Who Is Buying Japan’s Government Debt?

Japan’s government debt has a familiar set of owners: the central bank, banks, insurers, pensions and households. But a familiar ownership base does not guarantee a familiar price. As the Bank of Japan reduces bond purchases, the willingness of investors to buy the next bond—and the yield they require—matters increasingly alongside the immense stock already held.

Foreign demand is part of that story. A September 10 Reuters Breakingviews commentary reported ¥17.9 trillion of foreign purchases of Japanese fixed-income securities over six months. That broad description should not be treated as a verified total for government bonds alone.^[9]

This report uses information available through September 10, 2026, for the September 12 Japan edition. Its ownership snapshot is older by design: the BOJ’s first-quarter preliminary Flow of Funds release, published June 25, describes end-March holdings. It does not show a real-time September allocation.^[2]

Start with the owners—and the denominator

Source: BOJ, chart 6-2. The first column combines Treasury discount bills with JGBs and Fiscal Investment and Loan Program bonds. The second excludes bills. Public pensions are separate from insurance and pension funds. Figures include issuance by the Fiscal Loan Fund; rounding may affect totals.^[1]

The BOJ is the largest holder in both measures, and domestic sectors hold most of the debt. Foreign ownership is more prominent when bills are included. A claim about foreign participation therefore needs to specify which securities it covers.

Market-value measurement introduces another distinction. A portfolio’s value can fall because bond prices decline, even if the investor sells nothing. Redemptions and changes in total outstanding securities can also alter ownership shares. A shrinking percentage is not, on its own, evidence of an exit.

Three different meanings of “buyer”

A holder is an investor with bonds on its balance sheet. A net buyer acquires more than it disposes of over a specified period. An active market participant may trade repeatedly and help establish prices while retaining relatively little inventory. The largest holder need not be the most influential participant in a particular trading session.

Buying at issuance also differs from buying an existing security. A secondary-market purchase normally pays the seller, rather than sending new money directly to the government. Yet secondary yields influence how dealers value the next issue and the terms investors will accept. Ownership, transactions and financing conditions are connected without being interchangeable.

Japan.co.jp’s analysis is that this distinction avoids two misleading conclusions: that foreign buying certifies fiscal safety, or that a predominantly domestic ownership base makes foreign activity irrelevant. The important question is how different sources of demand respond when price, maturity and currency conditions change.

The BOJ is buying less, not disappearing

The BOJ’s June 16 plan set monthly long-term JGB purchases at around ¥2.5 trillion in July–September 2026, ¥2.3 trillion in October–December and ¥2.1 trillion in January–March 2027. From April 2027, it plans around ¥2 trillion a month. It retained flexibility to respond to a sharp rise in long-term yields.^[3]

These are planned purchases, not net increases in holdings. If maturing bonds exceed new purchases, the portfolio can shrink while the central bank remains an active buyer. Nor is a reduction in monthly purchases equivalent to selling the same amount of bonds into the market.

For private investors, the adjustment poses a price question: what return makes additional exposure worthwhile? There is no single answer across the maturity spectrum. A buyer seeking a short-term place to park cash does not automatically provide demand for a thirty-year bond. Replacing demand requires attention to the risk being absorbed, not merely the cash being spent.

Domestic institutions do not share one investment mandate

Banks balance securities against deposits, lending opportunities and liquidity needs. Higher yields can improve the prospective income from new purchases while reducing the market value of older holdings. Their ability to buy depends on capital and risk capacity as well as the attractiveness of the next coupon.

Life insurers have another consideration: assets that can help meet payments far into the future. Long-duration bonds may suit long-duration liabilities. But an insurer that has already acquired the exposure it needs does not necessarily buy more simply because yields rise. The phrase “domestic institutional demand” conceals different investment problems.

Pensions add portfolio constraints. The Government Pension Investment Fund’s policy portfolio for the five-year period beginning in fiscal 2025 assigns 25% each to domestic bonds, foreign bonds, domestic equities and foreign equities. The domestic-bond target is not a government-bond-only target, and it is not a quarterly purchase instruction.^[4]

Consequently, higher Japanese yields do not mechanically trigger an equal-sized sale of overseas assets. Portfolio targets, movements in other asset prices, currency exposure and implementation choices intervene. Rebalancing to maintain an allocation can resemble a directional investment bet in transaction statistics, even though its purpose is different.

The foreign investor’s return is not the yen yield

Foreign investors also encompass different strategies: long-horizon pension and sovereign funds, bond funds, bank treasury operations and investors trading relative prices. An overseas statistical category does not reveal the identity, nationality, holding period or hedge of every ultimate investor.

An investor measuring performance in dollars or euros must consider currency as well as the bond. A hedge changes the terms on which yen proceeds return to the investor’s home currency. BIS research explains that cross-currency pricing reflects hedging demand and the balance-sheet capacity of financial intermediaries, as well as interest-rate relationships.^[8]

A simple hypothetical example shows why this matters. Suppose a yen asset earns 3% over a year, but the exchange rate moves from ¥150 to ¥165 per dollar. Before fees and taxes, the unhedged dollar return is approximately minus 6.36%. This is Japan.co.jp’s calculation, not a current exchange-rate forecast or an observed investment result.

Hedging can change that outcome, and its effect is not always simply a deduction from the yen yield. The forward exchange rate matters. But a currency hedge does not eliminate a bond-price loss. A strategy that repeatedly renews short-dated hedges also faces uncertainty about future renewal terms. Comparisons with US or European debt must therefore use the same currency, horizon and risk assumptions.

A proposed allocation is not an executed order

Reuters reported September 4 that Norges Bank Investment Management, which manages Norway’s sovereign wealth fund, was proposing changes to its bond strategy that would increase Japanese government-bond investments. The report concerned a proposed strategy change, not a disclosed completed purchase of that amount.^[10]

Benchmark discussions can matter because large portfolios may eventually adjust to them. But approval, timing and implementation remain distinct steps. A percentage allocation can also change in value as markets move. Treating every proposal as cash already committed would overstate the certainty of future demand.

The policy history behind today’s market

The BOJ’s role developed through successive policy frameworks: quantitative easing in 2001, quantitative and qualitative easing in 2013, then negative interest rates and yield-curve control in 2016. Under yield-curve control, the central bank directly influenced the long end of the interest-rate structure as well as short rates.^[6]

In March 2024, the BOJ ended that framework and returned short-term rates to the center of monetary policy. It nevertheless continued long-term JGB purchases. The shift in policy did not instantaneously transfer the accumulated bond portfolio to private investors.^[7]

Today’s market is negotiating the consequences over time. Investors must reassess inflation risk and the extent of policy support against their own liabilities and return objectives. The Ministry of Finance describes smooth, reliable issuance and control of medium- to long-term funding costs as core debt-management goals, with diversification of holders among its tools.^[5]

A broader investor base can help because buyers need not all react in the same way. Some want a particular maturity, others a liquid instrument or an exposure required by a benchmark. But diversification does not mean demand becomes insensitive to price. It changes the sources of demand, not the basic need to offer acceptable terms.

A buyer at a higher yield is still a higher funding cost

A bond-market selloff and new buying are not contradictory. Each completed trade has both sides. Lower prices can attract investors who previously found the yield inadequate. The existence of buyers therefore does not establish that confidence or financing conditions are unchanged.

For the government, the relevant issue is the volume it can issue across maturities and the terms available. A rise in market yields does not immediately reset the coupons on all outstanding fixed-rate debt. It reaches the budget through new issuance and refinancing over time. Multiplying today’s yield increase by the entire debt stock would misrepresent that timing.

Domestic ownership also does not remove inflation risk, opportunity costs or fiscal trade-offs. A government borrowing in its own currency and an investor concerned about the purchasing power of that currency can both be looking at the same bond rationally. A deep domestic base is a feature of the market, not a promise that future funding will remain cheap.

What the next data should establish

Read ownership statistics with their date, valuation method and security coverage. Read transaction figures with their period and net-versus-gross definition. Read auction outcomes with the maturity and amount offered. The location of a financial intermediary need not identify the ultimate investor taking the risk.

The useful test is not whether “foreigners” or “Japan” wins a contest to finance the state. It is whether demand remains broad enough across maturities, and at what cost, as central-bank support changes. Japan’s government debt still rests on a substantial domestic base. The next bond must nevertheless find an investor willing to accept its particular price and risk.