Executive Summary: Why Japan's $1.1 Trillion Treasury Position Matters in 2025
Japan holds $1.1 trillion in U.S. Treasuries—the largest foreign holder. For decades, this was stable. Japanese institutions bought Treasuries because domestic yields were zero or negative, and dollar assets offered both yield and currency appreciation.
That arbitrage is ending. Eventually.
What changed:
• JGB 10-year yields: 1.82% (highest since 2013)
• Bank of Japan officially ended negative rates in March 2024
• Currency-hedged Treasury returns turned negative in October 2025 (though recent UST sell-off has narrowed the gap to essentially flat as of November 22)
• Japanese life insurers and pension funds facing first attractive domestic alternatives in 30 years
The critical reality check:
• TIC data through October 2025: Japan has been buying Treasuries at +$10–15B monthly avg through H1 2025, slowing to +$5–10B in Q3
• Hedged return differential widening (~280 bps favoring JGBs); USD/JPY at 155.83 (November 27)
• Historical precedent (2016-2018) showed minimal repatriation when JGB yields improved
• These institutions move on 2–3 year planning cycles, not quarters
Consensus estimates (JPM, Goldman, Nomura): 5–12% reduction over 2–3 years ($55–135 billion total), starting sometime in 2026 at earliest—not Q1 2026, possibly not until late 2026 or 2027.
Why it still matters: Even if repatriation is slow and gradual, the directional shift from consistent buyer (+$15–20B monthly) to neutral or modest seller creates a $30–50B annual swing in marginal Treasury demand. That's enough to pressure 10-year yields 15-25 bps higher over 12–18 months.
The positioning:
1. Curve steepener (5s30s): Modest size (2–3% capital), 12–18 month horizon
2. Japanese bank call options: Small lottery ticket (1–2% capital) on rising JGB yields benefiting banks
3. Dollar/yen: Tactical only—fade extreme moves rather than directional bets
Key insight: This is a prepare and wait theme, not a "load up now" trade. The math suggests repatriation should happen. History suggests Japanese institutions will take their sweet time deciding. Position small, watch the data, add when flows confirm.
This isn't a Q1-Q2 2026 story. It's a 2026-2028 story. Patience required.
Japanese Treasury Holdings Explained: Who Owns the $1.1 Trillion
Japan Treasury Holdings Data: October 2025 TIC Report
Japanese holdings (October 2025 Treasury data):
• Total U.S. Treasury securities: $1.13 trillion
• Percentage of total foreign holdings: 15.2%
• Percentage of publicly traded U.S. debt: 4.1%
Recent flow data (TIC releases through October 2025 + November preliminary):
• 2024 average monthly purchases: +$12-15 billion
• H1 2025 average: +$10-12 billion (Japan remains a buyer, not seller)
• Q3 2025: Pace slowed to +$5-10 billion monthly (e.g., +$7B October, +$5B September)
• November 2025 (preliminary): +$3-5 billion estimated — continued buying, no acceleration of selling
• 2025 YTD cumulative: +$45-50 billion (vs. +$180B full-year 2024)
Key observation: Despite JGB yields surging 62 bps in November (1.20% → 1.82%), preliminary flow data shows Japanese institutions continued buying Treasuries. The slowing pace (+$3–5B vs. +$10–15B earlier) suggests caution, but there is no evidence of net selling beginning. Institutional friction remains dominant over yield arbitrage—for now.
Japanese Treasury Flows vs. JGB Yields (2024–2025):
Quarter | Monthly Avg Flow ($B) | JGB 10Y (%) | Trend |
Q1 2024 | +$15B | 0.70–0.75% | Strong buyer |
Q2 2024 | +$12B | 0.85–1.00% | Steady buyer |
Q3 2024 | +$12B | 0.90–1.00% | Steady buyer |
Q4 2024 | +$10B | 1.00–1.10% | Slight slowing |
Q1 2025 | +$12B | 1.10–1.15% | Steady buyer |
Q2 2025 | +$10B | 1.15–1.20% | Modest slowing |
Q3 2025 | +$7B | 1.20% | Clear slowing |
Q4 2025 (prelim) | +$4B est. | 1.82% | Near inflection? |
Pattern: As JGB yields rose from 0.70% to 1.82%, monthly purchases declined from +$15B to +$4B—but never turned negative. The correlation is evident but the inflection to net selling hasn't occurred. Source: TIC monthly releases; JGB yields from Bloomberg.
Critical reality check: Despite JGB yields rising to 1.82% and hedged Treasury returns turning negative, Japanese institutions have continued buying Treasuries through October 2025 (+$45B YTD). Summer 2025 showed slowing pace (+$5–10B monthly vs. +$12-15B earlier), but flows remain net positive. This confirms the "extremely sticky" characterization.
Who holds them in Japan:
• Life insurance companies: ~40% ($452B)
• Pension funds (GPIF and corporate): ~30% ($339B)
• Banks (especially regional): ~20% ($226B)
• Other institutions: ~10% ($113B)
Why this composition matters: These are long-term, buy-and-hold institutions with specific liability-matching requirements. They don't trade actively. When they allocate, they allocate for years. And when they reallocate, they move glacially.
The fact that they've continued buying through 2025 despite improved domestic alternatives tells you everything about their pace of decision-making.
JGB Yields 2025: How Bank of Japan Policy Normalization Changes Everything
10-Year JGB Yield History: From Negative Rates to 1.2%
JGB 10-Year Yield Progression (2020–2025):
Date | 10Y JGB | Change | Event / Context |
2020 | -0.10% | — | Peak negative rates |
Dec 2022 | 0.25% | +35 bps | BOJ widens YCC to ±0.50% |
Oct 2023 | 0.50% | +25 bps | BOJ loosens YCC further |
Mar 2024 | 0.75% | +25 bps | BOJ ends negative rates |
Jul 2024 | 1.00% | +25 bps | BOJ hikes to 0.25% |
Early Nov 2025 | 1.20% | +20 bps | Pre-November surge |
Nov 26, 2025 | 1.82% | +62 bps | Highest since July 2013 |
November's move is remarkable: +62 bps in a single month (1.20% → 1.82%). For context, JGB 10Y moved only +70 bps in all of 2024. This acceleration suggests markets are pricing faster BOJ normalization than official guidance implies. If this pace continues, JGB yields could reach 2.0–2.5% by mid-2026—dramatically strengthening the repatriation case.
Currency Hedging Costs 2025: Why Japanese Institutions Face Negative Returns
Let's do the math from a Japanese life insurer's perspective:
Hedged Treasury returns (November 27, 2025):
• 10-year Treasury yield: ~4.01%
• 3-month USD/JPY basis swap (hedge cost): ~4.15–4.35%
• Net hedged return: approximately -0.34% (persistently negative)
Domestic alternative:
• 10-year JGB yield: 1.82%
• Currency risk: Zero (liability-matched)
• Net return: +1.82%
Current differential: ~216 bps hedged basis; ~280 bps unhedged basis favoring JGBs
Critical context: Hedged Treasury returns have been persistently negative or near-zero throughout H2 2025, fluctuating between -0.50% and +0.10% depending on daily yield/basis moves. The key insight isn't any single snapshot—it's that the structural advantage has durably shifted to favor JGBs for the first time since 2007. November's JGB surge (1.20% → 1.82%) widened this advantage further, but the trend was established before that move.
What this means for timing:
• The repatriation incentive has strengthened materially as JGB yields surged in November
• At 1.82% JGB yields, the case for domestic allocation is now compelling on pure return basis
• However, institutional decision cycles remain slow—actual repatriation flows lag yield moves by 6–12 months
• Yen appreciation risk: If USD/JPY moves sharply below 150 (currently 155.83), unhedged FX losses could create urgency that overrides institutional friction
Bottom line — Monitoring over prediction: The fundamental case for repatriation is now stronger than at any point in the past decade. But preliminary November data shows Japan still buying (+$3–5B estimated). The value in this analysis is identifying what to watch (TIC data, JGB yields, USD/JPY, BOJ policy) rather than predicting exactly when flows turn. Institutional friction (liability matching, switching costs, muscle memory) remains a powerful buffer—until it isn't.
Japanese Repatriation Mechanics: How and When Capital Flows Home
Will Japan Sell US Treasuries? Consensus Estimates for 2026-2028
First, let's be clear: Japan isn't dumping Treasuries. There's no crisis, no political motivation, no emergency.
This is gradual portfolio reallocation driven by:
1. Improving domestic yield alternatives
2. Currency hedge costs making dollar assets uneconomical
3. Asset-liability matching favoring yen-denominated bonds
4. Regulatory encouragement for domestic capital deployment
Base case assumption (consensus sell-side estimates): 5–12% reallocation over 2–3 years
• Amount: $55–135 billion total
• Pace: $2-4 billion monthly starting Q2 2026 (Japanese fiscal year begins April 1)
Critical context: Japanese life insurers are extremely slow and sticky. Historical precedent from 2016–2018 (when JGB yields rose to 0.10–0.15%) showed almost no net selling despite improved domestic alternatives—in fact, Japan added +$150B to Treasury holdings during that period. These are buy-and-hold institutions with decade-long investment horizons. The BOJ balance sheet remains at ¥696T (~$4.47T at ¥155.83/USD)—still extraordinarily large despite modest reductions from the ¥720T peak.
Why even modest repatriation matters: It's not the absolute size—it's the directional change. For 30 years, Japan was a consistent marginal buyer (+$30-80B annually). Shifting to neutral or modest seller (-$20-40B annually) is a $50–120B swing in the marginal buyer/seller dynamic.
Upside Risk Scenarios: What Could Accelerate Repatriation
The base case (5–12% over 2–3 years) assumes gradualism. But several scenarios could accelerate flows materially:
Scenario A: Faster BOJ Rate Hikes (Goldman scenario)
Goldman Sachs' upside case projects BOJ policy rate reaching 1.0% by end-2026 (vs. consensus 0.5%). If JGB 10Y yields rise to 2.5%+, the hedged return differential could widen to 300+ bps, creating overwhelming incentive for repatriation. Probability: 20–25%.
• Impact: Could accelerate timeline to 2026-2027 with 15–20% reallocation ($170–230B)
• Monitoring: BOJ policy statements, Tokyo CPI prints, wage negotiation outcomes (Shunto)
Scenario B: Sharp Yen Appreciation (USD/JPY < 140)
Current USD/JPY ~155.83 (November 27). If yen strengthens sharply toward 140 or below—whether from Fed cuts, risk-off sentiment, or BOJ hawkishness—unhedged Treasury losses compound. Japanese institutions may accelerate repatriation to avoid further FX losses. Probability: 15–20%.
• Impact: Creates urgency; could compress 3-year timeline into 12–18 months
• Monitoring: Fed policy trajectory, USD/JPY technicals, MOF intervention signals
Scenario C: Japanese Regulatory Push for Domestic Allocation
Japanese FSA or GPIF could signal preference for domestic asset allocation as part of broader economic policy. Soft guidance to life insurers and pensions to support domestic bond market could accelerate repatriation beyond pure economic incentives. Probability: 10–15%.
• Impact: Policy-driven repatriation could override institutional stickiness
• Monitoring: FSA statements, GPIF allocation reviews, LDP economic policy
Black Swan Mitigation: If multiple upside scenarios converge (faster BOJ + yen appreciation + regulatory push), repatriation could reach 20–25% ($230–285B) over 18–24 months—materially larger than base case. Position sizing must account for this tail risk. Keep total theme exposure at 4–6% of portfolio maximum. Use options for leveraged expressions to define maximum loss. Monitor the upside triggers as closely as the base case.
Q1 2026 Flow Inflection: Quantifying Acceleration Risk
November's JGB surge (+62 bps to 1.82%) combined with yen stability (¥155.83) has amplified the repatriation incentive to ~280 bps. The critical question: what triggers the shift from 'slowing purchases' to 'net selling'?
BOJ Policy Path — Market Pricing (November 27, 2025):
• January 2026 BOJ hike probability: ~60% (CME/Bloomberg consensus)
• Implied policy rate by June 2026: 0.50–0.75% (vs. current 0.25%)
• Goldman upside case: 50 bps cumulative hikes in 2026 (to 0.75%)
Flow Inflection Scenario Analysis:
Scenario | JGB 10Y Target | Flow Impact | Probability |
Base: No Jan hike | 1.80–2.00% | Continued +$3–5B/mo | 40% |
Central: Jan hike 25bp | 2.00–2.20% | Flat to -$2B/mo | 45% |
Hawkish: 50bp by H1 | 2.30–2.50% | -$3–5B/mo begins | 15% |
Key inflection point: If BOJ hikes in January AND JGB 10Y breaches 2.20%, the hedged return differential widens to ~350 bps—a threshold that historically triggers institutional allocation reviews. Watch for TIC data showing first negative monthly print as confirmation.
Q1 2026 Tail Risk Quantification:
BOJ hike odds stand at ~60% for January 2026 (CME FedWatch / Bloomberg consensus). If the hawkish scenario materializes (50 bps cumulative hikes by H1 2026), Japanese institutions could shift from +$4B monthly buyers to -$4B monthly sellers—a $96B annual swing in Treasury demand. Combined with continued China reduction (-$25B/yr) and ECB QT pressures, Q1 2026 could mark the beginning of a structural foreign demand deficit for Treasuries—potentially +$50B net outflow in Q1 alone if acceleration triggers cascade behavior.
Tail scenario (15% probability): BOJ hikes 50 bps by June 2026 + JGB 10Y reaches 2.50% + USD/JPY breaks below 150 = potential -$50B Japanese outflow in Q1 2026 alone. This would represent the largest quarterly foreign selling since 2022 and could add 30–40 bps to 10Y yields in a compressed time frame. Position sizing must account for this acceleration risk.
Global Context: Foreign Demand Fragility Beyond Japan
Japan isn't the only foreign holder facing repatriation incentives. The broader context of synchronized global QT creates systemic fragility in Treasury demand:
ECB QT Parallel:
The ECB has shed €3.2 trillion (44% reduction) since mid-2022 peak. European institutions face similar hedged-return calculations—though EUR/USD hedging costs are lower than JPY/USD, Bund yields have also risen. The September 2022 UK gilt crisis demonstrated how quickly 'gradual repatriation' narratives can accelerate when market structure meets policy shock.
UK Gilt Blowout Precedent:
The Truss mini-budget triggered 100+ bps gilt yield spikes in days—not because fundamentals changed, but because leveraged LDI structures amplified initial moves. Japanese life insurers don't use similar leverage, but the precedent shows how 'sticky' institutional flows can become 'unsticky' rapidly when thresholds breach.
China Holdings (~$770B):
China has reduced Treasury holdings from $1.1T (2021) to ~$770B—a steady -$20–30B annual pace driven by geopolitical diversification rather than yield differentials. Unlike Japan, China's reduction is structural and unlikely to reverse regardless of rate environment.
Systemic implication: If Japan shifts from buyer to seller while China continues steady reduction and European institutions face their own repatriation math, the marginal Treasury buyer increasingly becomes domestic: MMFs, banks, hedge funds, and eventually the Fed (if crisis emerges). This isn't 2015 anymore—foreign demand cannot be assumed.
Yield Pressure Sensitivity: Historical Backtest
The 15–25 bps yield pressure estimate draws from TIC flow sensitivity analysis. Here's the historical relationship:
Period | Japan Flow (Annual) | 10Y Yield Move | Implied Sensitivity |
2016–2018 | +$50B/yr avg | +85 bps (1.8→2.65%) | Buyer = rates rose anyway |
2022 (QT start) | +$20B | +235 bps (1.5→3.85%) | Fed QT dominated |
2023 (bank crisis) | +$80B (flight to safety) | -50 bps (3.9→3.4% trough) | ~6 bps per $10B flow |
2026E (base case) | -$25–40B | +15–25 bps | ~5–6 bps per $10B |
Methodology note: The ~5–6 bps per $10B sensitivity is derived from 2023 TIC flow reversals during the banking crisis, when Japanese flight-to-safety buying visibly compressed yields. The relationship is nonlinear—larger flows have disproportionate impact due to liquidity constraints—but provides reasonable order-of-magnitude estimate.
How to Trade Japanese Repatriation: Three Setups for 2026-2028
Trade #1: Dollar/Yen Forecast 2026 — Tactical Mean Reversion Strategy
Revised approach: Given that repatriation hasn't materialized yet (Japan still buying +$15–20B monthly through October 2025), the original two-phase directional trade is premature.
Better strategy: Fade extremes
If dollar/yen rallies to ¥155–160:
• Consider small short position (1–2% capital)
• Rationale: Extreme dollar strength creates optionality for eventual repatriation theme
• Stop loss: ¥162 | Target: ¥150-152 | Timeline: 3-6 months
If dollar/yen falls to ¥145–147:
• Consider small long position (1–2% capital)
• Rationale: Extreme yen strength likely met with MOF intervention concerns
• Stop loss: ¥143 | Target: ¥150-152 | Timeline: 3-6 months
Key change: This is mean-reversion trading around ¥150 fair value, NOT a directional multi-phase trade. The repatriation theme creates asymmetry at extremes but doesn't justify large directional bets yet.
Position size: 1–2% of capital maximum | Conviction level: Low-to-moderate | Patience required: High
Trade #2: Treasury Yield Curve Steepener (5s30s) — Positioning for Japanese Selling
Thesis: IF Japanese repatriation materializes (big if, given they're still buying), it will concentrate in 10-year sector where institutions hold duration-matched assets.
Trade structure:
• Long: 30-year Treasury bonds (current yield: 4.38%)
• Short: 5-year Treasury notes (current yield: 3.92%)
• Current spread: 46 basis points
• Target spread: 58–63 basis points (spread widens by 12–17 bps over 12–18 months)
Reality check: This is a very gradual theme. Even if repatriation begins in 2026, flows will be modest ($2–4B monthly per consensus). The curve move will be slow and subject to many other factors.
Revised positioning:
• Position size: 2–3% of capital (down from 4–6%)
• Timeline: 12–18 months minimum (this is 2026-2028 theme, not 2026 trade)
• Stop loss: Spread tightens below 40 bps sustained
Trade #3: Japanese Bank Stocks (MUFG, Mizuho) — Rising JGB Yield Beneficiaries
Thesis: IF Japanese repatriation accelerates and JGB yields continue rising, Japanese regional banks benefit from rising domestic yields (improved net interest margins).
Specific opportunities:
• Mitsubishi UFJ Financial (MUFG): Largest holder of foreign securities
• Mizuho Financial: Significant Treasury holdings
• Regional bank index (TOPIX Banks): Leveraged to domestic yield curve steepening
Trade structure:
• Buy: 9–12 month call options (Sep 2026 or Dec 2026 expiry) on MUFG ADRs
• Strike: 5-10% out-of-the-money (reduces premium cost)
• Sizing: 1–2% of total portfolio capital in premium
Realistic expectations: If thesis plays out: 100–200% gain on options (2–4x return on 1–2% capital = 2–8% portfolio gain). If thesis doesn't materialize: Lose 1–2% premium, manageable.
FAQ: Japan Treasury Holdings and Repatriation Risk
How much US debt does Japan own in 2025?
Japan holds approximately $1.14 trillion in U.S. Treasury securities as of end-October 2025 (up from $1.10T end-September), making it the largest foreign holder of US government debt. This represents 15.2% of all foreign-held Treasuries and approximately 4.1% of total publicly traded U.S. debt. Japanese life insurance companies hold roughly 38–42% (~$450B), pension funds including GPIF hold 28–32% (~$340B), banks hold 18–22% (~$230B), and other institutions hold the remainder.
Why would Japan sell US Treasuries?
Japanese institutions may reduce Treasury holdings because domestic Japanese Government Bond (JGB) yields have surged to 1.82%—the highest since 2013—while currency hedging costs make dollar-denominated Treasuries unattractive. When Japanese investors hedge their dollar exposure back to yen, the cost is approximately 4.15–4.35% annually, which means a 4.01% Treasury yield nets approximately -0.34% after hedging (negative return). Meanwhile, JGBs yield 1.82% with zero currency risk. This ~216 bps hedged advantage (~280 bps unhedged) favoring domestic bonds creates significant incentive to repatriate capital.
What is the JGB yield in 2025?
The 10-year Japanese Government Bond (JGB) yield reached 1.82% on November 26, 2025, the highest level since 2013. Notably, JGB yields surged from 1.20% in early November to 1.82% by month-end—a 62 basis point move in just weeks. This follows the Bank of Japan's policy normalization, including ending negative interest rates in March 2024 and gradually allowing yields to rise from -0.10% in 2020. The sharp November move suggests accelerating normalization expectations.
Will Japanese repatriation crash the Treasury market?
No. Consensus estimates from JPMorgan, Goldman Sachs, and Nomura suggest gradual reallocation of 5–12% over 2–3 years ($55–135 billion total), not a sudden liquidation. Japanese life insurers and pension funds are extremely slow-moving, buy-and-hold institutions that work on multi-year planning cycles. Historical precedent from 2016–2018, when JGB yields rose to 0.10–0.15%, showed almost no net selling despite improved domestic alternatives. Even modest repatriation would likely raise Treasury yields by 15–25 basis points over 12–18 months—not cause a crisis.
How does Japanese repatriation affect the dollar/yen exchange rate?
When Japanese institutions sell Treasuries and convert dollars to yen for domestic investment, it creates selling pressure on the dollar and buying pressure on the yen. If repatriation materializes at scale ($2–4B monthly), dollar/yen could strengthen from current levels around ¥149–150 toward ¥145–148 over 12–18 months. However, many competing factors influence dollar/yen including Fed policy, US economic outperformance, risk sentiment, and potential Japanese Ministry of Finance intervention. The relationship is not mechanical.
What is the Bank of Japan doing with its balance sheet?
The Bank of Japan's balance sheet stands at ¥696T (~$4.47T at ¥155.83/USD) as of November 25, 2025, down modestly from a peak of ~¥720T. JGB holdings have declined approximately 6% from peak. This follows the BOJ ending negative interest rates in March 2024 and gradually normalizing policy after decades of extraordinary accommodation. Governor Ueda has signaled continued gradual normalization. The BOJ's balance sheet remains extraordinarily large—still the world's largest relative to GDP—but the direction is clearly toward normalization.
When will Japanese Treasury selling begin?
Despite improved domestic alternatives, Japanese institutions have continued buying Treasuries through October 2025—though at a slowing pace (+$5–10B monthly in Q3 vs. +$10–15B in H1). Cumulative 2025 purchases total +$45B YTD (vs. +$180B full-year 2024). Actual repatriation has not yet begun. Consensus estimates suggest potential net selling could begin in Q2 2026 at earliest (after Japanese fiscal year begins April 1), but given institutional stickiness, meaningful sustained outflows may not materialize until late 2026 or 2027. Watch monthly TIC data for confirmation.
How do I track Japan's Treasury holdings?
The U.S. Treasury International Capital (TIC) system releases monthly data on foreign holdings of U.S. securities with approximately a 6-week lag. Data is available at treasury.gov/tic. Watch for consistent monthly declines of $3–5B or more for 3+ consecutive months as confirmation that repatriation is materializing. Additionally, the Japanese Ministry of Finance publishes Balance of Payments statistics showing portfolio investment flows by asset class.
What are the best trades for Japanese repatriation?
Three positioning strategies for this theme: (1) Dollar/yen mean reversion: Fade extreme moves—small short at ¥155–160, small long at ¥145–147, with 1–2% position sizes. (2) Treasury 5s30s steepener: Long 30-year, short 5-year Treasuries targeting 12–17 bps spread widening over 12–18 months with 2–3% position size. (3) Japanese bank call options: 9–12 month calls on MUFG or Mizuho ADRs with 1–2% premium exposure as rising JGB yields benefit bank net interest margins.
What happens to Treasury yields if Japan sells?
If Japanese repatriation materializes at consensus estimates ($2–4B monthly), 10-year Treasury yields would likely rise 15-25 basis points over 12–18 months as the market absorbs reduced foreign demand. This would push 10-year yields from current 4.44% toward 4.60–4.70% range. The move would concentrate in the 7–10 year sector where Japanese institutions hold duration-matched assets, potentially steepening the yield curve. This is normal price discovery, not a crisis—Treasury markets have absorbed larger supply shocks.
Conclusion: Japanese Treasury Selling — When Patience Beats Precision
The fundamental analysis is sound: Japanese institutions face the worst hedged Treasury returns relative to domestic alternatives in decades. The math says repatriation should happen.
But the data says it hasn't happened yet.
Through October 2025, Japan continued buying Treasuries—+$45B YTD, though at a slowing pace (+$5–10B monthly in Q3 vs. +$10–15B in H1). Preliminary November data suggests +$3–5B—still buying, not selling. Despite JGB yields surging to 1.82% (highest since 2013). Despite hedged Treasury returns turning definitively negative (~-0.34%). Despite all the economic rationale for reallocation.
Why? Because these are the slowest-moving pools of capital on earth.
Life insurance companies and pension funds don't pivot on quarterly data. They work on 2–3 year strategic asset allocation reviews. They have enormous switching costs. They prioritize stability over optimization. And they've been buying Treasuries for 30 years—that's institutional muscle memory.
The 2016-2018 precedent is instructive: JGB yields rose to 0.10–0.15%. Everyone predicted massive repatriation. Instead, Japan added +$150B to Treasury holdings. Japanese institutions shrugged and kept buying.
So what's the trade?
Not: "Load up now for imminent Q1-Q2 2026 repatriation"
Instead: "Position small, prepare infrastructure, watch the data obsessively, add when flows confirm"
Three rules for this theme:
1. Size tiny (2–3% total across all trades)—this could take 2–3 years to play out, or might not happen at all
2. Watch monthly TIC data religiously—if Japan goes from +$5–10B monthly buyer to -$5B monthly seller, that's your signal to add
3. Extend timeline to 2026-2028—this isn't a quarterly trade, it's a multi-year positioning theme
The asymmetry is still attractive: If repatriation accelerates in 2026-2027, small positions compound into 20–30% gains over 18–24 months. If repatriation doesn't materialize, small losses, easy to cut, preserve capital for themes that do develop.
But acknowledge the tail risk: If upside scenarios converge (faster BOJ hikes + yen appreciation + regulatory push), repatriation could reach 20–25% ($230–285BB) in 18–24 months—2-3x base case. This would vindicate the thesis spectacularly but also move faster than gradual positioning allows. Keep some optionality (calls on Japanese banks, tail hedges on USD/JPY) to capture acceleration scenarios without requiring perfect timing.
Patience beats precision on multi-year institutional themes.
Legal Disclaimer
Data current to: November 27, 2025. TIC holdings through October 2025 (November 18 release); November flows preliminary estimates; JGB yields November 26 close (1.82%); USD/JPY spot November 27 (¥155.83); BOJ balance sheet November 25 (¥696T).
This content represents research and analysis, not investment advice. TMS Capital Research shares our perspective on liquidity markets and positioning considerations. You bear full responsibility for your investment decisions.
Critical reminders:
• Markets can remain irrational longer than you can stay solvent
• Size positions appropriately for your risk tolerance
• Use stop-losses on directional trades
• Don't be a hero with concentrated positions
• Never YOLO your retirement account on any single trade idea
Be smart. Manage risk. Think in probabilities, not certainties.
About TMS Capital Research
Mission: Liquidity-driven macro analysis. Actionable trade ideas. No BS.
What we do: Track Federal Reserve operations, global central bank balance sheets, money market conditions, and the metrics that actually drive market movements.
What we don't do: Doom-mongering, hype cycles, vague predictions, or weekly previews that are obsolete by Tuesday.
Follow on Twitter: @TMSCapResearch
Data Sources and Methodology
Primary Data Sources:
• U.S. Treasury International Capital (TIC) System — Monthly foreign holdings data
• Japanese Ministry of Finance — Balance of Payments statistics
• Bank of Japan — Financial statements, policy board minutes, balance sheet data
• Federal Reserve H.4.1 — Weekly balance sheet releases
• Bloomberg Terminal — Real-time JGB yields, FX rates, Treasury curves
• Sell-side research — JPMorgan, Goldman Sachs, Nomura consensus estimates
Methodology: We track cross-border capital flows, currency hedging costs, and yield differentials to identify when mechanical portfolio rebalancing may occur. This analysis combines fundamental flow analysis with historical precedent and positioning data. Consensus estimates are weighted toward conservative assumptions given Japanese institutional stickiness.
Related Articles:

