Japan Just Broke the Global Bond Market

Japan's 30-year government bond yield reached a record high on May 26, 2026, closing at 3.93%. This marks the highest level since the series began in 1999, following a brief period above 4.03% the previous week. This development reflects a global cost-of-capital shift with implications extending well beyond Tokyo.

Repatriation is now observable rather than merely anticipated. In this context, repatriation refers to Japanese investors selling foreign bonds and reallocating capital to domestic assets. During the first quarter of 2026, Japanese institutions sold $29.6 billion in foreign bonds and directed record monthly inflows into domestic JGB funds. The departure of the marginal, price-insensitive buyer of US duration is beginning to impact market dynamics, as this investor segment previously purchased for yield irrespective of price, and its withdrawal necessitates lower prices to attract replacements.

The hedged-yield calculation has reversed. While the US 30-year Treasury yield stands at 5.07%, the approximately 290-basis-point cost of hedging dollar exposure into yen reduces the net yield to about 2.17%. This figure is notably lower than Japan's domestic 30-year yield of 3.93%. For Japanese life insurers, domestic bonds now offer superior returns, credit ratings, and liability matching.

The August 2024 unwind provides a recent and illustrative example of market disruption. A surprise Bank of Japan (BOJ) rate hike on July 31, 2024, triggered a cross-asset selloff. By August 5, the Nikkei 225 had declined by 12% in a single session, the S&P 500 had fallen by 3%, and the VIX had spiked to 65. Current conditions are characterized by elevated debt levels, higher yields, persistent inflation, and historically extreme short yen positioning.

The upcoming Bank of Japan meeting on June 16 warrants close attention. The April 28 board vote was 6-3 in favor of holding rates, with three members already supporting a hike. Deputy Governor Himino's statement that "real rates remain extremely low" also signals a potential bias toward tightening. The conditional probability of a rate hike in June or July is estimated at approximately 70%.

Japan's bond market is no longer providing implicit support to global fixed-income markets. For the past three decades, the yen carry trade, borrowing at low rates in yen and investing in higher-yielding foreign bonds, helped anchor long-term yields in developed markets. With Japanese government bond (JGB) yields at multi-decade highs, the Bank of Japan signaling further rate increases, and Japanese institutions now net sellers of foreign bonds, this dynamic is reversing. As a result, US and European term premia must rise to reach a new market equilibrium.

Conviction in the directional thesis is high, while conviction regarding timing is moderate. Data confirms that repatriation flows are already occurring. The precise timing of further catalysts depends on the Bank of Japan's meeting schedule.

The anticipated time horizon for this thesis is six to eighteen months. The scenario is expected to unfold over several Bank of Japan policy decisions and multiple Treasury data release cycles, rather than as a single event.

Potential invalidation scenarios include the Bank of Japan maintaining its current policy in June and signaling an indefinite pause, combined with a ceasefire in Iran that reduces oil prices by $20 or more. Such developments would likely dampen inflation and narrow the hedged-yield differential. If both occur simultaneously, the repatriation thesis would lose momentum.

Three factors that historically emerged sequentially are now occurring simultaneously. The Bank of Japan has concluded its prolonged period of accommodative monetary policy and set its policy rate at 0.75%. The April 28 board vote was 6-3 to hold, with three members favoring a rate increase, making the June 16 meeting particularly significant compared to earlier meetings in 2025.

At the same time, Japan is running an expansionary fiscal stance under Prime Minister Sanae Takaichi. The headline is a ¥21.3 trillion stimulus package layered with a $19 billion supplementary budget. For the first time in a generation, Japan's fiscal and monetary impulses both lean inflationary at once.

Layered on top of that, the Iran-driven oil shock is directly importing inflation into Japanese consumer prices. April core inflation moved above the BOJ's 2% target. Brent crude is still near $109 per barrel, roughly 50% above pre-war levels. Oil is a price-level event for an economy that imports nearly all of its energy.

The cleanest way to see how this changes the world is to look at the same day from two desks. On May 18, US headlines focused on the 10-year Treasury yield touching its highest level in a year and tied it to the Iran story. On the same day, in Tokyo, the 30-year JGB yield set a record dating back to 1999. US commentary framed the global selloff around the US fiscal picture. The marginal mover was Japan: the buyer whose bid had just stepped back.

The narrative anchor for this thesis is the May 14 auction of 30-year JGBs, which printed a record auction yield of 3.915% despite drawing solid demand. The data behind why that auction matters tells a simple story. It is the first hard signal that the bond market is now demanding an actual risk premium for Japanese duration. That is what makes this week different from any week in the prior 26 years.

Exhibit 1: The hedged-yield math has inverted

Headline yields say the US 30-year at 5.07% beats Japan's 30-year at 3.93% by 114 basis points (again, a basis point is one hundredth of a percent, so 114 basis points is 1.14%). That is the number on most fixed-income screens. Japanese institutions act on a different one. Life insurers and pension funds with yen-denominated liabilities do not size foreign bond positions on headline yields. They size them on hedged yields.

The hedge cost is roughly the rate differential between the Fed Funds policy rate and the BOJ policy rate. With the Fed at 3.64% and the BOJ at 0.75%, that gap is about 290 basis points. Pull 290 basis points off a 5.07% US 30-year, and the hedged yield is approximately 2.17%. That sits 176 basis points below the 3.93% available domestically.

Figure 1. Hedged yield math, May 26, 2026. The hedged US 30-year now sits below the domestic JGB 30-year.

For a Japanese investor with yen liabilities, home now wins on three dimensions at once: rate, currency match, and a domestic regulatory framework that already favors JGBs. That is the math the rest of this report builds on.

Exhibit 2: Repatriation is a flow already in motion

In the first quarter of 2026, Japanese institutional investors sold a combined $29.6 billion in US Treasuries and other foreign bonds. March 2026 was the largest monthly inflow on record into Japanese sovereign bond funds. Foreign investors, separately, turned net sellers of ultra-long JGBs for the first time in 16 months, offloading ¥81.3 billion on inflation concerns.

The largest single pool worth tracking is the Government Pension Investment Fund. GPIF manages about $1.8 trillion, of which roughly 24% ($430 billion) is in foreign bonds. Reports from the Japanese financial press indicate GPIF is weighing a shift toward domestic government bonds. Even a modest reallocation, say five percentage points, would mean roughly $90 billion of selling pressure on foreign bond markets.

The headline matters: $1.24 trillion in US Treasuries makes Japan the single largest foreign holder of US debt. The flow data shows it is already on the move. Japan does not need to dump the position for this to bite. The base case is a steady reversal at the margin, where prices clear.

Exhibit 3: Japan's fiscal math forces the yield higher

Japan's debt-to-GDP ratio is 236.7%, more than double that of the United States (125%), the United Kingdom (115%), and France (101%). On its own, that level was sustainable for two decades because the BOJ owned roughly half of the JGB market and yields were anchored near zero. Both of those facts are now in transition.

Debt-servicing costs in the FY2026 budget jumped 10.8% to ¥31.3 trillion, roughly a quarter of total Japanese government spending. That projection assumes an interest rate of 3%, the highest assumed rate in 29 years. Every additional 50 basis points on the JGB curve makes that projection look low. Each turn of the screw on yields is also a turn of the screw on the fiscal arithmetic.

Figure 2. Japan's FY2026 debt-servicing cost. The number assumes a 3% rate, a 29-year high.

On January 20, 2026, the 40-year JGB yield breached 4% in a single session, the largest one-day move in 30-year JGB yields since the series began in 1999. The May 14 auction of new 30-year debt drew solid demand, but it still cleared at 3.915%, also a record. Read together, those auctions say something simple. The market is willing to buy Japan's debt, but it now demands a real risk premium to do so.

Exhibit 4: The US side of the equation is already moving

The US 30-year Treasury yield is at 5.07%. The 20-year is at 5.06%. The US 10-year term premium is 0.83%. The federal deficit is running above 7% of GDP, with no fiscal slack to absorb a duration shock. The Federal Reserve's implied year-end Fed Funds path is 4.02%, which prices in roughly two hikes over 12 months.

In the prior cycle, a marginal Japanese seller of $20 billion in Treasuries per month was associated with roughly a 10-15 basis-point pickup in the 10-year term premium. Q1 2026 has already seen a net foreign bond sale by Japan of about $30 billion. If the run rate persists, the second-quarter impulse is larger and lands during a period when the Fed is on hold, and primary dealers (the banks obligated to bid at Treasury auctions) are absorbing larger-than-usual auctions. The chain from Tokyo to Treasury auctions runs more directly than the consensus narrative is currently pricing.

Exhibit 5: Yen positioning is at a historic short extreme

Asset-manager net positioning in yen futures, as reported by the Commodity Futures Trading Commission (CFTC), is at an all-time low of -10.56%, which is 1.6 standard deviations below its historical average (a genuine outlier). Translated into plain language: the world's institutional money has never been this short of the yen, in this data series. That positioning becomes fuel if the catalyst arrives. A confirmed BOJ hike plus continued repatriation flow gives the yen two simultaneous drivers. Crowded short positioning gives any squeeze a force multiplier.

Each stage is a separate transmission channel; each step depends on the prior step holding.

Stage 1: BOJ normalization raises JGB yields. The BOJ exited yield curve control (its policy of capping long-term yields) in 2024 and lifted the policy rate to 0.75% by late 2025. With inflation above 2% and exports up 14.8% in April, the case for another hike is the strongest in a generation. The 10-year JGB is at its highest level since 1997; the 30-year is at a record going back to 1999. Transmission: domestic yields drive the hedged-yield math.

Stage 2: Hedged foreign yields invert versus domestic. With the BOJ at 0.75% and the Fed at 3.64%, the 290-basis-point hedge cost erases the headline US-Japan yield gap. A Japanese life insurer comparing a hedged 30-year Treasury (about 2.17%) against a domestic 30-year JGB (3.93%) now picks home, even before counting the liability match. Transmission: the math drives the rebalancing decision.

Stage 3: Institutions rotate to domestic. Life insurers, pension funds, and banks shift portfolios. Q1 2026 already shows $29.6 billion in net foreign bond selling and record JGB fund inflows. The flow is already in the data. Transmission: portfolio flows hit marginal demand for Treasuries.

Stage 4: Term premium rises globally. With Japan absent at the margin, US Treasury auctions must clear at higher yields to attract domestic and other foreign buyers. The US 20- and 30-year yields are at 5.06% and 5.07%. The 10-year term premium is 0.83%, and the federal deficit is above 7% of GDP, leaving no fiscal slack to absorb the shock. Transmission: term premium feeds into discount rates worldwide.

Stage 5: The yen strengthens, carry unwinds. Higher JGB yields plus repatriation flow put upward pressure on the yen. CFTC asset-manager net positioning is at an all-time low of -10.56%, an extreme short that becomes fuel for a squeeze when the trigger arrives. Transmission: the FX move hits carry-funded positions.

Stage 6: Global risk assets reprice. Carry-funded long positions in US tech, emerging-market bonds, and leveraged credit are forced to delever. The August 2024 mini-unwind is the clearest recent example: the Nikkei 225 down 12% in a single session, the S&P 500 down 3%, the VIX touching 65. Transmission: deleveraging breaks equity, credit, and emerging-market correlations.

The weakest link in this chain is Stage 5. The yen has not yet broken. Carry positioning is at extremes, but the spark requires BOJ confirmation. If the BOJ delays in June and July, the chain stalls at Stage 4. The other stages would still operate (yields would remain elevated, repatriation flows would continue at a slower pace), but the August 2024-style risk-asset move would not detonate without the FX catalyst.

The closest historical parallel ran from the BOJ's July 31, 2024, hike into the August 5 unwind. Governor Ueda lifted the policy rate by 15 basis points in a move the market read as more hawkish than expected. The yen surged roughly 11% against the dollar in three weeks. The Nikkei dropped 12% in a single session. The S&P 500 lost 3%. The VIX touched 65. Carry-funded positions unwound violently before the BOJ walked back its hawkishness, allowing markets to stabilize.

Figure 3. The August 2024 carry unwind, the cross-asset move that followed the July 31 BOJ hike. One of the sharpest unwinds since 2020.

Parallels to today

The same actors are still in the chain. Japanese institutions are still the marginal holders of global duration. The yen carry trade still funds long positions in higher-yielding assets globally. The mechanism by which a BOJ hike forces a yen rally, a carry unwind, and a repricing of risk assets is structurally intact. None of the plumbing changed in the intervening eighteen months.

Critical differences this time

Four differences cut both ways and shape how a 2026 episode might rhyme without repeating.

Signaling versus surprise. The 2024 hike was a genuine surprise; the 2026 setup is telegraphed. A 6-3 board vote, with three pro-hike dissents and explicit guidance from Deputy Governor Himino that "real rates" are "extremely low," reduces gap risk. That argues for a smaller initial shock and a more orderly unwind.

Policy commitment. The 2024 BOJ paused after the unwind and walked back rhetorical hawkishness. The 2026 BOJ is publicly committed to continued rate hikes when data warrant them. That extends the runway for repatriation, even if the headline price move is smaller per hike.

Inflation regime. In 2024, inflation was demand-driven and cooling. In 2026, inflation is supply-driven (oil) and structural. A supply shock that anchors above target is harder for the BOJ to ignore than a demand impulse that is already fading.

Debt-and-yield context. Debt-to-GDP and yield levels are both higher today than in 2024. That makes the fiscal-premium component of yields larger. And critically, repatriation flows in 2026 are already happening before any trigger. The August 2024 episode had a step-function trigger. The 2026 setup is a slow burn that is already alight.

The implication

The historical pattern suggests two scenarios are most plausible. Scenario A is a slow grind: the BOJ delivers a 25-basis-point hike in June or July, the market absorbs it, and repatriation flows continue to leak into US duration over the next year. Term premium rises step by step, with no single dramatic event. Scenario B is a 2024-style unwind: the BOJ surprises with more hawkish guidance than expected, the yen rallies hard, the short positioning panics, and cross-asset correlations break for a session or two. Both scenarios are consistent with the same thesis. Both pressure long-duration assets, leveraged carry positions, and emerging-market carry; the difference is just velocity.

Each of the major asset classes responds to a different leg of the mechanism. The framing below is educational. It describes patterns that have historically held when these macro conditions emerge, and how institutional desks have typically been positioned in similar regimes. It is not a recommendation to buy or sell any specific security.

Equities

In past episodes when the BOJ tightened, driving a steeper JGB curve, Japanese domestic financials have historically outperformed Japanese exporters. The mechanism is intuitive: a steeper curve helps net interest margins for banks (the gap between what a bank earns on loans and pays on deposits), while a stronger yen pressures exporters' translated earnings. Outside Japan, the historical pattern points to pressure on carry-trade-correlated assets, namely US large-cap tech that benefits from cheap funding, emerging-market equities, and leveraged equity strategies. The Nikkei 225 closed above 65,000 for the first time on May 25, leaving it priced for a benign outcome rather than a 2024-style episode.

Rates and Fixed Income

Term premium tends to reprice higher across developed-market bond markets when a marginal, price-insensitive holder of duration exits. Yield curves typically steepen, and duration risk becomes asymmetric toward higher yields. Institutional allocators have historically responded to that asymmetry by shortening duration in core bond portfolios and rotating from long-end to short-end exposures.

Credit

Investment-grade spreads have historically widened modestly in this regime; high-yield spreads widen more due to a combination of growth concerns and forced deleveraging. The most exposed structure is levered credit funded by yen carry, because such positions face simultaneous pressure on both the carry and spread legs. Today, US investment-grade spreads (the extra yield corporate bonds pay over comparable Treasuries) sit near 74 basis points, in the lowest percentile of the historical distribution, and high-yield spreads are at 274 basis points, in the 9th percentile, classified as tight. The mean-reversion risk in both is asymmetric to wider spreads.

FX and Emerging Markets

Confirmed BOJ tightening has historically produced yen strength on a multi-week horizon, especially when positioning is one-sided. Emerging-market currencies and emerging-market sovereign bonds funded by yen carry have typically come under pressure first because they were the marginal carry destinations. The historical pattern points to long-yen exposure against a broader dollar and an emerging-market basket as the most direct expression of the thesis, with the caveat that crowded short-yen positioning introduces sharp two-way risk.

Commodities

Gold has historically been supported during bond-regime resets because central-bank diversification flows intensify when sovereign duration is in question. Gold currently trades near $4,500 per ounce, 16.5% off its January peak. The World Gold Council has documented a sustained 700 to 900 tonne annual official-sector bid. That bid is essentially a neither-bond-nor-dollar trade; a bond-regime rerating reinforces rather than undermines it. Oil sensitivity is dominated by the Iran outcome rather than by Japan dynamics, with Brent crude near $109 per barrel.

Counter-Argument 1: The BOJ delays the hike

The Bank of Japan has been described as "almost ready" to hike for the better part of eighteen months. The bar to actually pull the trigger is therefore higher than it looks. If an Iran ceasefire drives oil down by $20 per barrel, Japanese inflation could cool meaningfully and allow the BOJ to pause. With inflation moderating, the hedged-yield advantage shrinks, and the case for repatriation weakens in proportion.

The data tilts the other way. The base rate for following through on telegraphed BOJ moves is around 60%. The 6-3 board vote and explicit guidance from Himino push the conditional probability of a June or July hike toward 70%. Oil-related relief is a real swing factor, but it would need to be both large (a $20-plus drop) and sustained (multiple months) to fully neutralize the inflation impulse. The repatriation flow data also weakens this counter-argument: institutional rebalancing is already happening at current yields, before any further BOJ move. The flow is unlikely to fully reverse on a single dovish meeting.

Estimated probability counter-argument is correct: 25%

Counter-Argument 2: Other Treasury buyers backfill the Japanese exit

Foreign holdings of US Treasuries total roughly $9.2 trillion. Japan is only about 13% of the foreign pool. If Japan exits at the margin, plausible substitutes include US domestic banks (whose deposit base is growing again), other foreign reserve managers (China, Saudi Arabia, Gulf sovereigns), and the growing stablecoin and money-market complex, which has become a structural source of short-end Treasury demand. Backfill is the historical norm: large flow shifts almost always find a clearing buyer at some price.

Every plausible alternative buyer is also pulling back or constrained. The Fed's portfolio is flat under quantitative tightening, not absorbing duration. China's Treasury holdings have been declining for years rather than expanding. Stablecoin demand is concentrated in Treasury bills and the front end, not in 20- and 30-year durations. Some backfill is likely. A full offset at current yields is not. The Japanese flow need not fully evaporate to push the term premium higher; it only needs to shrink at the margin while every substitute also fades.

Estimated probability counter-argument is correct: 20%

Counter-Argument 3: The yen weakens despite the hike

If the BOJ delivers a 25-basis-point hike while the Fed holds and US growth outperforms Japan, the dollar-yen rate could rise counterintuitively on a growth-differential argument rather than a rate-differential argument. That would keep the carry trade alive, blunt the repatriation flow, and prevent the cross-asset unwind from triggering.

The 2024 episode actually showed the opposite: the yen strengthened on a confirmed BOJ hike because the carry unwind dominated growth-differential reasoning. The historical base rate favors yen strength following telegraphed BOJ tightening. The pattern could break if global risk-on persists and equity volatility stays compressed through the BOJ window. That is possible, but the cross-currents argue against it. Short yen positioning is at an all-time extreme, and equity volatility is already creeping up from its lows. A hike into a hot inflation print is far more credibly hawkish than the same hike into a cooling one.

Estimated probability counter-argument is correct: 15%

These are the seven indicators most likely to confirm or invalidate the thesis. Each has a current level, a bullish trigger (the move that would push conviction higher), a bearish trigger (the move that would force a reassessment), and a traffic-light status as of the report date.

If the 30-year JGB sustains a break above 4.00% in the weeks after the June 16 BOJ meeting, the thesis accelerates. If the BOJ holds in June and follows with dovish guidance in September, and if the dollar-yen rate trades back above 155, it is time to reassess. Three of the seven indicators are already in green status. None is red. That is the configuration the thesis predicts in its early innings.

Two models test the thesis claims against the historical record. The first asks whether Japanese repatriation reliably moves the US term premium; it does, with statistical significance. The second asks whether BOJ rate hikes systematically produce carry-unwind episodes in the August 2024 mold; the data says they do not, and the August 2024 episode looks driven by extreme positioning and surprise rather than the rate hike itself.

Model 1: Japan Repatriation and the US Term Premium

Model 1 tests the thesis elasticity that $100 billion of Japanese repatriation lifts the US 10-year term premium by roughly 10 to 15 basis points. It pairs the ACM 10-year term premium (a standard estimate of the term premium, published as FRED series THREEFYTP10) with Treasury TIC data on Japan's US Treasury holdings, quarterly from 2003 to the present (88 quarters), in an ordinary-least-squares regression of the term-premium change on the change in Japan's holdings, supplemented by an event study around quarters of net selling versus net buying, a lead-lag correlation scan, and a permutation test to confirm the coefficient is not an artifact of the sample. The result supports the thesis: p = 0.017, a point estimate of roughly 17 basis points of term premium for a $100 billion impulse, and a 95% confidence interval of [2, 32] basis points, inside which the 10 to 15 basis point claim sits.

Supporting findings:

Japan's UST holdings peaked at $1,303 billion in December 2021 and have since fallen to $1,001 billion. That is a $302 billion drawdown from the peak, the largest sustained selling stretch in the post-2003 sample.

In quarters with Japanese net selling, the term premium rose 64% of the time, with a median move of plus 4.9 basis points.

In quarters with Japanese net buying, the term premium rose 46% of the time, with a median move of minus 4.2 basis points.

The concurrent correlation between flow and term premium is r = -0.255. The sign is negative because Japanese selling shows up as a negative flow and is paired with a rising term premium.

Lag 0 is the only statistically significant lag in the cross-correlation scan. The effect is contemporaneous: it lands in the same quarter.

Figure 4. Japan's UST holdings plotted against the ACM 10-year term premium since 2003. The peak in late 2021 maps to the trough in term premium; the drawdown since maps to the term premium climb.

Figure 5. Quarterly change in Japan's UST holdings against the same-quarter change in the 10-year term premium. The downward-sloping fit is the visual statement of the model's coefficient.

Figure 6. Term premium moves in selling quarters versus buying quarters. The buying-quarter median sits below zero; the selling-quarter median sits well above it.

Figure 7. Lead-lag correlation between Japanese UST flow and the US term premium. Only lag 0 is statistically significant, which says the effect lands in the same quarter rather than leading or lagging.

The model explains only part of the picture. Its R-squared is 6.5%, so Japanese flows account for about 6.5% of the quarterly variation in the term premium. Japanese flow is one input among several; Fed policy expectations, fiscal deficits, and global risk sentiment all share the explanatory load. What the model does establish is that the Japanese flow channel is statistically real, and its sensitivity is in the ballpark claimed by the thesis. What it does not establish is that Japanese flow alone forecasts the term premium with precision.

Model 2: BOJ Rate Hikes and the Carry Unwind

Model 2 tests the framing that the August 5, 2024, episode (the Nikkei down 12% in a session, the S&P 500 down 3%, the VIX touching 65) is the systematic template for BOJ hikes, meaning BOJ tightening reliably produces yen strengthening, a carry unwind, and global equity stress. It studies the dollar-yen rate (FRED series DEXJPUS), the Nikkei 225, the S&P 500, and the VIX around the six BOJ rate-hike events between 2006 and 2025, in an event study at 1-day, 5-day, 10-day, 21-day, and 63-day windows around each hike, with bootstrap confidence intervals and a permutation test against random non-hike dates. The data does not support the claim of a systematic pattern, which clarifies how the August 2024 episode should be read: this report's Historical Precedent section treats it as the clearest recent example; the pattern does not recur reliably on every hike.

Key findings across the six-hike sample:

The yen strengthened in only 3 of 6 hikes at the 21-day horizon. The hit rate is 50%.

The Nikkei declined in 4 of 6 hikes at 21 days. That is directional, but not statistically significant in a sample this small.

Median dollar-yen 21-day change: -1.24%, with a 95% confidence interval of [-3.88%, +2.82%] and p = 0.268. The interval crosses zero.

Median Nikkei 21-day change: -1.79%, with a 95% confidence interval of [-4.65%, +5.05%] and p = 0.278. The interval crosses zero.

Median S&P 500 21-day change: -0.16%. The interval crosses zero.

The July 2024 hike (the run-up to the August 5 episode) saw the dollar-yen rate down 3.6% and the Nikkei down 1.2% over 21 days. The move was violent, but markets recovered within a month.

The January 2025 hike was the larger market mover in the sample: dollar-yen -4.2%, Nikkei -4.5%, S&P 500 -2.4%, VIX +4.1 points at 21 days.

The June 2025 hike went the other way. The yen weakened, equities rallied, and the market absorbed the move without stress.

Figure 8. Event study across six BOJ hikes, four assets, five horizons. The cross-event dispersion is the visual statement of the result: outcomes are not consistent enough to call this a template.

Figure 9. The July 2024 hike plotted against the other five hikes. The episode is an outlier in the sample, well outside the typical experience around a BOJ move.

Figure 10. Forward return distributions for the dollar-yen rate, the Nikkei, and the S&P 500 across the six hikes. The confidence intervals straddle zero on every series.

The mechanism is real: carry trades can unwind in response to BOJ moves. The data show that the unwind is not automatic. The August 2024 episode required crowded positioning plus surprise to produce that magnitude. The 2026 setup carries the positioning ingredient cleanly: CFTC asset-manager net positioning in the yen is at an all-time low of -10.56%. Whether the next BOJ move qualifies as a surprise depends on how well-telegraphed the June 16 decision actually is. With a 6-3 hold vote already on the board, the surprise component is less pronounced than in August 2024.

Read together, the two models cut in opposite directions, and that is the point. Model 1 strengthens the thesis: the repatriation-to-term-premium link is statistically real, and the elasticity sits within the stated range. Model 2 disciplines the thesis: a BOJ hike is not a clean systematic trigger for a 2024-style cross-asset unwind. The setup for the next BOJ move has the positioning ingredient locked in at the all-time low CFTC reading. The surprise ingredient is the more fragile leg. Holding the Historical Precedent framing to that asymmetry, treating August 2024 as a positioning-plus-surprise event rather than a mechanical consequence of the hike, is the honest reading of what these models say.

All data points in this report come either from the internal Benjamin Capital Research briefing trail (Weekly Macro Sitreps, Daily Macro News Briefs, and the macro briefing JSON) or from the public sources listed below. Where this report cites yields, holdings, or auction data, the figures match the corresponding internal memo to the basis point. Where this report references sell-side narrative or consensus pricing, those are descriptive characterizations of public commentary, not direct attributions to proprietary research products.

Methodology Notes

Hedged-yield calculation. This report uses the simple convention of approximating the FX hedge cost as the front-end policy-rate differential (Fed Funds minus the BOJ policy rate). In practice, a forward-points calculation is closer to the cross-currency basis, but for a directional comparison, the policy-rate differential is the standard institutional shortcut.

Repatriation flow estimates. Q1 2026 Japanese net foreign-bond sales of $29.6 billion combine Treasuries and other foreign sovereign bonds. The TIC data for US Treasury holdings has a two-month lag and was last released through February 2026 as of the report date, which is why the tracker shows $1.24 trillion for February 2026.

Asset-class framing. All asset-class commentary describes historical patterns and educational framing. It is not a recommendation regarding any specific security, fund, or ticker.

Models. The two quantitative models use public series: the ACM 10-year term premium (FRED THREEFYTP10), Treasury TIC holdings, the dollar-yen rate (FRED DEXJPUS), the VIX (FRED VIXCLS), the S&P 500, and the Nikkei 225. Statistical inference uses bootstrap confidence intervals and permutation tests; the small BOJ-hike sample (six events) limits the power of the Model 2 tests, which is reflected in the NOT SUPPORTED verdict.

This report is for informational and educational purposes only. It does not constitute investment advice, a recommendation, or a solicitation to buy or sell any security. All asset class commentary reflects historical patterns and educational analysis, not personal investment advice. Past performance does not guarantee future results. Readers should consult a qualified financial advisor before making investment decisions.

Benjamin Capital Research | June 6, 2026