. One answer is demand. It is China.
China's refinery throughput in July was 15.8% lower than a year earlier, while crude imports were approximately 24% below year-earlier levels.
So an extraordinary geopolitical supply shock is meeting relatively weak demand from the world's largest crude importer.
That tells us something important about inflation.
High oil prices can raise CPI. But they do not automatically prove that the U.S. economy is overheating. And a central bank needs to distinguish the two.
The Fed can suppress American demand. It cannot reopen the Strait of Hormuz.
This is one of the clearest examples of high politics entering the domain of monetary policy.
7. Why My Base Case Is a September Hold
Put the pieces together.
The labor market is cooling beneath a relatively stable unemployment rate. Wage growth is weakening. Productivity is improving. Unit labor cost growth is relatively subdued. GDP remains positive. AI-related capital spending continues to support the economy. Recent CPI and PPI readings have moderated. But PCE inflation remains well above the Fed's 2% objective.
This gives the Fed a surprisingly logical middle ground. Inflation is still too high to cut. The labor market is becoming too soft to hike. Productivity is improving enough to wait.
As of August 18, futures markets priced only about a 35% probability of a September rate hike, down from more than 50% a week earlier.
A Reuters survey conducted August 12–17 likewise found an overwhelming majority of economists expecting the Fed to keep rates unchanged.
The September FOMC decision arrives on September 16.
My base case is therefore Fed holds. But investors should not make the mistake of translating that immediately into Financial Easing.
Those are no longer necessarily the same thing.
8. The Fed Controls the Policy Rate. The Market Prices the 10-Year.
This may be the most important distinction in the entire analysis.
The Federal Reserve directly controls an overnight policy rate. But households and corporations do not finance everything at the federal funds rate.
Mortgage pricing, corporate borrowing costs, equity discount rates and long-duration investment decisions are heavily influenced by the Treasury curve.
And longer Treasury yields contain much more than expectations for Fed policy.
They incorporate expectations about inflation, real growth, fiscal deficits, Treasury supply, foreign demand and term premium.
As of August 18, the U.S. 10-year Treasury yield was approximately 4.72%, while the 30-year yield had risen above 5.32%, its highest level since 2007.
This is occurring while expectations for a September Fed hike are declining. That is not a trivial observation.
The market is already demonstrating that Fed expectations and long-term borrowing costs can move in opposite directions.
Which means the most important interest rate after September's FOMC may not be the federal funds rate at all.
It may be the 10-year Treasury yield.
9. Then Comes Japan
The timing is remarkable.
The Federal Reserve meets on September 15–16. The Bank of Japan meets immediately afterward on September 17–18.
And while expectations for Fed tightening have been falling, expectations for BOJ tightening have been rising.
Recent market pricing suggests roughly an 80% probability of a September BOJ hike.Japan's bond market has already moved aggressively.
On August 18, the 10-year Japanese government bond yield reached 2.945%, its highest level since 1996. The two-year yield reached around 1.7%.
This fundamentally changes the economics of global capital allocation. For decades, Japan was one of the world's largest suppliers of cheap funding.
The basic trade was simple,
Borrow yen cheaply → sell yen → buy a higher-yielding foreign asset.
That is the core logic behind the yen carry trade.
But what happens when the funding currency no longer remains almost free?
10. Japan Owns More Than $1 Trillion of U.S. Treasuries
Japan is not a marginal participant in the U.S. Treasury market.
It remains the largest foreign holder of U.S. government debt.
At the end of June, Japanese holdings stood at approximately $1.116 trillion, down 2.3% from May.
China's holdings, meanwhile, fell to roughly $633 billion, their lowest level since 2008.
A single month of declining Japanese holdings does not prove the beginning of a structural exodus.
That would be an irresponsible conclusion. But the incentive structure is clearly changing.
When Japanese government bonds yielded close to zero, the attraction of foreign fixed income was obvious.
When a 10-year JGB approaches 3%, the calculation becomes different.
A Japanese institution must ask. How much additional U.S. yield compensates me for currency risk, hedging costs and duration risk?
If that required premium rises, U.S. Treasury yields may have to rise with it. So a plausible future combination is Fed Hold, 2-year Treasury stable or lower while JGB yields up, Japanese foreign-bond demand down, U.S. Treasury supply up, term premium up, which leads to 10-year and 30-year Treasury yields higher.
The Fed can pause. The bond market does not have to.
11. But Here High Politics Changes the Calculation
A purely financial model might stop here and predict:
BOJ tightening → Japanese repatriation → Treasury selling → U.S. yields explode.
I think that analysis is incomplete.
Japan is not merely an institutional investor maximizing portfolio yield. Japan is one of America's most important strategic allies.
The United States is not simply the issuer of a bond Japan happens to own. It is the foundation of Japan's external security architecture.
And the United States has an equally strong interest in keeping Japan economically and financially stable as strategic competition intensifies across Asia.
This is precisely where high politics begins constraining low politics.
Would Japan prefer a stronger yen? Increasingly, yes.
Would Washington tolerate a more normal Japanese monetary policy? Almost certainly.
But would either government benefit from Japan dumping hundreds of billions of dollars of Treasuries into a fragile market and sending American mortgage rates, corporate borrowing costs and long-term yields sharply higher? It is difficult to see why.
Japan would damage its own remaining Treasury portfolio. The United States would suffer tighter financial conditions. And two strategic allies would create a global financial accident against their own interests.
That is why I believe the probability of a deliberate, disorderly Japanese government liquidation of U.S. Treasuries is much lower than a purely mechanical financial model might suggest.
Recent policy actions reinforce that view.
12. Bessent's Yen Intervention Is Financial Diplomacy
The United States and Japan recently carried out an unusual coordinated intervention to support the yen.
U.S. Treasury Secretary Scott Bessent subsequently said he was prepared to repeat coordinated action if necessary and publicly argued for expanding the Federal Reserve's FIMA Repo Facility.
The Fed describes FIMA as a facility allowing approved foreign monetary authorities to temporarily obtain dollars against U.S. Treasury securities rather than selling those securities directly into the open market.
That mechanism exists specifically because stressed foreign dollar demand can otherwise create disorder in the U.S. Treasury market.
In other words:
Japan can obtain dollar liquidity without necessarily dumping Treasuries into the market at precisely the worst possible moment.
It is financial diplomacy.
Washington wants to prevent disorderly yen depreciation. But it also wants to avoid an intervention process that destabilizes its own government bond market.
Tokyo wants a more sustainable yen. But it does not want to destroy the value of its dollar reserves or destabilize its principal strategic ally.
The implicit strategic objective therefore appears to be a stronger yen, but an orderly Treasury market.
That is exactly what I mean when I say high politics is beginning to dominate low politics.
13. There Is One Problem: Hedge Funds Do Not Have Allies
Governments can coordinate but private markets are different.
Washington can speak to Tokyo. The Treasury Department can communicate with Japan's Ministry of Finance. The Fed can provide liquidity facilities. The BOJ can control the pace and language of monetary tightening. Central banks can intervene in currencies.
But none of them can issue instructions to every hedge fund, CTA, bank trading desk, leveraged asset manager or derivatives portfolio around the world.
Private capital does not care about the U.S.-Japan security alliance.
Its decision function is different, return, volatility, funding cost, leverage, margin and liquidity.
This is where the real risk appears. If the yen strengthens gradually, carry positions can be reduced gradually. But if USD/JPY begins moving violently lower, investors that borrowed yen to finance foreign positions can start losing money very quickly.
Then a reflexive process becomes possible:
Yen rises → carry losses rise → positions liquidated → foreign assets sold → yen bought back → yen rises further
At that point the question is no longer whether Washington and Tokyo understand what is happening.
“Can they control the speed?”
14. The Real Risk Is the Speed of the Yen Carry Unwind
A carry trade does not become dangerous merely because it eventually closes.
Every trade closes eventually. The dangerous variable is how quickly leverage has to be removed.
Imagine the September sequence.
The Fed announces a hold. Markets initially interpret it positively. The two-year yield falls. The dollar weakens. Technology stocks rally. Then, one or two days later, the BOJ tightens policy or signals a faster normalization path. The yen strengthens.
Initially, this is exactly what policymakers want. But suppose USD/JPY falls faster than expected. Leveraged investors start reducing positions. The yen strengthens further. Stop-losses trigger. Volatility rises. Risk limits tighten. More positions have to be sold.
At that point, the market itself becomes the transmission mechanism.
And this exposes the limit of the high-politics thesis. High politics can constrain government behavior but it cannot eliminate market reflexivity.
Or, put differently:
Governments can coordinate policy. They cannot coordinate every leveraged position in global markets.
15. The Market Signal I Would Fear Most
If this thesis is correct, there is one combination investors should watch particularly closely.
A rapidly strengthening yen combined with rising U.S. 10-year and 30-year Treasury yields. Normally, stronger expectations for Fed easing would produce something like U.S. yields down, weaker dollar, stronger Yen.
But imagine if USD/JPY sharply fall while yields up.
That is a different signal. It could indicate that long-duration Treasury demand is weakening even while expectations for Fed tightening are falling.
Then two separate deleveraging mechanisms could occur simultaneously.
The first is Yen carry deleveraging. The second is duration and valuation compression from rising U.S. long-term rates.
Higher Treasury yields raise the discount rate applied to future corporate cash flows.
That is particularly important for expensive, long-duration growth companies.
It also raises mortgage rates and corporate financing costs.
So even if the Fed leaves rates unchanged: Financial conditions could tighten anyway.
This is why the bond market may matter more than the FOMC headline.
16. This Is Also Why “Fed Hold = Stocks Up” May Be Too Simple
Markets have been conditioned to react to central banks almost mechanically.
Dovish Fed leads to stocks up, hawkish Fed leads to stocks down. But that framework becomes less reliable when the long end of the yield curve is being driven by forces outside the Fed's immediate control.
On August 18, U.S. long-term Treasury yields rose sharply even though weak economic data were simultaneously reducing expectations for Fed tightening. That is precisely the type of divergence investors should pay attention to.
Suppose the Fed holds in September and the equity market rallies for several days.That does not necessarily tell us the tightening cycle is over.
If the 10-year subsequently moves from roughly 4.7% toward 5% while the 30-year remains above 5%, the market could effectively impose another round of tightening without the Fed changing its target rate at all.
The relevant question would no longer be:
“What did the Fed do?”
It would be:
“Who is willing to finance the United States at the long end of the curve, and at what price?”
17. Why This Matters From Korea and Asia
From New York, the yen is often discussed primarily as a currency or funding variable.
From Asia, it is much more than that.
Japan and Korea compete across autos, machinery, industrial equipment, chemicals, materials, components, tourism and parts of advanced manufacturing.
A persistently weak yen can improve Japanese exporters' relative price competitiveness.
A gradual appreciation of the yen can therefore improve the competitive position of some Korean companies.
But the word gradual is crucial.
If the yen strengthens because Japan is normalizing monetary policy in an orderly manner, that can be constructive for parts of Korea's export economy.
If the yen strengthens because leveraged carry trades are being forcibly unwound, the effects are completely different.
Global investors reduce leverage. High-beta assets are sold. Asian equities are often hit harder than the most defensive U.S. assets. Foreign flows can reverse quickly. So Gradual yen appreciation can help Korean competitiveness. Violent yen appreciation can become a global risk-off event.
Direction is important but speed is more crucial.
18. Three Possible Paths Through September
I see three broad paths.
The third scenario is possible. But possibility should never be confused with probability.
There are powerful reasons Washington and Tokyo would try to prevent it. That is precisely why I do not regard a catastrophic yen-driven crash as the base case.
Yet there is an equally powerful reason not to dismiss it.
Policymakers control policy. They do not control every balance sheet.
19. What I Will Watch After the Fed Meeting
Heading into September, I will continue to watch employment, wages, PCE inflation, productivity and unit labor costs to determine whether the domestic case for a Fed hold remains intact.
But after the FOMC decision, my attention will increasingly move elsewhere.
I will watch USD/JPY for the speed of yen appreciation.
I will watch 10-year JGB yields for the changing attractiveness of domestic Japanese assets.
I will watch Japanese holdings and purchases of U.S. Treasuries for evidence of structural repatriation rather than a one-month portfolio adjustment.
I will watch the relationship between the U.S. 2-year and 10-year Treasury yields to determine whether the bond market is pricing Fed policy or a rising term premium.
And, above all, I will watch one unusual combination:
Yen stronger + U.S. long-term yields higher.
If that relationship appears forcefully and persists, I would regard it as a much more important warning signal than the Fed's policy statement itself.
Conclusion: The Fed May Pause. The World May Not.
My base case for September remains a Fed hold.
Inflation is still too high to justify a cut.
But labor-market deterioration, weaker wage pressure, improving productivity and softer recent inflation data increasingly weaken the case for another immediate hike.
AI-related investment and capital deepening may allow the U.S. economy to continue expanding even with much weaker employment growth than investors have traditionally associated with a healthy expansion.
So the Fed can wait.
But investors should not confuse a Fed pause with falling borrowing costs.
Japan is beginning to normalize monetary policy.
Japanese yields are rising.
Japan owns more than $1 trillion of U.S. government debt.
And the yen remains one of the world's most important funding currencies.
The strategic relationship between Washington and Tokyo gives both governments powerful reasons to prevent a disorderly transition.
Recent coordinated yen intervention and Bessent's support for a larger FIMA backstop suggest that both sides understand how closely the yen and the Treasury market have become connected.
This is exactly what we should expect in the age of supply-chain realignment.
Economics is no longer operating independently of geopolitics.Currencies are strategic. Government bonds are strategic. Supply chains are strategic.Capital flows increasingly have strategic consequences.
High politics is increasingly shaping low politics.
But high politics has limits.
Washington can negotiate with Tokyo.
The Treasury Department can coordinate with Japan's Ministry of Finance.
The Federal Reserve can provide liquidity.
The BOJ can adjust monetary policy.
Governments can intervene in currencies and design mechanisms to discourage disorderly Treasury selling.
What they cannot do is negotiate with every leveraged portfolio in the global financial system.
And that may be the most important risk heading into September.
The danger is not necessarily that Washington and Tokyo fail to understand the problem. The danger is that private markets move faster than diplomacy can manage them.
So after September's Fed meeting, I will not be watching only the federal funds rate.
I will be watching the yen. I will be watching Japanese bonds. And, most importantly, I will be watching the U.S. 10-year Treasury.
Because the Fed may pause.
The bond market may not.
And if high politics can manage governments but cannot control the speed of private deleveraging, the next major global market shock may not begin in Washington.
It may begin in Tokyo.