Limited room for Japanese savings to support JGBs, with the long end more likely to benefit from signaling effects

Overview

The report discusses recent comments from Japanese policymakers about increasing domestic purchases of Japanese assets, focusing on two channels that could support Japanese government bonds: first, GPIF raising allocations to domestic assets, especially JGBs; second, allowing JGBs into NISA, the tax-exempt savings vehicle for households. Goldman Sachs believes both measures could increase JGB demand, but foreign ownership of Japanese government bonds is already very low within the G10, leaving limited direct room to further redirect domestic savings into JGBs.

Core views

The core judgment is that the direct buying impact of GPIF rebalancing and NISA expansion on JGBs is limited, and the market impact would come more from policy signaling, follow-on buying by other domestic investors, and positive feedback loops. GPIF currently has about USD75bn of room to add JGBs, and if the domestic bond allocation rises from 27% to 31%, this could correspond to about a 3-6bp move in 30-year swap spreads; however, the pace of rebalancing may be slow, since the domestic bond allocation has historically moved by only 1-2 percentage points per year. On NISA, tax incentives may bring household demand, but the low bond holding ratio is more likely the result of persistently low yields. The report believes 20-year and 30-year JGBs have more room to strengthen, while the cheapness in 10-year bonds and the belly of the curve is mainly driven by macro policy and cannot be resolved by supply-demand measures alone.

Analysis framework

The report uses GPIF asset allocation, bond holding structure, maturity distribution, rebalancing flows, free-float supply, and swap spread sensitivity frameworks, combined with NISA account scale and household asset allocation, to assess the impact of potential policy changes on the JGB curve and cross-market spreads.

Methodology notes

• Rates supply-demand frameworkFree-float supply and 30-year swap spread sensitivity

  Measure the impact of supply shocks on long-end swap spreads through changes in the share of debt held by the private sector.

  The report cites Goldman Sachs' swap spread framework, under which a 1 percentage point change in the free-float share in UST, Bund, and Gilt markets typically corresponds to about 1bp in 30-year swap spreads; if flows are more concentrated in 30-year bonds or induce other buyers, the impact can rise to 2-4bp.

• Institutional asset allocation analysisGPIF strategic allocation and rebalancing flows

  Estimate potential JGB demand through GPIF's allocation across the four major asset classes, permitted deviation ranges, and duration management behavior.

  GPIF allocates roughly 25% each to domestic bonds, foreign bonds, domestic equities, and foreign equities, with domestic bonds allowed to fluctuate by about plus or minus 6 percentage points; because equities and foreign assets have performed strongly, the yen has weakened, and JGB yields have risen, GPIF needs to keep rebalancing into JGBs to maintain target allocations.

• Household flow analysisNISA tax-incentive channel

  Assess whether tax-exempt investment accounts can redirect household cash savings into JGBs.

  There are more than 20 million NISA accounts, with transaction value exceeding JPY71tn by the end of 2025; the growth quota currently allows investment in JGB funds but not direct purchases of JGBs. The report believes allowing direct investment could increase demand, but households' low JGB allocation is mainly related to historically low yields.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• JGBs

  Core subject of the research, potentially supported by GPIF rebalancing and NISA policy.

  Strengths

  Strong domestic investor base, low foreign ownership in Japan, and policy signaling can drive local demand.

  Weaknesses

  There is limited room to further direct domestic savings, and yields and volatility are influenced more by macro policy.

  Comparison

  Compared with USTs, Bunds, and Gilts, GPIF has a higher holding share in the JGB market, giving it a more direct impact on the domestic curve.

  Risks

  If fiscal and monetary policy do not adjust, supply-demand measures will struggle to deliver lasting stability.

• 20y and 30y JGBs

  Maturity sectors the report sees as more likely to outperform.

  Strengths

  Supply-related cheapness is more pronounced at the ultra-long end, and GPIF has the ability to direct funds to the long end to maintain weighted average duration.

  Weaknesses

  The pace of rebalancing may be slow, and actual buying size may not be enough to change the long-term trend.

  Comparison

  Compared with 10-year bonds and the belly of the curve, 20-year and 30-year sectors are more likely to be supported by supply-demand and signaling effects.

  Risks

  If macro policy continues to cheapen the curve, support for the long end may be offset.

• JPY swap spreads

  An important market variable for measuring long-end JGB supply-demand shocks and relative value.

  Strengths

  The sensitivity of 30-year swap spreads to potential GPIF additions can be estimated.

  Weaknesses

  Recent volatility has already been large, and policy expectations may reverse quickly.

  Comparison

  The report compares the potential 3-6bp impact with the roughly 8bp round-trip move in 30-year swap spreads that week.

  Risks

  If the market extrapolates GPIF rebalancing too aggressively to the entire pension industry, short-term mispricing may result.

• NISA household savings

  A potential channel for incremental JGB demand.

  Strengths

  Broad account coverage and large transaction volumes; if direct investment in JGBs is allowed, household participation could rise.

  Weaknesses

  The underweight in JGBs is more likely due to low yields rather than tax restrictions.

  Comparison

  The Italian experience shows that a rise in retail bond holdings usually comes with higher yields, not lower yields.

  Risks

  If household buying is a response to higher yields, it may not push yields sustainably lower.

• USTs, Bunds and Gilts

  Used as comparison objects for GPIF foreign bond holdings and G10 supply-demand sensitivity.

  Strengths

  Useful for calibrating the historical impact of free-float supply changes on swap spreads.

  Weaknesses

  With GPIF's holding share below about 1%, the direct cross-market impact of a small shift from foreign bonds into JGBs is limited.

  Comparison

  Compared with JGBs, overseas markets are less directly affected by GPIF marginal rebalancing.

  Risks

  Only if Japanese domestic investors adopt a broader policy of reducing foreign bond holdings could cross-market effects become larger.

Key data

• GPIF asset sizeabout USD2.8tnAs of end-March 2026.
• GPIF domestic bond target allocationabout 25%, with an allowable range of about +/-6 percentage pointsThe current strategy is expected to be reviewed in 2030.
• GPIF room to add JGBsabout USD75bnEstimated under the current strategy framework.
• GPIF share of JGB market holdingsabout 5%At the same time, its holding share in G10 markets such as USTs, Bunds, and Gilts is below about 1%.
• Share of JGBs within GPIF domestic bondsabout 80-90%Estimated based on GPIF annual single-security holding data.
• Impact of raising domestic bond allocation from 27% to 31%about 3-6bp in 30-year swap spreadsGoldman Sachs estimate; comparable with the roughly 8bp round-trip move in 30-year swap spreads that week.
• Foreign ownership ratio of Japanese government bondsabout 14%A relatively low level among G10 government bond markets.
• NISA accounts and transaction valuemore than 20 million accounts, with transaction value exceeding JPY71tnAs of end-2025.
• Share of NISA growth quota in transaction valueabout 75%This quota can invest in stocks, ETFs, REITs, active funds, and JGB funds, but cannot invest directly in JGBs.
• Japanese households' direct JGB holding ratioabout 4%As a share of portfolio assets excluding NISA accounts.

Impact & implications

From an investment perspective, the policy discussion can support JGB sentiment in the short term, especially at the ultra-long end and in 20-year and 30-year swap spreads, but it is not a fundamental solution to JGB volatility. If policy remains only at the level of supply-demand and savings redirection, it may do more to reduce marginal volatility or improve market psychology; only if fiscal tightening, the monetary policy path, or inflation expectations also change is a sustained decline in JGB yields and volatility more likely.

Risks

• GPIF's actual rebalancing pace may be slower than the market expects, causing short-term JGB gains to reverse.
• If Japan's fiscal and monetary policy do not shift in tandem, JGB volatility may struggle to decline sustainably.
• Including JGBs in NISA may generate only limited incremental demand, because the root cause of households' underweight in bonds may be historically low yields.
• The market may extrapolate GPIF's potential rebalancing too aggressively to the entire pension system, creating excessive expectation-driven reactions.
• If the inflation path does not move lower, supply-demand improvement may be insufficient to suppress yields.

What to watch

• Follow-up policy comments from Japan's Ministry of Finance and Prime Minister on encouraging purchases of domestic assets.
• Whether GPIF adjusts its domestic bond allocation before 2030 or releases early signals of a strategic review.
• Whether NISA rules will allow direct holding of JGBs, and the specific quotas and product scope.
• Performance of 20-year and 30-year JGBs relative to 10-year bonds and the belly of the curve.
• Changes in 20-year and 30-year JPY swap spreads.
• Whether Japan's fiscal policy, monetary policy, and inflation trajectory shift enough to reduce JGB volatility.