## _wp1019

## Source details

**Canonical URL:** [_wp1019](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1019.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1019.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1019.pdf.json)

---

### Introduction
- Since the early 1990s, 10-year JGB yields declined from 7 percent to below 2 percent while net public debt rose from 20 percent of GDP to 60 percent of GDP.
- Since 2000, net public debt climbed to 90 percent of GDP, while long-term yields remained fairly stable at below 2 percent.
- In early 2009 JGB yields picked up following stimulus announcements but remained low by historical standard.
- With the general government deficit projected to stay around 10 percent of GDP in 2010, public debt will exceed 110 percent of GDP in net terms and 225 percent of GDP in gross terms.
- Near-term conclusion: structural factors (large household savings, stable institutional investors, strong home bias) are likely to persist and keep JGB yields low.
- Longer-horizon conclusion: population aging reducing savings inflows and financial reforms enhancing risk appetite will likely diminish market capacity to absorb public debt; fiscal consolidation becomes critical over the long horizon.

### Theoretical and Empirical Background
- Standard theory predicts a positive relation between deficits or debt and government bond yields; neoclassical frameworks and finite-horizon models produce a positive relation between debt and yields.
- Empirical literature ranges:
  - Impact of a 1 percent of GDP increase in fiscal deficit on long-term government bond yields: 10 to 60 basis points.
  - Impact of a 1 percent of GDP increase in debt on yields: at most 10 basis points.
- Ricardian equivalence findings are mixed in the literature.

### Cross-Country Regression Approach
- Dependent variable: nominal (or real) 10-year government bond yield.
- Main explanatory variables: general government primary deficit (percent of GDP) or general government debt (percent of GDP), measured at end of previous year (debt in net and gross terms).
- Controls: short-term interest rate (3-month deposit rate), inflation, real GDP growth. Data measured annually covering 1971–2008 (Japan, U.S., Canada) and shorter for other countries.
- Estimated equations:
  - nominal 10-year bond yield = β0 + β1 primary deficit + controls
  - real 10-year bond yield = β0 + β1 primary deficit + controls
  - nominal 10-year bond yield = β0 + β1 net debt + controls
  - nominal 10-year bond yield = β0 + β1 gross debt + controls

### Main Country-by-Country Findings
- Japan
  - Coefficients on primary deficit at most 0.1 and statistically insignificant.
  - Coefficients on net or gross debt turn negative.
  - Results robust to inclusion of U.S. 10-year yield as proxy for international liquidity and to excluding central bank-held central government debt.
  - Restricting sample to pre-deflation period (before 1995) gives coefficient on primary deficit around 0.2.
- U.S., U.K., Canada
  - Coefficients on primary deficit in the range of 0.20-0.40 and statistically significant at the 1 percent level.
  - Interpretation: an increase in the primary deficit by 1 percent of GDP leads to an increase in the 10-year yield by 15–40 basis points.
  - Coefficients on net/gross debt are negative or close to zero.
- France and Germany
  - Coefficients on primary deficit in the range of 0.15-0.30 and statistically significant at least at the 10 percent level.
  - Coefficients on net/gross debt are negative.
- Italy
  - Coefficients on primary deficit are negative.
  - Coefficients on net/gross debt are close to zero.

### Selected Exact Regression Results (highlights)
- Japan OLS nominal 10-year bond yield with primary deficit: R^2 = 0.08; Residual Unit Root Test (DF t stat) = -3.93? (table shows -4.03); sample size = 38.
- U.S. OLS nominal 10-year bond yield with primary deficit: R^2 = 0.39; residual t-statistic (reported) = (8.50)***; sample size = 38.
- U.K. OLS nominal 10-year bond yield with primary deficit: R^2 = 0.28; residual t-statistic (reported) = (3.92)***; sample size = 30.

### Panel Regressions (G7 pooled)
- Specification includes primary deficit and primary deficit * Japan dummy (instrumented with its first lag in IV specifications).
- Key results:
  - Coefficient on primary deficit for all countries: about 0.15–0.20.
  - Coefficient on Primary deficit * Japan dummy: between -0.1 and -0.25 and statistically significant.
  - Interpretation: an increase in the primary deficit by 1 percent of GDP has a smaller impact on JGB yields by 10–25 basis points compared to other countries; overall impact on JGB yields close to zero given the aggregate coefficient.
- Example: Instrumental specification reports primary deficit * Japan dummy = -0.23; R^2 = 0.16; sample size = 231.

### High-Frequency Analysis (monthly forecast-based)
- Data source: Consensus Economics Inc. means of private institutions' forecasts; sample period August 2006 - September 2009.
- Dependent variable: 12-month-ahead forecast for nominal 10-year government bond yield (percent).
- Main explanatory variable: forecast for overall fiscal deficit (percent of GDP) in next U.S. fiscal year.
- Controls: 12-month-ahead forecast for short-term interest rate (3-month deposit rate), forecasts for inflation and real GDP growth in next calendar year.
- Results:
  - Japan: coefficient on forecasted overall fiscal deficit ≈ -0.01; R^2 = 0.85; Residual Unit Root Test (DF t stat) = -3.41***; sample size = 38.
  - U.S.: coefficient on forecasted overall fiscal deficit = 0.09; R^2 = 0.92; Residual Unit Root Test (DF t stat) = -3.33***; sample size = 38.
  - Interpretation: coefficient for Japan is about zero and well below the U.S. coefficient (0.09).

### Interpretation and Caveats
- Results may be subject to omitted variable bias even after using the instrumental variable method.
- Today’s yield may be affected by various current business cycle variables not fully captured by growth or short-term monetary conditions included as controls.
- Long-horizon official deficit forecasts for Japan were unavailable for much of the sample; Consensus Economics forecasts have a short horizon (one year), so high-frequency results are likely significantly affected by the current business cycle.
- Cointegration among variables is assumed; p-values of the Dickey-Fuller t-statistics are typically over 0.7 across dependent and explanatory variables except for the real GDP growth forecast.

### Japan-specific factors explaining low sensitivity of JGB yields
- Large pool of household assets
  - Household saving rates were "over 10 percent" until around 1999 and declined sharply thereafter.
- Strong home bias
  - JGBs financed largely by domestic investors: "94 percent of holdings as of end-2008".
  - Share of currency and deposits in households’ financial assets is "55 percent (at end FY2008)" — compared with "16 percent in the U.S.".
- Large and stable institutional holders
  - Japan Post Bank and the Government Pension Investment Fund have invested about "¥250 trillion in JGBs (around 35 percent of the total JGBs)".
  - Bank of Japan (BoJ) held nearly "¥60 trillion of JGBs as of end-2008".
  - Specific holdings: Japan Post Bank invested "¥156 trillion in JGBs as of September 2008"; Government Pension Investment Fund held "¥82 trillion of JGBs as of end-2008".
- Recent large saving flows from the corporate sector
  - Corporate sector recorded financial surpluses during recovery from the 1990s crisis; surpluses driven by higher profits and declining prices of capital goods.
- FILP liabilities and gross public debt dynamics
  - FILP reform reduced FILP liabilities by "around half to about ¥200 trillion (40 percent of GDP) since 2000".
  - Including FILP liabilities, public gross debt has not increased over the past 10 years.

### Empirical re-estimation including Japan-specific factors (flow of funds, Q1 1998–Q1 2009)
- Data advantages: quarterly FILP liabilities and each sector’s financial assets available; 45 observations.
- Main explanatory variable: gross debt including FILP liabilities (percent of GDP).
- Additional controls: central government debt held by the BoJ (percent of GDP); net financial wealth held by the household and corporate sectors (percent of GDP); share of foreign holdings of JGBs; short-term interest rate; inflation rate; real GDP growth.
- Key estimation results (OLS):
  - Coefficient on gross debt (including FILP): around "0.02" — implying an increase in public debt by "1 percent of GDP" would result in a rise in JGB yields by "2 basis points".
  - Coefficient on central government debt held by the BoJ: positive and insignificant in some specifications.
  - Coefficients on net household and corporate financial wealth and the share of foreign JGB holdings: signs as expected and statistically significant.
- Implication: including Japan-specific factors explains why yields remained low despite increases in JGBs.

### Share of JGB holdings (selected figures)
- Government (includes FILP): Dec-00 "23.1", Dec-05 "15.9", Dec-06 "14.3", Dec-07 "13.8", Dec-08 "13.2" (percent).
- Bank of Japan (BoJ): Dec-00 "11.9", Dec-05 "14.0", Dec-06 "11.2", Dec-07 "9.5", Dec-08 "8.3" (percent).
- Financial institutions: Dec-00 "55.1", Dec-05 "59.9", Dec-06 "62.2", Dec-07 "62.0", Dec-08 "64.2" (percent).
- Overseas holdings: Dec-00 "5.9", Dec-05 "4.7", Dec-06 "5.5", Dec-07 "7.0", Dec-08 "6.4" (percent).
- Households: Dec-00 "2.2", Dec-05 "4.0", Dec-06 "4.8", Dec-07 "5.3", Dec-08 "5.3" (percent).
- Total JGBs outstanding (in trillions of yen): Dec-00 "391.2", Dec-05 "672.1", Dec-06 "674.6", Dec-07 "682.4", Dec-08 "697.3".

### Outlook for absorption capacity of the JGB market (risks and structural shifts)
- Household sector
  - Households finance at least "50 percent of the total JGBs" through banks (including Japan Post Bank) and pension funds when indirect channels are included; direct holdings are "5 percent" of the total outstanding.
  - Basic public pension finances about "10 percent of the total JGBs".
  - Household saving rate has been on a declining trend; "2.2 percent" is the household saving rate in FY2007.
  - Simulation: assuming household saving rate remains at "2.2 percent", gross public debt (including FILP liabilities) in "2015" could exceed gross households’ financial assets. Excluding FILP liabilities, cross-over could emerge around "2020".
  - After 2014 assumptions: real GDP growth and inflation converge to "1.2" and "1.0 percent", respectively; no consumption tax increase assumed; FILP liabilities assumed to stay at the same level as in 2008 for 2009 onwards.
- Financial sector
  - Government Pension Investment Fund no longer obligated to purchase FILP bonds and is looking to expand investment in risky assets; Japan Post Bank allowed to expand non-JGB investment.
  - A "10 percent" shift by these institutions from JGBs to other assets would amount to "¥20–30 trillion (4–6 percent of GDP)".
  - BoJ holds about "8 percent" of total JGBs (as of end-2008); BoJ’s increased monthly purchases in early 2009 helped stabilize markets, but unwinding monetary easing may require BoJ to scale back JGB holdings.
- Other domestic sectors
  - Corporate sector financial surpluses declined to around "1 percent of GDP" in both 2007 and 2008.
  - FILP lending has been cut substantially; further constraint on FILP bond issuances may be difficult.
- Foreign sector
  - Foreign holdings of JGBs are small; shifts in foreign investor behavior unlikely to have a significant impact for the time being.
  - Sharp increases in sovereign bond issuances globally could negatively impact JGB market through crowding-out.

### Conclusions and Policy Implications
- JGB yields could be more strongly tied to changes in the deficit and debt looking forward due to ongoing structural shifts in the JGB market.
- Gross public financing requirement is likely to remain substantial—reaching 50 percent of GDP in 2009 (including rollover of financing bills) and will increase further with rising public debt.
- To ensure smooth refinancing and stable absorption of debt:
  - The timing of debt issuance and maturity structure should be carefully matched to market conditions.
  - The government should maintain regular contacts with market participants to identify demand for maturities and communicate intentions.
- Over the medium-term, it is critical to establish a credible framework for ensuring fiscal sustainability, featuring a clear timetable for comprehensive tax and expenditure reforms to be implemented once the economy recovers.
- Structural and demographic risks (rapid population aging; shifts in institutional investors’ behavior) make sound public debt management and fiscal consolidation critical to maintain market stability.

*Source: IMF staff analysis in chapter 13 of the provided PDF content.*

### References .............................................................................................................

### _wp1019 - References .............................................................................................................

### Introduction
- Since the early 1990s, 10-year JGB yields declined from 7 percent to below 2 percent while net public debt rose from 20 percent of GDP to 60 percent of GDP. Since 2000, net public debt climbed to 90 percent of GDP, while long-term yields remained fairly stable at below 2 percent.
- In early 2009 JGB yields picked up following stimulus announcements but remained low by historical standard.
- With the general government deficit projected to stay around 10 percent of GDP in 2010, public debt will exceed 110 percent of GDP in net terms and 225 percent of GDP in gross terms.
- Near-term conclusion: structural factors (large household savings, stable institutional investors, strong home bias) are likely to persist and keep JGB yields low.
- Longer-horizon conclusion: population aging reducing savings inflows and financial reforms enhancing risk appetite will likely diminish market capacity to absorb public debt; fiscal consolidation becomes critical over the long horizon.

### Theoretical and Empirical Background
- Standard theory predicts a positive relation between deficits or debt and government bond yields; neoclassical frameworks and finite-horizon models produce a positive relation between debt and yields.
- Empirical literature ranges:
  - Impact of a 1 percent of GDP increase in fiscal deficit on long-term government bond yields: 10 to 60 basis points.
  - Impact of a 1 percent of GDP increase in debt on yields: at most 10 basis points.
- Ricardian equivalence findings are mixed in the literature (e.g., Seater (1993) vs. Gale and Orszag (2003)).

### Cross-Country Regression Approach
- Dependent variable: nominal (or real) 10-year government bond yield.
- Main explanatory variables: general government primary deficit (percent of GDP) or general government debt (percent of GDP), measured at end of previous year (debt in net and gross terms).
- Controls: short-term interest rate (3-month deposit rate), inflation, real GDP growth. Data measured annually covering 1971–2008 (Japan, U.S., Canada) and shorter for other countries.
- Estimated equations:
  - (1) nominal 10-year bond yield = β0 + β1 primary deficit + controls
  - (2) real 10-year bond yield = β0 + β1 primary deficit + controls
  - (3) nominal 10-year bond yield = β0 + β1 net debt + controls
  - (4) nominal 10-year bond yield = β0 + β1 gross debt + controls

### Main Country-by-Country Findings
- Japan:
  - Coefficients on primary deficit at most 0.1 and statistically insignificant.
  - Coefficients on net or gross debt turn negative.
  - Results robust to inclusion of U.S. 10-year yield as proxy for international liquidity and to excluding central bank-held central government debt.
  - Restricting sample to pre-deflation period (before 1995) gives coefficient on primary deficit around 0.2 (relatively large and significant), but still generally smaller than other countries’ coefficients.
- U.S., U.K., Canada:
  - Coefficients on primary deficit in the range of 0.20-0.40 and statistically significant at the 1 percent level.
  - Interpretation: an increase in the primary deficit by 1 percent of GDP leads to an increase in the 10-year yield by 15–40 basis points.
  - Coefficients on net/gross debt are negative or close to zero.
- France and Germany:
  - Coefficients on primary deficit in the range of 0.15-0.30 and statistically significant at least at the 10 percent level.
  - Coefficients on net/gross debt are negative.
- Italy:
  - Coefficients on primary deficit are negative.
  - Coefficients on net/gross debt are close to zero.

### Selected Exact Regression Results (highlights from tables)
- Japan OLS nominal 10-year bond yield with primary deficit: R^2 = 0.08; Residual Unit Root Test (DF t stat) = -3.93? (table shows -4.03) and sample size = 38. (See Table II.2 for full cell values.)
- U.S. OLS nominal 10-year bond yield with primary deficit: R^2 = 0.39; residual t-statistic (reported) = (8.50)***; sample size = 38.
- U.K. OLS nominal 10-year bond yield with primary deficit: R^2 = 0.28; residual t-statistic (reported) = (3.92)***; sample size = 30.
- France and Germany entries report R^2 values and unit-root test statistics per Table II.2; many debt coefficients reported negative with significant t-statistics for debt regressions.

### Panel Regressions (G7 pooled)
- Specification includes primary deficit and primary deficit * Japan dummy (instrumented with its first lag in IV specifications).
- Key results:
  - Coefficient on primary deficit for all countries: about 0.15–0.20.
  - Coefficient on Primary deficit * Japan dummy: between -0.1 and -0.25 and statistically significant.
  - Interpretation: an increase in the primary deficit by 1 percent of GDP has a smaller impact on JGB yields by 10–25 basis points compared to other countries; overall impact on JGB yields close to zero given the aggregate coefficient.
- Example table entries (Table II.3):
  - (1) Real 10-year bond yield OLS: primary deficit coefficient ~0.15-0.20; primary deficit * Japan dummy coefficient ~-0.21? (table reports ranges).
  - (2) Real 10-year bond yield Instrumental: primary deficit * Japan dummy = -0.23 (instrument = 1st Lag of Main Explanatory Variable); R^2 = 0.16; sample size = 231.

### High-Frequency Analysis (monthly forecast-based)
- Data source: Consensus Economics Inc. means of private institutions' forecasts; sample period August 2006 - September 2009.
- Dependent variable: 12-month-ahead forecast for nominal 10-year government bond yield (percent).
- Main explanatory variable: forecast for overall fiscal deficit (percent of GDP) in next U.S. fiscal year.
- Controls: 12-month-ahead forecast for short-term interest rate (3-month deposit rate), forecasts for inflation and real GDP growth in next calendar year.
- Results (Table II.4):
  - Japan: coefficient on forecasted overall fiscal deficit ≈ -0.01; R^2 = 0.85; Residual Unit Root Test (DF t stat) = -3.41***; sample size = 38.
  - U.S.: coefficient on forecasted overall fiscal deficit = 0.09; R^2 = 0.92; Residual Unit Root Test (DF t stat) = -3.33***; sample size = 38.
  - Interpretation: coefficient for Japan is about zero and well below the U.S. coefficient (0.09).

### Interpretation and Policy Implications (as presented)
- Empirical evidence supports that JGB yields are less sensitive to fiscal deficits than yields in other G7 countries.
- Structural factors underpinning low JGB yields: large household savings pool, stable institutional investors, strong home bias.
- Near-term policy implication: sound public debt management can help preserve market stability while structural factors persist.
- Long-term policy implication: fiscal consolidation will be critical to ensure smooth financing as demographic changes and financial reforms reduce market capacity to absorb public debt.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1019.pdf*

### 13.      All these results suggest that Japanese government bond yields are less sensitive to

### 13.      All these results suggest that Japanese government bond yields are less sensitive to fiscal deficits than in other countries, but they should be interpreted with caution

### Interpretation and caveats
- Results may be subject to omitted variable bias even after using the instrumental variable method.
- Today’s yield may be affected by various current business cycle variables not fully captured by growth or short-term monetary conditions included as controls.
- Ideally, regressions would use future (expected) yields regressed on official forecasts for deficits 5–10 years ahead, but for Japan such long-horizon official deficit forecasts are unavailable because the Japanese Cabinet Office started publishing 5-year forecasts for the deficit and the stock of debt only a few years ago.
- Consensus Economics forecasts have a short horizon (one year), so results (e.g., Table II.4) are likely significantly affected by the current business cycle.
- Cointegration among variables (except for the real GDP growth forecast) is assumed. P-values of the Dickey-Fuller t-statistics are typically over 0.7 across dependent and explanatory variables except for the real GDP growth forecast.

### Japan-specific factors explaining low sensitivity of JGB yields
- Large pool of household assets
  - Japan had household saving rates "over 10 percent" until around 1999 when they began to decline sharply.
  - Ample saving flows contributed to large accumulation of household financial assets, helping finance the build-up of public debt.
- Strong home bias
  - JGBs financed largely by domestic investors: "94 percent of holdings as of end-2008".
  - Share of currency and deposits in households’ financial assets is "55 percent (at end FY2008)" — compared with "16 percent in the U.S." — and a large part of these funds is invested in JGBs mainly through the banking sector.
- Large and stable institutional holders
  - Japan Post Bank and the Government Pension Investment Fund have invested about "¥250 trillion in JGBs (around 35 percent of the total JGBs)".
  - Bank of Japan (BoJ) held nearly "¥60 trillion of JGBs as of end-2008".
  - Specific holdings: Japan Post Bank invested "¥156 trillion in JGBs as of September 2008"; Government Pension Investment Fund held "¥82 trillion of JGBs as of end-2008".
- Recent large saving flows from the corporate sector
  - Corporate sector recorded financial surpluses during recovery from the 1990s crisis; surpluses driven by higher profits and declining prices of capital goods (IMF, 2006).
  - Financial surpluses defined as "a change in financial assets minus a change in financial debt".
- FILP liabilities and gross public debt dynamics
  - FILP reform reduced FILP liabilities by "around half to about ¥200 trillion (40 percent of GDP) since 2000".
  - Including FILP liabilities, public gross debt has not increased over the past 10 years.

### Empirical re-estimation including Japan-specific factors (flow of funds data, Q1 1998–Q1 2009)
- Data advantages: quarterly FILP liabilities and each sector’s financial assets (including foreign sector) available, yielding 45 observations.
- Main explanatory variable: gross debt including FILP liabilities (percent of GDP).
- Additional controls: central government debt held by the BoJ (percent of GDP); net financial wealth held by the household and corporate sectors (percent of GDP); share of foreign holdings of JGBs; short-term interest rate; inflation rate; real GDP growth.
- Estimated equation (dependent variable = nominal 10-year government bond yield):
  - nominal 10-year bond yield = β0 + β1 gross debt including FILP liabilities + controls
- Key estimation results (Table II.6, OLS):
  - Coefficient on gross debt (including FILP): around "0.02" — implying an increase in public debt by "1 percent of GDP" would result in a rise in JGB yields by "2 basis points".
  - Coefficient on central government debt held by the BoJ: positive and insignificant in some specifications.
  - Coefficients on net household and corporate financial wealth and the share of foreign JGB holdings: signs as expected and statistically significant.
- Implication: omitting Japan-specific factors leads to biased results; including them explains why yields remained low despite increases in JGBs.

### Share of JGB holdings (selected figures from Table II.5)
- Government (includes FILP): Dec-00 "23.1", Dec-05 "15.9", Dec-06 "14.3", Dec-07 "13.8", Dec-08 "13.2" (percent).
- Bank of Japan (BoJ): Dec-00 "11.9", Dec-05 "14.0", Dec-06 "11.2", Dec-07 "9.5", Dec-08 "8.3" (percent).
- Financial institutions: Dec-00 "55.1", Dec-05 "59.9", Dec-06 "62.2", Dec-07 "62.0", Dec-08 "64.2" (percent).
- Overseas holdings: Dec-00 "5.9", Dec-05 "4.7", Dec-06 "5.5", Dec-07 "7.0", Dec-08 "6.4" (percent).
- Households: Dec-00 "2.2", Dec-05 "4.0", Dec-06 "4.8", Dec-07 "5.3", Dec-08 "5.3" (percent).
- Total JGBs outstanding (in trillions of yen): Dec-00 "391.2", Dec-05 "672.1", Dec-06 "674.6", Dec-07 "682.4", Dec-08 "697.3".

### Outlook for absorption capacity of the JGB market (risks and structural shifts)
- Household sector
  - Households finance at least "50 percent of the total JGBs" through banks (including Japan Post Bank) and pension funds when indirect channels are included; direct holdings are "5 percent" of the total outstanding.
  - Basic public pension finances about "10 percent of the total JGBs" and is categorized as general government’s assets.
  - Household saving rate has been on a declining trend; "2.2 percent" is the household saving rate in FY2007.
  - Simulation: based on current trends and assuming household saving rate remains at "2.2 percent", gross public debt (including FILP liabilities) in "2015" could exceed gross households’ financial assets. Excluding FILP liabilities, cross-over could emerge around "2020".
  - After 2014 assumptions: real GDP growth and inflation converge to "1.2" and "1.0 percent", respectively; no consumption tax increase assumed; FILP liabilities assumed to stay at the same level as in 2008 for 2009 onwards.
- Financial sector
  - Government Pension Investment Fund no longer obligated to purchase FILP bonds and is looking to expand investment in risky assets; Japan Post Bank allowed to expand non-JGB investment.
  - A "10 percent" shift by these institutions from JGBs to other assets would amount to "¥20–30 trillion (4–6 percent of GDP)".
  - Decline in home bias among private financial institutions could reduce domestic absorption (Walker, 2005).
  - BoJ holds about "8 percent" of total JGBs (as of end-2008); BoJ’s increased monthly purchases in early 2009 helped stabilize markets, but unwinding monetary easing may require BoJ to scale back JGB holdings.
- Other domestic sectors
  - Corporate sector financial surpluses declined to around "1 percent of GDP" in both 2007 and 2008; profits and surpluses may turn negative near-term amid global slowdown.
  - FILP lending has been cut substantially; constraining FILP bond issuances further may be difficult, potentially tying future supply more directly to the overall deficit.
- Foreign sector
  - Foreign holdings of JGBs are small; shifts in foreign investor behavior unlikely to have a significant impact for the time being.
  - However, sharp increases in sovereign bond issuances across the world could negatively impact JGB market in the near-term (through crowding-out).

*Source: IMF staff analysis in chapter 13 of the provided PDF content.*

### 19.      As discussed in the previous section, JGB yields could be more strongly tied to

### V. CONCLUSIONS

### Market capacity, financing requirements, and debt issuance
- JGB yields could be more strongly tied to changes in the deficit and debt looking forward due to ongoing structural shifts in the JGB market.
- The gross public financing requirement is likely to remain substantial—reaching 50 percent of GDP in 2009 (including rollover of financing bills) and will increase further in line with rising public debt.
- To ensure the smooth refinancing of the debt, both short-term and medium-term measures are needed.
- To ensure stable absorption of debt by the market:
  - The timing of debt issuance and maturity structure should continue to be carefully matched to market conditions.
  - The government is making regular contacts with market participants to identify their demand for maturities; such efforts to communicate with market participants will remain critical.
- Over the medium-term, it is critical to establish a credible framework for ensuring fiscal sustainability, featuring a clear timetable for comprehensive tax and expenditure reforms to be implemented once the economy recovers.

### Historical context and risks to market stability
- Historically, Japan’s public debt has been financed in a fairly smooth manner. The large pool of household savings and the stable domestic institutional investor base have contributed to keeping yields steady despite the rapid rise in public debt.
- Structural and demographic risks:
  - Japan is undergoing rapid population aging, which will likely limit the market’s absorptive capacity of public debt.
  - Shifts in institutional investors’ behavior could serve to reduce inflows to the market.
- Policy implication: To maintain market stability, sound public debt management and fiscal consolidation will be critical.

### Empirical evidence on stationarity and unit roots (Appendix)
- A unit root test is clearly not rejected across countries for the nominal 10-year government bond yield and the general government net/gross debt (high p-values). The test results for the nominal short-term interest rate are more ambiguous, but all of these four variables are assumed to be nonstationary.
- Dickey-Fuller t Statistics (by country and variable):

  - Nominal 10-year bond yield
    - Japan: -0.60 (p-value 0.87)
    - U.S.: -0.86 (p-value 0.80)
    - U.K.: -0.68 (p-value 0.85)
    - France: -0.83 (p-value 0.81)
    - Germany: -1.21 (p-value 0.67)
    - Canada: -0.64 (p-value 0.86)
    - Italy: -0.84 (p-value 0.81)

  - Real 10-year bond yield
    - Japan: -2.80 (p-value 0.06)
    - U.S.: -1.91 (p-value 0.33)
    - U.K.: -2.78 (p-value 0.06)
    - France: -1.29 (p-value 0.63)
    - Germany: -4.28 (p-value 0.00)
    - Canada: -1.79 (p-value 0.38)
    - Italy: -1.65 (p-value 0.46)

  - Nominal short-term interest rate
    - Japan: -1.55 (p-value 0.51)
    - U.S.: -1.49 (p-value 0.54)
    - U.K.: -0.94 (p-value 0.78)
    - France: -1.39 (p-value 0.59)
    - Germany: -2.58 (p-value 0.10)
    - Canada: -1.51 (p-value 0.53)
    - Italy: -1.32 (p-value 0.62)

  - Real short-term interest rate
    - Japan: -2.14 (p-value 0.23)
    - U.S.: -2.01 (p-value 0.28)
    - U.K.: -2.28 (p-value 0.18)
    - France: -1.52 (p-value 0.52)
    - Germany: -4.60 (p-value 0.00)
    - Canada: -2.11 (p-value 0.24)
    - Italy: -2.07 (p-value 0.26)

  - Inflation
    - Japan: -2.18 (p-value 0.21)
    - U.S.: -1.92 (p-value 0.32)
    - U.K.: -1.82 (p-value 0.37)
    - France: -1.12 (p-value 0.71)
    - Germany: -4.86 (p-value 0.00)
    - Canada: -1.48 (p-value 0.54)
    - Italy: -1.27 (p-value 0.64)

  - Real GDP growth
    - Japan: -3.28 (p-value 0.02)
    - U.S.: -4.67 (p-value 0.00)
    - U.K.: -4.17 (p-value 0.00)
    - France: -4.01 (p-value 0.00)
    - Germany: -4.25 (p-value 0.00)
    - Canada: -4.07 (p-value 0.00)
    - Italy: -5.46 (p-value 0.00)

  - General government primary deficit
    - Japan: -1.47 (p-value 0.55)
    - U.S.: -2.24 (p-value 0.19)
    - U.K.: -1.50 (p-value 0.53)
    - France: -2.52 (p-value 0.11)
    - Germany: -2.82 (p-value 0.06)
    - Canada: -1.71 (p-value 0.43)
    - Italy: -1.20 (p-value 0.67)

  - General government net debt
    - Japan: 1.32 (p-value 1.00)
    - U.S.: -1.02 (p-value 0.75)
    - U.K.: -1.21 (p-value 0.67)
    - France: -1.64 (p-value 0.46)
    - Germany: -0.89 (p-value 0.79)
    - Canada: -0.92 (p-value 0.78)
    - Italy: -3.30 (p-value 0.01)

  - General government gross debt
    - Japan: 2.01 (p-value 1.00)
    - U.S.: -0.73 (p-value 0.84)
    - U.K.: -1.23 (p-value 0.66)
    - France: -1.08 (p-value 0.72)
    - Germany: -0.77 (p-value 0.83)
    - Canada: -0.84 (p-value 0.81)
    - Italy: -1.07 (p-value 0.73)

- Other variables (real 10-year bond yield, general government primary deficit, real short-term interest rate, inflation, and real GDP growth) are assumed to be stationary.

### Regression assumptions used in benchmark estimations
- (1) Dependent variable: Nominal 10-year bond yield
  - Main explanatory variable: General government primary deficit
  - Assumption: The 10-year bond yield and the nominal short-term interest rate are cointegrated.
- (2) Dependent variable: Real 10-year bond yield
  - Main explanatory variable: General government primary deficit
  - Assumption: All the variables (real 10-year bond yield, general government primary deficit, real short-term interest rate, inflation, and real GDP growth) are stationary.
- (3) Dependent variable: Nominal 10-year bond yield
  - Main explanatory variable: General government net debt
  - Assumption: The nominal 10-year bond yield, the net debt, and the short-term interest rate are cointegrated.
- (4) Dependent variable: Nominal 10-year bond yield
  - Main explanatory variable: General government gross debt
  - Assumption: The nominal 10-year bond yield, the gross debt, and the short-term interest rate are cointegrated.

*Source: IMF Working Paper (excerpt provided in the content unit).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1019.pdf_