Back to Columns

Japan Yield Curve Bear-Flattens as Rates Rise

Japan’s August 2026 JGB yields rose most at the short end, driving bear flattening. The analysis examines real rates, r-g dynamics, and debt sustainability.

IRTracker
12 min read
JGB YieldsYield CurveFiscal PolicyJapan Economy

In Japan’s JGB market in August 2026, the main driver of rising yields shifted from the long and super-long ends to the short end, and the yield curve turned to bear flattening. According to data released by the MOF (Ministry of Finance), month-end yields rose most sharply for the 1-year tenor, up 24.7bp month on month to 1.502%, and the 2-year tenor, up 23.6bp to 1.743%, while the increase in the 30-year yield was limited to 11.0bp. The 10-year yield rose to 2.943%, and the real 10-year rate recovered to the 1% range at 1.043% (a provisional estimate using July’s 1.9% year-on-year CPI). Meanwhile, 25- to 40-year yields remained clustered at 4.09–4.10%, clearly indicating flattening at the super-long end.

1. Yield Curve Dynamics — Yields Rose Across All Tenors, with the Largest Increases at the Short End

Yields rose across all tenors in August, but the change took the form of bear flattening, with the increases larger at shorter maturities.

TenorMonth-end (%)Change from prior month (bp)Change from 6M earlier (bp)
1Y1.502+24.7+47.9
2Y1.743+23.6+49.3
5Y2.233+18.9+63.9
10Y2.943+14.2+81.1
20Y3.815+12.5+83.6
30Y4.092+11.0+75.0
40Y4.094+12.7+65.4

This table shows that the main driver of rising yields has shifted depending on the time horizon. Compared with six months earlier, the 10- to 20-year sector recorded the largest increases, exceeding +80bp, indicating that the rise was previously led by the medium- to long-term sector. By contrast, over the latest month, the 1- to 2-year sector recorded the largest increases, at around +24bp, showing that the center of the rise had shifted to the short end. The curve as a whole shifted upward, but its shape changed from steepening over the six-month horizon to flattening during the latest month. These should be distinguished as separate phenomena based on different reference periods.

The curve maintains a monotonic upward slope from 1 year through 25 years. However, compared with 25-year yields at 4.102%, 30-year yields at 4.092% and 40-year yields at 4.094% show a very slight inversion of the slope after the 25-year peak. There are two possible interpretations of super-long yields becoming saturated near 4.1%. One is that demand from domestic investors with long liability durations is creating an upper limit; the other is that the market has begun to value the long-run equilibrium nominal rate plus a risk premium at around 4%. Because this article’s data do not include trading trends by holder type, it is impossible to determine which interpretation is dominant. What can be confirmed is that the 30-year tenor recorded the smallest increase in the month, while the 40-year increase of 12.7bp was also below the 10-year increase of 14.2bp.

2. Term Spread Analysis — A Divergence between Economic Signals and Fiscal Confidence

Both the 10Y-2Y and 30Y-10Y spreads narrowed, meaning that the two spreads moved in the same direction.

Month10Y-2Y30Y-10Y40Y-10Y
2026-010.9961.3301.434
2026-061.3081.1831.102
2026-071.2941.1811.166
2026-081.2001.1491.151

The 10Y-2Y spread peaked at 1.308% in June and narrowed for the second consecutive month, reaching 1.200% in August. As a business-cycle signal, this is consistent with flattening during a tightening phase, in which short-term rates rise more than long-term rates. An opposing interpretation—that it is a precursor of an economic slowdown—is also possible, but the coincident index discussed below rose in June, and there is currently no evidence indicating a recession.

The 30Y-10Y spread narrowed by 18.1bp, from 1.330% in January to 1.149% in August. The fact that the spread between the super-long and long ends narrowed while yields rose across all tenors is consistent with the absence of an increase in the term premium specific to the super-long end. However, this article does not contain data that identify the factors driving the change in the spread, so the presence or absence of fiscal factors cannot be asserted. What can be noted is that the 30-year yield’s absolute level has become established in the 4% range, creating a heavy funding cost for new issuance and refinancing, and that the 40Y-10Y spread at 1.151% is almost identical to the 30Y-10Y spread at 1.149%, indicating that little additional term premium is being paid beyond 30 years.

3. Real Interest Rate Analysis — Positive Rates Become Established and Real Funding Costs Rise

The real 10-year rate was 1.043% (provisional), remaining consistently in positive territory since the beginning of 2026.

MonthNominal 10Y (%)CPI YoY (%)Real 10Y (%)
2026-052.6571.51.157
2026-062.6901.61.090
2026-072.8011.90.901
2026-082.9431.9※1.043※

※Because August CPI had not yet been released, the July year-on-year figure of 1.9% was used as a substitute, making this a provisional value. The August increase merely reflects the rise in nominal rates assuming unchanged CPI; if August CPI accelerates, the figure will be revised downward. Therefore, the trend in real rates cannot be determined solely from the current-month figure.

The important point is the level. Since the beginning of 2026, nominal yields have remained around 1 percentage point above inflation, meaning that this is not a period in which negative real rates erode the real value of debt. For the government, the real cost of new funding is positive, and stabilizing the debt ratio will require real growth to exceed the real interest rate or an improvement in the primary balance (PB). For households, this is an environment in which investing in long-term bonds can generate real returns above inflation. According to data from the Statistics Bureau of the MIC (Ministry of Internal Affairs and Communications), core-core inflation slowed from 2.5% in January to 1.7% in June before reaccelerating to 1.9% in July, meaning that the inflation denominator for real interest rates remains unstable.

4. Connection with BOJ Policy — What a 1.5% One-Year Yield Signals about Policy Expectations

Because the BOJ (Bank of Japan) call rate could not be obtained for this analysis, no direct assessment of the policy rate level is made. As an alternative indicator, the 1-year yield has risen to 1.502% (up 24.7bp month on month and 47.9bp from six months earlier). The fact that the maturity most sensitive to the policy rate has risen by approximately 48bp over six months is consistent with the market continuing to price in additional monetary tightening. The spread over the 2-year yield is 24.1bp, suggesting that pricing for rate hikes within one year is substantial.

The quantitative scale of QT—the reduction of JGB holdings—is outside this article’s data coverage. However, the fact that 10- to 20-year yields recorded the largest increases over the six-month horizon is not inconsistent with a channel through which reduced central-bank purchases transfer duration risk in the medium- to long-term sector to private investors. Because data on the holding structure and purchase volumes were not obtained, the contribution of this channel cannot be evaluated quantitatively. A period such as the current month, in which yields rise mainly at the short end and the increase at the super-long end is smallest, indicates that the focus of price formation has shifted from the term premium to the policy-rate path.

5. Consistency with the Real Economy — Rising Rates Are Compatible with Economic Expansion

The Cabinet Office’s business-cycle indicators support the view that rising interest rates are occurring alongside economic expansion.

MonthLeading indexCoincident indexLagging index
September 2025108.0115.1112.9
April 2026116.1118.1111.7
May 2026116.5117.9111.3
June 2026116.5118.5111.8

The coincident index rose to 118.5 in June, its highest level during the period covered by the data. The leading index was flat at 116.5 in May and June, indicating that upward momentum has temporarily reached a plateau. The lagging index declined from 113.2 in April 2025 to 111.8, showing that the lagging series has not followed the economic expansion. The data run only through June, creating a two-month lag relative to the interest-rate developments in August; this limitation should be noted.

The implication is clear. As long as nominal rates rise at the same time as the economy expands, the denominator in the debt-dynamics equation—nominal GDP growth—will also be pushed higher. The worst scenario for public finances is one in which rising rates occur alongside economic deterioration, and there are currently no confirmed signs of that. The key indicator to watch is whether the leading index’s stagnation turns into a decline.

6. Corporate Sentiment — Current Improvement Coexists with Forward-Looking Caution

The BOJ Tankan for Q2 2026 shows simultaneous improvement in current assessments and greater caution about the outlook.

SurveyLarge manufacturers(Outlook)Large nonmanufacturersAssumed exchange rate
2025-Q4151234147.06
2026-Q1171536150.10
2026-Q2221437152.57

The diffusion index for large manufacturers improved by 5 points from the previous survey to 22, while the index for nonmanufacturers remained high at 37. The indices for medium-sized manufacturers at 17 and small manufacturers at 9 also moved in the direction of improvement. At the same time, the outlook index for large manufacturers was 14, eight points below the current assessment. Companies do not expect current favorable conditions to persist.

The assumed exchange rate for all enterprises and industries was 152.57 yen, revised toward yen depreciation from 150.10 yen in the previous survey. A weaker yen supports the earnings of exporters, while also creating an upward pressure on CPI through import prices. This tends to strengthen expectations for continued monetary-policy normalization and exert upward pressure on short-term rates, while on the fiscal side it leads to higher interest expenses. Improved corporate earnings tend to increase tax-revenue elasticity, but because actual tax-revenue data were not obtained for this analysis, the impact on corporate tax receipts cannot be verified in amount terms.

7. Implications for the Current Account — Insufficient Evidence to Assess the External Balance

The MOF’s trade statistics run through December 2025, limiting the assessment of the latest external balance. The balance recorded a surplus of 306.0 billion yen in November and 94.8 billion yen in December, moving into surplus for the second consecutive month after deficits of 277.7 billion yen in September and 242.9 billion yen in October. However, these data are eight months old and cannot be used to confirm the trend in 2026.

Japan’s fiscal sustainability conditions are determined not by monthly fluctuations in the trade balance but by two factors: the persistence of a current-account surplus and the structure in which JGBs are absorbed mainly domestically. Because the holder composition was not included in the data for this analysis, no level is presented; however, as long as domestic holdings remain predominant, higher interest expenses represent a domestic income transfer—from the government to households and financial institutions—rather than an external outflow. Under this structure, the fiscal cost of rising rates appears not as a currency-crisis risk but as pressure on other policy expenditures within the budget. The establishment of super-long yields in the 4% range this month indicates that this pressure is set to intensify over the medium term.

8. Outlook — The Direction of the r-g Differential and the Refinancing Time Lag

From a debt-dynamics perspective, this month’s data show that upward pressure on r, the effective interest rate, is continuing. The funding rate for a new 10-year bond is 2.943%, while super-long yields are in the 4.09% range; the 10-year yield has risen 81.1bp from six months earlier. However, the effective interest rate on the total debt stock is supported by the low coupons on outstanding bonds, and the increase will emerge gradually as refinancing progresses. This time lag provides a buffer for public finances.

g, nominal growth, is supported by the combination of July CPI inflation at 1.9% and improvement in the coincident index. Actual nominal GDP data are outside this article’s data coverage, so the sign of the r-g differential cannot be established numerically. What can be confirmed is an asymmetry: r is on a clear upward trend, while inflation appears to be plateauing in the upper 1% range. This suggests that the r-g differential may be moving in the direction of narrowing—that is, in an unfavorable direction for the debt ratio.

In the debt-dynamics equation, if the PB (primary balance) improves and pb(t) turns positive, the debt ratio will move in the direction of improvement even if r-g is unchanged. However, if the PB surplus remains below interest expenses, the fiscal deficit will continue and gross financing needs will remain elevated. Because the actual PB figure was not included in the data for this analysis, this article evaluates only changes in the interest-rate conditions. The three points of focus going forward are as follows.

  • The released August CPI figure: Whether the provisional real 10-year rate of 1.043% is revised upward or downward. This directly determines the real borrowing cost
  • The direction of the 10Y-2Y spread: It has narrowed for two consecutive months from 1.308% in June. If the narrowing continues, its interpretation as a business-cycle signal will need to be reconsidered
  • Whether the 30Y-10Y spread widens again: If the 18.1bp narrowing from January reverses, it would be an initial sign of a reassessment of the term premium at the super-long end

The combination seen this month—short-end-led rate increases and flattening at the super-long end—indicates that the focus of price formation has shifted from the term premium at the super-long end to the policy-rate path. Whether this structure persists will depend on maintaining nominal growth and on the pace at which the effective interest rate rises as refinancing proceeds.

Glossary

TermDefinition
Bear flatteningA phenomenon in which yields rise across all tenors (bear market = falling bond prices), while short-term yields rise more than long-term yields, flattening the yield curve. It typically occurs when expectations for monetary tightening strengthen.
Term spreadThe yield difference between different maturities. The 10Y-2Y spread primarily reflects the business cycle and policy-rate expectations, while the 30Y-10Y spread reflects the super-long-end term premium and fiscal confidence.
Real interest rateThe interest rate calculated by subtracting the inflation rate from the nominal yield. In this article, it is calculated by subtracting year-on-year CPI inflation from the nominal 10-year JGB yield. A negative real interest rate represents financial repression that erodes the real value of debt.
Financial repressionA condition in which a combination of suppressed nominal interest rates and inflation makes real interest rates negative, reducing the real burden of government debt. It involves an income transfer from savers to borrowers.
r-g differentialThe difference between the effective interest rate (r) and nominal GDP growth (g). It is the most important variable determining debt-to-GDP dynamics: when r<g, the debt ratio tends to decline automatically; when r>g, it tends to rise.
Primary balance (PB)The balance between revenue and expenditure excluding interest payments. A primary surplus means that the government’s finances are self-sustaining excluding interest payments on existing debt.
Gross financing needsThe total amount of financing required in a fiscal year, comprising the fiscal deficit plus the amount of maturing debt to be refinanced. It measures reliance on market funding and rollover risk.
QT (quantitative tightening)A policy under which a central bank reduces its holdings of government bonds. It transfers duration risk remaining in the market to private investors and can push up the term premium in the medium- to long-term sector.
Business conditions diffusion indexIn the BOJ Tankan, an index calculated by subtracting the percentage of companies responding “poor” from the percentage responding “good.” A larger positive figure indicates stronger corporate sentiment.

This column was automatically generated by AI integrating Ministry of Finance JGB interest rate data (and tax revenue data when available), Bank of Japan statistics, and e-Stat public statistics as a fiscal analysis resource. This is not a recommendation to buy or sell any financial instruments or government bonds. Please make investment decisions at your own responsibility and consult professionals as needed.

Source: Ministry of Finance Japan

This service uses government bond interest rate data published by the Ministry of Finance Japan, but the content of this service is not guaranteed by the Ministry of Finance.

Source: Bank of Japan

This service uses statistical data published by the Bank of Japan, but the content of this service is not guaranteed by the Bank of Japan.