Japan: can the return of savings raise borrowing costs elsewhere?

The return of Japanese savings makes for a compelling story: higher yields at home could draw capital away from overseas markets and raise other borrowers’ funding costs. Documents available on 22 October 2026 make that mechanism plausible, while describing a much more selective transition. Its pressure point lies in the reinvestment decision, when the proceeds of a maturing investment reach an account and several markets compete for their next use.

The figure that brought the question back into focus comes from Japan’s Ministry of Finance. In September, life insurers in its reporting panel recorded JPY288.6 billion in net disposals of foreign medium- and long-term debt securities. The release was published on 8 October. Disposals include securities redeemed at maturity as well as sales; acquisitions are deducted to arrive at the balance. The statistic therefore records net disinvestment in a category of securities. Establishing where the money went requires further evidence. ^{1, 3}

Our earlier analysis of Japanese life insurers’ balance sheets examined how interest-rate risk enters their accounts. This investigation follows the next stage: reported flows and the decision to reinvest overseas or in Japan. That distinction defines the investigation. An insurer can let a US bond mature, buy foreign equities, change its currency hedge or choose a Japanese government bond. Each decision shifts a different set of risks. Assessing the implications for global borrowing costs means following the return received in yen, the promise made to the policyholder and the investor available to take the insurer’s place.

September extends an earlier shift

The monthly series puts September in perspective. Life insurers in the panel recorded JPY1,757 billion in net disposals during the first nine months of 2026, compared with JPY314.9 billion in the corresponding period of 2025. February and March 2026 had each exceeded JPY600 billion. These totals add the published monthly balances for foreign medium- and long-term debt, with redemptions included in disposals. ^{2, 3}

Method and scope

Foreign medium- and long-term debt held by life insurers in the MOF designated reporting panel, October 2025 to September 2026. Published monthly balances in JPY billion, acquisitions minus disposals; disposals include redemptions. The January–September total sums already rounded published net values. The destination of proceeds requires additional data. Sources released on 8 October 2026.

The flows document net disinvestment in these securities. Reconstructing the underlying investment decisions requires retaining the survey’s scope. The ministry collects reports from designated institutions and classifies securities by the issuer’s residence. The foreign classification therefore does not, by itself, identify dollar exposure. The table does not identify each investment’s destination country, currency or hedge ratio. A negative balance cannot directly identify either US Treasury sales or purchases of yen. ^{1, 3}

Other investor categories make a single national narrative harder to sustain. From January to September, banks’ own banking accounts recorded approximately JPY11,739 billion in net disposals, while trust banks’ trust accounts recorded approximately JPY11,415 billion in net acquisitions of the same category of foreign debt. The aggregate conceals substantial movements in opposite directions. Trust accounts describe an institutional account category; the entire amount cannot be attributed to pension funds from this series alone. ^{2, 5}

Method and scope

January–September 2026, the same class of securities, JPY billion. Banks’ own accounts, life insurers and trust banks’ trust accounts have distinct scopes. Trust accounts can serve several types of beneficiary; that breakdown is outside this table. The seven displayed categories cover the panel, with several categories grouped under “other”. Their sum may differ slightly from the published total because of rounding.

That diversity matters for international risk. A bank reducing its portfolio, an insurer seeking a particular payment date and a trustee rebalancing assets for clients may trade different securities on different schedules. Adding them together measures a balance. Understanding prices requires identifying the segments involved. A withdrawal concentrated in a less liquid maturity can put more pressure on its yield than a larger flow spread across several deep markets.

The relevant return arrives in yen

Japan’s domestic market now offers substantial reference points. In the finance ministry’s official curve for 21 October 2026, the constant-maturity yield was 2.923% at ten years and 3.980% at thirty years. These are dated analytical benchmarks, calculated under the ministry’s methodology from secondary-market reference prices. They differ from an issue coupon or a price immediately available for execution. ^{6, 7}

For an insurer paying claims in yen, a dollar bond’s quoted yield is the beginning of the calculation. The eventual value also depends on the exchange rate at which dollars are converted. Yen appreciation reduces the amount received in the currency of the liabilities; depreciation increases it. A currency hedge can fix an exchange rate in advance for a specified amount and date by selling the expected dollars forward. Much of its economic cost is embedded in that forward rate, alongside transaction costs and, depending on the arrangement, collateral requirements. ⁹

Consider a wholly hypothetical example. An investor exchanges JPY15,000 for USD100, then buys a bond paying USD105 in one year, assuming contractual repayment. The investor hedges the entire USD105. At a forward rate of JPY147 per dollar, the proceeds are JPY15,435, a 2.90% return before costs. A forward rate of JPY148.5 produces JPY15,592.5, or 3.95%. A hypothetical yen reference investment returning 3% falls between the two. These numbers illustrate arithmetic; they are neither October market quotes nor a comparison of equivalent credit risks.

Method and scope

Entirely hypothetical one-year example: JPY15,000 invested, spot exchange rate JPY150/USD, contractual final payment of USD105 fully hedged. Hypothetical forward rates of JPY147 and JPY148.5/USD produce JPY15,435 and JPY15,592.5, returns of 2.90% and 3.95% before fees and default. The 3% yen benchmark is fictional. The comparison isolates currency conversion, leaving credit differences outside the calculation.

The exact calculation is straightforward when the final payment is known and fully hedged: multiply the dollar return factor by the ratio of the forward exchange rate to the initial spot rate. In the first scenario, 1.05 × 147 / 150 − 1 = 2.90%. Simply subtracting two percentages would produce a different approximation. Hedging changes the ranking because it changes the amount of yen ultimately recovered.

A thirty-year bond hedged with three-month contracts introduces another complication. The price of the next hedge renewal remains unknown, while the bond continues to respond to changes in yields. Carry calculated today indicates income under current conditions; the outcome over thirty years depends on successive hedges, prices, costs and payments. The currency contract’s term and the bond’s interest-rate sensitivity are separate dimensions of risk. ^{9, 10}

Higher Japanese rates work in two directions

On 18 September, the Bank of Japan set its overnight call-rate target at around 1.25%, effective from 24 September. The decision helps explain monetary normalisation, but a three-month hedge is priced using market interest rates for the corresponding tenor, financing conditions and the market-specific pricing adjustment known as the cross-currency basis. An executable hedge quote therefore requires market prices for the relevant maturities. ^{8, 9}

This produces an important counterweight. Other things equal, a rise in short-term yen rates narrows the gap with short-term dollar rates and can reduce the cost of hedging dollars. A hedged foreign investment then becomes more rewarding in yen. At the same time, a rise in long-term Japanese bond yields improves their appeal for new investments. The choice depends on the relative size of these changes, the slopes of the yield curves and the basis.

The relationship can be understood without predicting the next central-bank decision. If long-term Japanese yields rise while hedging conditions change little, domestic bonds gain ground. If US short-term rates fall more than US long-term yields, holding dollar bonds and renewing their hedges can become more attractive. Stress in the swap market can instead make protection more expensive even as the interest-rate differential improves. These scenarios follow the pricing relationships discussed by the BIS; whether they occur must be checked against contemporary market quotes. ⁹

Investors can also change how much of a portfolio they hedge. BIS research describes episodes in which investors adjusted currency derivatives while keeping their underlying assets, including the tensions of April 2025. Substantial dollar-market activity can therefore reflect a change in protection. Interpreting debt disposals and their currency effects requires examining both the asset portfolio and its hedges. ¹⁰

An insurer invests for the date of a promise

A life insurer’s choice also reflects the liabilities on its balance sheet. Its assets fund benefits payable over many years, according to policy terms and policyholders’ behaviour. The task is to have the required amounts in the relevant currencies when payments fall due. Asset-liability management organises that relationship and measures the mismatches that remain. Japan’s economic value-based solvency framework, applied from 31 March 2026, assesses assets and liabilities through this economic lens. ^{13, 14}

In its April report, the BOJ specifies that some foreign bonds classified as “unhedged” are earmarked for foreign-currency insurance policies. ¹⁵

Higher rates have two sides on the balance sheet. They improve returns on new purchases while reducing the market value of existing fixed-rate bonds. Other things equal, they also reduce the present value of fixed future benefits when those benefits are discounted at higher rates. The effect on the economic surplus depends on the amounts, sensitivities and options on both sides of the balance sheet. Looking only at bond valuation losses leaves a substantial part of the calculation outside the frame. ¹⁴

Method and scope

Qualitative diagram of present values and interest-rate sensitivity, other things equal. Both assets and fixed future benefits can lose present value as rates rise. The effect on the difference depends on their amounts, durations and embedded options. Arrows show relationships with no estimated magnitude.

Duration helps describe that sensitivity. Modified duration approximates a bond’s relative change in value for a small change in yield, other things equal. For an insurer, the sensitivities of assets and liabilities must be compared alongside their respective values. A very long bond can help bring the two profiles closer together. Once that adjustment is well advanced, buying progressively more long-dated assets serves the original matching need less directly. ^{14, 15}

The FSA’s August 2026 insurance monitoring report provides precisely this qualification. It observes that the accumulation of super-long JGBs ahead of the new framework had eased and that duration gaps had narrowed. It also describes replacements of low-yielding bonds. Regulation had already shaped portfolios before implementation; a fresh, uniform wave of purchases in October would require additional evidence. ¹³

Execution remains a management choice. Replacing an old bond with a higher-yielding one can crystallise an accounting loss, whereas holding the original to redemption spreads the adjustment over time. Natural maturities, incoming premiums and benefits payable provide several routes for reshaping a portfolio. Insurers differ in their payment schedules, liquidity resources and willingness to recognise an immediate loss. Company documents reveal these choices more clearly than a sector average. ^{11, 12}

Published strategies lead in several directions

In its June presentation, Nippon Life planned to reduce domestic bonds and increase currency-hedged foreign bonds in fiscal 2026, while actively managing older low-yielding securities. These were intentions announced at a particular date; their execution must be checked later. Their direction demonstrates that higher Japanese yields can coexist with a stated intention to expand selected foreign investments. ¹¹

Sumitomo Life’s September presentation describes another route: shifting part of its unhedged foreign bond allocation towards foreign equities. It also expects redemptions to weigh on its JGB portfolio for asset-liability management. A reduction in foreign debt can therefore accompany a change in the risk taken abroad, without an equivalent move into Japanese debt. ¹²

More recently, Reuters reported on 21 October that Meiji Yasuda planned purchases of twenty- and thirty-year JGBs while targeting a stable yen bond portfolio during its fiscal second half. Purchases and redemptions can offset one another. ¹⁹

These announcements illuminate possible decisions, distinct from the September transactions in the ministry’s panel. They also distinguish an intention to buy, gross purchases and a net increase in holdings. A manager buying after a redemption can be active without expanding the portfolio. When that manager moves between foreign asset classes, debt loses an investor while the capital remains exposed to other international risks.

Global prices turn on the next reinvestment

One transmission channel runs through reinvestment. A foreign issuer repays a bond; the Japanese investor compares available investments and may choose to participate less in the next issue. Another source of funding must take its place. If additional buyers require greater compensation, the market price adjusts downwards and the yield increases. That pressure can reach new issues and refinancing. The contractual coupon on an existing fixed-rate bond remains unchanged. Our analysis of the marginal Treasury buyer explains how new transactions set the funding price.

Method and scope

Conditional chain from reinvestment to marginal pricing. Proceeds from a redeemed bond may be reallocated; new borrowing then depends on available buyers and their required yield. Flows and delays have no numerical scale. Existing fixed-rate coupons remain contractual.

This competition is developing as the Bank of Japan reduces its government-bond purchases. A staff study published on 4 August describes rising JGB holdings among several investor groups and the time needed to adjust portfolios. It also notes that insurers’ major phase of super-long purchases in preparation for solvency regulation had largely run its course. Japan’s bond market must attract capital in an environment where each investor faces distinct constraints and alternatives. ¹⁶

Empirical research supports the existence of international transmission. An IMF working paper published in November 2025 examines surprises around 151 BOJ announcements between 2010 and 2023. It identifies effects on foreign sovereign yields; for asset-purchase shocks, the effects are stronger in markets with greater Japanese investor participation. The method concerns historical monetary-policy surprises and depends on its identification assumptions. Converting insurers’ September 2026 disposals into basis points of future yield increases would require a separate estimation. ¹⁷

The IMF’s April 2026 Global Financial Stability Report also discusses reallocation risk while emphasising the usually gradual adjustments of large investors. The effect depends on the market, its liquidity, the maturities affected and replacement demand. Local investors abroad, international funds or an improvement in hedged returns can absorb part of the adjustment. The risk concerns the price at which that absorption becomes possible and the speed at which it must happen. ¹⁸

The IMF’s October report adds context: sovereign yields also reflect expected monetary policy, term premiums compensating for interest-rate risk over time, and public borrowing needs. Its measures of volatility and bid-ask spreads point to more difficult JGB liquidity conditions, especially for long-dated bonds. Domestic reinvestment decisions also depend on those trading conditions. ²¹

The reasoning remains conditional. A persistent decline in Japanese demand for a particular segment, alongside substantial issuance and intermediaries reluctant to hold the risk, can raise the yield required to clear the market. A temporary flow quickly absorbed by other buyers leaves a different footprint. A monthly national aggregate cannot resolve these questions about maturity, counterparties and market depth.

Following the proceeds to their next investment

The weekly release of 22 October, covering transactions from 11 to 17 October, records JPY161.6 billion in net disposals of foreign medium- and long-term debt for the full reporting panel. It provides a recent observation with a broader scope than life insurers, without a corresponding investor breakdown. The two releases must retain their respective scopes. ⁴

US statistics provide another perspective with a different limitation. As our analysis of Asian purchases of US assets explains, holdings and transactions answer different questions. The TIC series distributed by FRED puts Japan’s Treasury holdings at approximately USD1,068 billion at the end of August 2026. It measures a market-value stock across investor categories, classified by legal residence. Custody through foreign intermediaries can obscure the final holder’s location. Changes combine transactions, valuation effects and other adjustments. Isolating any repatriation by life insurers would require sector-specific transactions. ²⁰

A more complete demonstration would connect investor-specific transactions to the countries, currencies and maturities of the securities, then to domestic purchases and currency hedges. It would also separate decisions taken before a rate increase from those responding to it and identify other forces moving foreign yields. The available evidence establishes sector-level net disinvestment and plausible allocation choices; much of the chain linking them remains to be documented.

The useful signal is a repeated reinvestment choice. Insurers that persistently favour JGBs after comparing yen income and balance-sheet needs can make it more expensive to find a creditor elsewhere. The documents also describe routes that could slow or redirect that movement: cheaper hedging, sustained international allocations and portfolios already adjusted to prudential requirements.

Japanese savings can contribute to higher foreign borrowing costs through the price investors require to keep providing finance. As of 22 October 2026, the evidence documents debt flows and diverse stated allocation plans. A generalised repatriation, followed by a rise in foreign yields attributable to that return, remains a hypothesis to test. The risk lies at this boundary between the observations and their interpretation.

The opening illustration is conceptual artwork. Research cut-off: 22 October 2026, Europe/Paris. Statistical periods, publication dates, company intentions and hypothetical examples are identified in the sources. The assumptions and currency-return formula are set out in this article. Contemporary hedged returns and any basis-point impact remain outside the study’s scope.

Sources and documents consulted

1. S01 · Ministry of Finance, Japan · Published: 2026-10-08 · Accessed: 22 October 2026MOF : International Transactions in Securities, September 2026
   Passages and scope

   Transactions from 1 to 30 September 2026; reports from designated major investors.

   p. 1, Portfolio Investment Assets, Total September 2026; p. 2, investor breakdown, Life insurance companies / Long-term debt securities and Short-term debt securities.

   Unit: JPY 100 million. Life insurers’ published net is −2,886, or −JPY 288.6 billion. Dispositions include redemptions. Rounded gross amounts may not reproduce the published net exactly. This table does not trace reinvestment in Japan.

   Official statistical producer; the flows describe its reporting coverage and do not establish market causality.

   Additional references: Ministry of Finance, Japan 1

2. S02 · Ministry of Finance, Japan · Published: 2026-10-08 · Accessed: 22 October 2026MOF : Monthly foreign long-term debt transactions by investor type
   Passages and scope

   History from 2005 to September 2026; extraction covers 2025–September 2026 and annual totals for 2021–2025.

   CSV: header rows 1–11; months in 2025–September 2026 rows 252–272; life insurers columns 43–45; trust banks’ trust accounts columns 37–39. 2026 cumulative: rows 264–272. Twelve months: rows 261–272. Annual totals for 2021–2025: rows 292–296.

   One-based row and column numbers. CP932 encoding; unit JPY 100 million. l0g cumulative figures sum published monthly nets. History is the 8 October vintage, not a reconstruction of every first release. Do not add aggregates to their own subcategories.

   Official statistical producer; the flows describe its reporting coverage and do not establish market causality.

   Additional references: Ministry of Finance, Japan 1

3. S03 · Ministry of Finance, Japan · Published: date unspecified · Accessed: 22 October 2026MOF : International Transactions in Securities: methodology and coverage
   Passages and scope

   Online definitions accessed on 22 October 2026.

   FAQ Q1–Q2: reporting coverage and difference from BOP; Q5: redemptions included in dispositions; Q6: execution and redemption dates.

   ITS does not cover all portfolio investment. Unlike BOP, it does not provide country or currency breakdowns. Net disposition establishes neither yen conversion, JGB purchases nor investment motivation.

   Official statistical producer; the flows describe its reporting coverage and do not establish market causality.

4. S04 · Ministry of Finance, Japan · Published: 2026-10-22 · Accessed: 22 October 2026MOF : International Transactions in Securities, release of 22 October 2026
   Passages and scope

   Latest week: 11–17 October 2026; table also includes the four previous weeks.

   p. 1, Portfolio Investment Assets, Long-term debt securities, October 11–October 17 row; note 1 for revision asterisks.

   Latest net: −1,616 units of JPY 100 million, or −JPY 161.6 billion. Five consecutive negative weeks for the full reporting panel. No weekly investor breakdown attributes this figure to life insurers. Figures marked * are revised.

   Official statistical producer; the flows describe its reporting coverage and do not establish market causality.

   Additional references: Ministry of Finance, Japan 1

5. S05 · Bank of Japan · Published: date unspecified · Accessed: 22 October 2026BOJ : Investor classification in balance-of-payments statistics
   Passages and scope

   Definitions accessed on 22 October 2026; FAQ linked by the MOF for investor classification.

   Q14, Table 2: banking accounts and trust accounts; Q16: BOP revisions and stock valuation.

   Trust banks’ trust accounts are an account category without an ultimate-beneficiary breakdown here. They cannot simply be relabelled “pension funds”. BOP revision rules do not automatically apply to ITS.

   Central bank and statistical compiler; methodological document without an additional ultimate-beneficiary breakdown for each flow.

6. S06 · Ministry of Finance, Japan · Published: 2026-10-22 · Accessed: 22 October 2026Japan Ministry of Finance : constant-maturity JGB yield curve
   Passages and scope

   Observations 1–21 October 2026; latest point 21 October

   Row 2026/10/21; columns 1Y to 40Y.

   Constant-maturity yields with semiannual compounding. These are neither coupons nor intraday quotes; no FX-hedged yield is calculated from this curve alone.

   Public issuer; publishes a curve calculated from JSDA reference prices.

   Additional references: Ministry of Finance, Japan 1

7. S07 · Ministry of Finance, Japan · Published: date unspecified · Accessed: 22 October 2026Japan Ministry of Finance : yield methodology Q&A
   Passages and scope

   Methodology accessed on 22 October 2026

   Questions on rate type, publication time, JSDA prices and constant maturities.

   The Q&A specifies a 3 pm close and release at 9:30 am on the next business day. Rates use semiannual compounding.

   Statistical producer explaining its own methodology.

8. S08 · Bank of Japan · Published: 2026-09-18 · Accessed: 22 October 2026Bank of Japan : monetary policy decision of 18 September 2026
   Passages and scope

   Decision on 18 September, effective from 24 September 2026

   p. 1, paragraphs 1–2 and footnotes 1–2.

   Uncollateralised overnight call-rate target around 1.25%. This is a policy rate, not a three-month FX hedge quote.

   Monetary authority announcing its own decision.

9. S09 · Bank for International Settlements · Published: 2025-06-29 · Accessed: 22 October 2026BIS Annual Economic Report 2025 : Chapter II
   Passages and scope

   Analysis published in June 2025; series mainly through end-2024

   Chapter II, pp. 52–56; Box A pp. 53–54; transmission pp. 62–65.

   Hedge cost: interest-rate differential at the derivative tenor plus basis. Every value in the teaching figure is fictional, not a quote from this report.

   Central-bank institution; economic and financial-stability analysis.

   Additional references: Bank for International Settlements 1

10. S10 · Bank for International Settlements · Published: 2025-12-08 · Accessed: 22 October 2026BIS : Global FX markets when hedging takes centre stage
    Passages and scope

    April 2025 FX survey; 2022–2025 comparison and hedging episodes

    BIS Quarterly Review December 2025, pp. 23–24, 28–30 and 35.

    Hedge ratios can adjust without selling underlying assets. The Japanese hedge-ratio change from about 60% to 40% concerns 2022–2024, not October 2026.

    BIS research; authors’ views need not represent member central banks.

    Additional references: Bank for International Settlements 1

11. S11 · Nippon Life Insurance Company · Published: 2026-06 · Accessed: 22 October 2026Nippon Life : FY2025 Financial Results and Management Strategy Briefing, June 2026
    Passages and scope

    Results for April 2025–March 2026 and investment plan for April 2026–March 2027.

    p. 49: interest-rate risk and asset-liability matching; p. 54: FY2026 investment plan. Physical PDF pages match printed pages.

    The insurer’s June plan calls for lower domestic bonds, higher hedged foreign bonds and stable unhedged foreign bonds. These are annual intentions, not an October transaction record. Replacing low-yield bonds does not measure repatriation.

    Corporate communication about strategy and results; intentions and assessments attributed to Nippon Life.

12. S12 · Sumitomo Life Insurance Company · Published: 2026-09-09 · Accessed: 22 October 2026Sumitomo Life : Investor Presentation, September 2026
    Passages and scope

    Results as of March 2026 and investment plan for April 2026–March 2027.

    pp. 17–21, especially p. 21: FY2026 plan; p. 20: replacements and ALM. p. 8: regulatory ESR; p. 29: internal ESR using a different methodology. Physical PDF pages match printed pages.

    Sumitomo plans, among other changes, to shift part of its unhedged foreign bonds into foreign equities. In its ALM portfolio, JGB holdings may decline despite purchases for new insurance contracts because of maturities. Foreign-bond sales, lower total foreign investment and repatriation are different concepts.

    Corporate communication about allocation and risks; intentions and assessments attributed to Sumitomo Life.

    Additional references: Sumitomo Life Insurance Company 1

13. S13 · Financial Services Agency of Japan · Published: 2026-08-06 · Accessed: 22 October 2026FSA : Insurance Monitoring Report 2026
    Passages and scope

    Supervision report published in August 2026; the ESR field-test results on p. 20 refer to 31 March 2025.

    Printed pp. 19–21, physical PDF pages 20–22; printed p. 20 / PDF p. 21 on super-long JGBs, duration gaps and bond replacements.

    The FSA says preparatory super-long JGB accumulation for ESR has subsided and asset-liability duration gaps have narrowed. Higher yields also increase accounting valuation losses on bonds and raise liquidity issues. The reported 215% average ESR is a March 2025 field test, not an October 2026 result.

    Prudential authority; institutional interpretation of sector supervision.

14. S14 · Financial Services Agency of Japan · Published: 2026-07 · Accessed: 22 October 2026FSA : Overview of the Economic Value-Based Solvency Regulation
    Passages and scope

    Regime applicable from 31 March 2026; explanatory document dated July 2026.

    Printed pp. 1–4 / PDF pp. 2–5: objectives, economic balance sheet and risks; printed p. 9 / PDF p. 10: timetable and application date.

    Exact application date: 31 March 2026. The regime revalues assets and liabilities and measures eligible capital against required capital, including interest-rate, currency, spread and lapse risks. It is not a general requirement to buy JGBs or repatriate assets. The solvency effect of higher yields depends, among other things, on the balance sheet and its matching.

    Authority responsible for the prudential regime; primary source for its operation and timetable.

    Additional references: Financial Services Agency of Japan 1, Financial Services Agency of Japan 2

15. S15 · Bank of Japan · Published: 2026-04-21 · Accessed: 22 October 2026Bank of Japan : Financial System Report, April 2026
    Passages and scope

    Primarily data available at end-March 2026; cited insurer valuations at December 2025; currency-hedging chart at September 2025.

    Chapter III.B, printed pp. 23–24 / PDF pp. 28–29: insurance; chart notes give their scope and dates.

    Latest published Financial System Report available when researched on 22 October 2026. The BOJ describes insurers as cautious about super-long JGBs and unhedged foreign bonds; ESR-related duration preparations have largely run their course. Some unhedged foreign assets match foreign-currency insurance contracts. Do not apply the report’s banking stress-test findings to insurers.

    Central-bank source assessing financial stability and its risks.

    Additional references: Bank of Japan 1, Bank of Japan 2

16. S16 · Bank of Japan · Published: 2026-08-04 · Accessed: 22 October 2026Impact of the Bank of Japan’s Reductions in JGB Purchases on the JGB Markets
    Passages and scope

    Holdings from June 2024 to March 2026; analysis published in August 2026.

    BOJ Review 2026-E-10, pp. 5–6, Chart 8; pp. 3–4 on limitations of term-premium estimates.

    Investor portfolios adjust as BOJ purchases decline. The paper distinguishes investor groups and says insurers’ earlier regulatory purchases of super-long bonds have largely run their course. Holdings do not directly measure international transactions.

    Central-bank primary source; assessment of its own policy, with estimates and interpretations attributed to the authors.

    Additional references: Bank of Japan 1

    BOJ, 30/09/2026

17. S17 · International Monetary Fund · Published: 2025-11-07 · Accessed: 22 October 2026IMF : Macroeconomic Effects and Spillovers from Bank of Japan Unconventional Monetary Policy
    Passages and scope

    151 BOJ announcements in 2010–2023; panel of 38 countries

    WP 25/227, pp. 18–24 (PDF pp. 21–27); identification p. 5.

    High-frequency policy surprises; stronger LSAP spillovers where Japanese participation is higher. Historical coefficients are neither a sales-to-yields elasticity nor a 2026 forecast.

    Authors’ research published for discussion; not necessarily the IMF’s institutional position.

    Additional references: International Monetary Fund 1

18. S18 · International Monetary Fund · Published: 2026-04 · Accessed: 22 October 2026IMF : April 2026 GFSR, Japanese yields and global asset allocation
    Passages and scope

    April 2026 analysis; international holdings at end-2024 in Figure 1.1.1

    Chapter 1, Box 1.1, pp. 34–35.

    Conditional risk of higher issuance costs abroad; mandates generally adjust gradually. Foreign-holdings comparison includes both public and private debt, not sovereign securities alone.

    Multilateral institution assessing global financial stability.

19. S19 · Reuters · Published: 2026-10-21 · Accessed: 22 October 2026Reuters / Yahoo Finance Japan : Meiji Yasuda’s FY2026 second-half investment plan
    Passages and scope

    Intentions for October 2026–March 2027; article published the day before the cutoff.

    Headline and paragraphs on 20/30-year JGB purchases and expected stable yen-bond holdings in the second half.

    Corporate statements reported by Reuters, not realized transaction data. Use in both articles is limited to continuing purchases and the effect of redemptions on holdings.

    News agency reporting an insurer briefing; strategy attributed to Meiji Yasuda.

20. S20 · Federal Reserve Bank of St. Louis / U.S. Treasury TIC · Published: 2026-10-16 · Accessed: 22 October 2026Foreign Portfolio Holdings of U.S. Long-Term and Short-Term Treasury Securities: Japan
    Passages and scope

    August 2026; monthly series in millions of dollars, market value, not seasonally adjusted.

    Series FORTREASPOS42609: April–August 2026 observations and methodological notes.

    The series measures market-value holdings by legal-residence country across investor types. Changes combine transactions, valuation and other adjustments; they are not life-insurer flows.

    Institutional dissemination of Treasury TIC data and Federal Reserve estimates; series subject to revision.

    Additional references: Federal Reserve Bank of St. Louis / U.S. Treasury TIC 1, Federal Reserve Bank of St. Louis / U.S. Treasury TIC 2

21. S21 · International Monetary Fund · Published: October 2026 · Accessed: 22 October 2026IMF: global financial stability, October 2026
    Passages and scope

    Chapter 1, pp. 5–6, Figure 1.5. Sovereign-yield context and JGB liquidity indicators.