## _wp12291

## Source details

**Canonical URL:** [_wp12291](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12291.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12291.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12291.pdf.json)

---

### Introduction: role and recent performance of SMEs
- SMEs account for nearly 70 percent of total employment and over half of manufacturing value-added.
- Over the last decade, SMEs "have no longer been a thriving source of growth": profitability and investment of SMEs have declined significantly and business registration of SMEs (in net terms) has been stagnant since the global financial crisis, despite very accommodative financing conditions.
- Paper's three questions:
  - What are the principal reasons of weak credit growth since the Lehman shock, especially among SMEs?
  - Why has SME restructuring been so slow and risk capital so limited?
  - What financial sector policies are needed to boost growth and what gains can be expected?

### Findings on weak credit growth
- Aggregate indicators of financial conditions:
  - Loan-to-deposit ratios at historic lows (near 70 percent) and the lowest among advanced countries.
  - Bond spreads returned to pre-crisis levels (less than ½ percent for AA-rated debt).
  - Bank of Japan Tankan diffusion index recovered to positive territory by late 2011, but bank attitudes towards larger firms recovered more than for small firms.
  - Credit flows: credit barely grew in early 2012 for large corporations and fell by 1 percent y/y for SMEs (decline narrowed later).
- Principal contributors to weak credit growth:
  - Macroeconomic conditions:
    - Weak growth prospects, slowing consumer demand, stagnant land prices (particularly outside metropolitan areas).
    - Lingering deflation and interest rates at the zero lower bound limited the decline of real interest rates relative to other countries (Figure 2).
    - Structural factors: population aging and shrinking rural economy served mainly by regional banks.
  - Firms deleveraging:
    - Firms scaled back capital investment and shifted toward internal financing.
    - Internal financing (retained earnings and depreciation) increased from 60 percent of external financing to 150 percent during 2008–10.
  - Low SME profitability and legacy effects:
    - SMEs have a lower net profit margin (adjusted by capital ratio) at about 1½ percent versus an average of 8 percent for other advanced economies.
    - SMEs still have high debt at 250–300 percent of equity; about 2½ times higher than large firms.
    - Loans to SMEs still account for about half of bank credit.
    - SMEs with credit guarantees tend to take longer to repay current debt and are more likely to incur losses.

### Evidence on SME financial structure and crisis impact
- Summary statistics (averages from 2000 by sector and size of capital, Table 1):
  - Capital / Total assets: All 6.7; Large 9.6; Medium 4.0; Small 3.8 (Nonmanufacturing: All 7.9; Large 10.4; Medium 3.9; Small 3.5).
  - Return on assets: All 0.8; Large 0.9; Medium 0.9; Small 0.7.
  - Debt equity ratio: All 222.1; Large 168.2; Medium 278.9; Small 307.3 (Manufacturing: All 138.6; Large 114.9; Medium 177.2; Small 208.1).
  - Interest rate coverage (in times): All 1.0; Large 0.9; Medium 1.0; Small 1.0.
  - Adjusted net profit margin: All 3.3; Large 6.2; Medium 1.6; Small 1.5 (Nonmanufacturing: All 4.4; Large 6.6; Medium 1.9; Small 1.4).
  - Labor productivity (value-added per employee): All 0.3; Large 0.8; Medium 0.3; Small 0.2.
- SME heterogeneity by risk rating (Table 2 highlights):
  - Full sample: Debt equity ratio 1.75; Return on assets 0.30; Interest coverage ratio 0.82; Borrowing cost 2.21; Repayment capacity 0.09.
  - After 2008: Debt equity ratio 1.28; Return on assets 0.14; Interest coverage ratio 0.44; Borrowing cost 2.06; Repayment capacity 0.08.
  - Risk-rating gradients example:
    - Risk rating 1: Debt equity ratio 11.24; Return on assets 3.57; Interest coverage 8.39; Borrowing cost 1.79; Repayment capacity 0.39.
    - Risk rating 10: Debt equity ratio -1.74; Return on assets -1.85; Interest coverage -0.30; Borrowing cost 2.99; Repayment capacity 0.02.
- Regression and time-series evidence on profitability (Table 3, Figure 3):
  - Lagged debt-equity ratio coefficient negative and statistically significant (e.g., -0.26 in specification (1) full sample).
  - Lagged value-added per employee positive (e.g., 1.60 in specification (1)).
  - Dummy variables indicate manufacturing and larger firms have higher profit margins than nonmanufacturing and smaller firms after controlling for yearly cycles.
  - Interactive-year dummy results:
    - Firm size: Profitability premium between large and small/medium firms widened steadily to about 2–3 percentage points over the last 30 years. Current profitability levels for small enterprises are at historic lows; large enterprises show an increasing trend since the early 1970s except for a temporary fall during the peak of the global financial crisis.
    - Sector: Manufacturing firms tended to have higher profitability than nonmanufacturing firms, while nonmanufacturing recorded steady improvement in profitability in the 2000s. The decline in manufacturing profit margin could be due to yen appreciation and weaker external demand.

### Financial performance of SMEs by risk groups (2000–11)
- Risk groups divided into ten deciles (1 = best creditworthiness; 10 = least). Groups "8" to "10" indicate nonperforming SMEs. Data separated into pre-crisis (2000-07) and post-crisis (2008-11).
- Key findings:
  - Median return on assets has fallen across all risk classes after the global financial crisis.
    - Decline for SMEs with good credit has been marginal.
    - Decline for SMEs in risk groups 7th-10th deciles was substantial and continued after 2008.
    - Overall median return on assets has been low at about 1 percent.
  - Median debt-equity ratio deteriorated significantly for high-risk SMEs.
    - Median debt-equity ratio has turned negative for SMEs in risk groups 7th to 10th—an indication of negative equity.
    - For SMEs with good credit worthiness, the debt-equity ratio has slightly declined to about 2 times on average.
  - Interest rate coverage declined after the global financial crisis, reducing repayment capacity.
    - SMEs in risk groups 8th-10th had interest coverage close to zero on average before the global financial crisis and saw coverage ratios turn negative after 2008.

### Borrowing costs, breakeven rates, and role of credit support
- Observations:
  - Median borrowing costs for SMEs declined by 20–30 basis points after the global financial crisis across risk groups by similar magnitudes.
  - The difference in median borrowing costs between the lowest and highest risk groups is about 100 basis points.
  - Borrowing costs eased across the board, and interest rates do not fully reflect underlying credit risks among weak SMEs.
- Comparison of measures:
  - Firms’ loanable rates are derived from interest coverage data (interest rates firms can afford).
  - Actual borrowing costs are largely flat across risk groups.
  - Bank’s breakeven rates (estimated by Shirota, Imakubo, and Nishioka (2011)) account for funding costs and credit risks based on nonperforming loans and CRD data.
- Specific findings:
  - SME borrowing rates slightly exceeded banks’ breakeven rates for strong SMEs — indicating a thin but positive net interest margin.
  - Borrowing costs of weak SMEs are far below banks’ breakeven rate once accounting for credit risks; some credit risks were borne elsewhere, such as through credit guarantees.
  - Firms’ loanable rates for risk groups 8th-10th are on average lower than actual borrowing costs, implying these SMEs are, on average, not able to afford borrowings at current interest rates and could have faced credit rationing if banks extended loans solely on market perspectives.
  - SMEs in the 8th risk group could afford higher payable interest rates in the early- and mid-2000s than in 2009-10, but in both periods their borrowing costs were much lower.

### Impact and design of credit support measures
- Credit support measures likely eased financing conditions but may have unintended consequences:
  - After FY2009-10, firms’ payable rates shifted down, suggesting general weakening of SME financial soundness while borrowing costs and banks’ breakeven rates remained largely unchanged.
  - Credit support measures likely prevented a systemic tightening of financing conditions among SMEs across all risk groups but tended to protect SMEs in higher risk groups that could be nonviable and should have exited or been restructured.
  - Part of the risks was borne by various credit support measures beyond banks and SMEs.
- Program magnitudes and design:
  - Public guarantees increased from about 4 to 6 percent of GDP over the past four years.
  - Guarantee limit: ¥280 million (of which: unsecured loans ¥80 million).
  - Guarantee percentage: 80 percent of loans under accountability-sharing system; 100 percent (in some programs).
  - Guarantee fees: 1.15 percent on average; about 0.8-1.0 percent on average (for special/emergency).
- Indicators: cumulative use of safety-net guarantees shown in trillions of yen; number of guarantees tracked cumulatively.

### Economic costs and behavioral effects of public credit support
- Weakening credit risk assessment:
  - Guarantees covering a large share of loans (in some cases up to 100 percent) reduce banks’ incentives to assess and take on credit risk.
- Creating negative stigma:
  - Over half of SMEs that received some support (e.g., loan modification) reported the support would have a negative impact on their ability to obtain new borrowing.
- Understating credit risk:
  - Relaxation of classification requirements for restructured loans could understate true credit risk.
  - Bank of Japan estimates banks’ NPL ratio would rise by about 1 percentage point if firms with restructured loans failed to recover.
  - Higher proportion of loans reclassified under the Act: about 3-6 percent of total credit in regional banks and credit cooperatives relative to 1.7 percent in major banks.
- Fiscal cost precedent:
  - A previous special guarantee program in the late 1990s and early 2000s cost the government about ¥2 trillion (½ percent of GDP), relative to total credit guaranteed of ¥40 trillion.
- Crowding out and reduced restructuring incentives:
  - Keeping nonviable SMEs afloat through credit support can reduce credit supply for startups and limit incentives for creditors and borrowers to restructure.
  - Creditors can roll over guaranteed loans without requiring business improvements that restructuring would necessitate.

### Slow restructuring dynamics and bank incentives
- Entry and exit:
  - Entry and exit rates of SMEs are on average at about one-third of those in other advanced countries.
  - About 7,000 SMEs (less than 1 percent of total SMEs) applied for business rehabilitation during 2000–09 under new procedures; only 40 percent of those achieved rehabilitation.
- Factors limiting restructuring:
  - Small size of individual SME loans makes case-by-case restructuring costly.
  - Recognizing losses reduces profit and capital, especially for smaller regional/shinkin banks that:
    - Generally lack expertise in business restructuring.
    - Face higher nonperforming loan ratios: 4–6 percent relative to 2 percent among major banks.
    - More than half the larger (tier 1) regional banks and most small (tier 2) regional banks have capital adequacy ratios below that of the major banks.
  - Multiple creditors and widespread use of personal guarantees complicate voluntary workouts or business transfers.
  - Bankruptcy carries a high social stigma in Japan.
- Aggregate assessment:
  - Slow restructuring of SMEs is attributable to limited incentives and some credit support measures.
  - Continued survival of nonviable SMEs could impede credit intermediation and growth potentials by limiting risk capital and reducing credit availability to startups and viable firms when demand recovers.

### Availability of risk-based capital: findings and constraints
- Start-ups rely predominantly on self-finance; available funds (median value) raised from this source are small relative to other financing sources.
- Venture capital availability limited: in 2009, Japan ranked second to last in venture capital investment as a share of GDP among OECD countries.
- Business transfer rates, initial public offerings, and securitization of SME loans are low by international standards.
- Market- and institution-level practices constraining risk-based financing:
  - Banks’ preference for fixed-asset collateral and personal guarantees limits risk-based lending.
  - Personal guarantees (historically unlimited) legally constrained by the Act for Partial Revision of the Civil Code (2005) to require maximum guarantee amounts and to limit guarantee obligations to within five years (within three years when no period is specified).
  - Claims with personal guarantees are not easily transferable; more than a quarter of SMEs consider personal guarantees an obstacle to business transfers (JSBRI 2011).
- Institutional investor behavior limits risk funding:
  - Public pension funds hold a much higher share of assets in government bonds and have no exposure to property or alternative investments, contributing to low provision of risk capital.
  - Household financial assets are about five times disposable income, the highest among the OECD, but concentrated in cash and deposits.
  - Financial institutions allocate most assets to government securities.
- Asset-based lending (ABL) and ERMC:
  - GFIs and BOJ have promoted ABL, but many financial institutions—particularly smaller regional/shinkin banks—have limited expertise in ABL.
  - Receivable-backed financing is well established; inventory-based and current-asset based financing are limited by difficulty in assessing fair value and double assignment risks.
  - The Act on Electronically Recorded Monetary Claims (ERMC) introduced in FY2007 may help stipulate rights on claims and avoid double assignment, but banks face sizeable investment and operating costs to use it.
- Credit information and market practices:
  - Over half of surveyed financial institutions consider assessing SME credit risks challenging due to limited information and untimely disclosure.
  - Credit registries include the Credit Risk Database (CRD), Risk Data Bank of Japan (RDB), Credit Risk Information Total Services (CRITS), and shinkin data bank (SDB).
  - Designated credit bureaus under the Money Lending Business Act include Japan Credit Information Reference Center Corp. (JICC) and Credit Information Center (CIC).
  - No centralized credit bureau for consumer loans; absence of a tax identification number hinders credit information sharing and likely creates a financing gap for the “middle-risk” group (Schaede, 2005).

### Policy recommendations (financial sector measures)
- Gradually phase out credit support measures:
  - Phase out special credit guarantees with full coverage of loan value as the recovery takes hold.
  - Reduce the normal guarantee coverage ratio from 80 to 60 percent over time (in line with international averages (IMF, 2006)) and scrutinize the rollover of guarantees to ensure market discipline.
  - Phase out the temporary SME Financing Facilitation Act by March 2013 so banks would not have an obligation (but retain the option) to restructure loans at SMEs’ request, reducing incentives to classify restructured loans as ‘normal.’
- Encourage consolidation of regional and shinkin banks by raising capital requirements:
  - Increase the capital requirement for domestic-oriented banks from 4 percent at present.
  - Example: less than 5 percent of smaller banks would have a capital shortfall if the capital requirement for those banks was raised to 6 percent of risk-weighted assets (still below the minimum requirement of internationally-active banks at 8 percent).
  - Near-term risk: possible contraction in credit as regional banks deleverage; mitigation options include equity issuance, lower dividend payouts, or temporary public capital injection.
  - Medium-term benefit: higher capital would strengthen institutions’ capacity to take on risk.
- Accelerate SME restructuring:
  - Consider refocusing the public asset management company to advance SME restructuring.
  - Use debt-equity swaps to incentivize banks and SMEs for out-of-court voluntary workouts (Laryea, 2010).
- Promote risk-based financing and deepen capital markets:
  - Authorities (government and municipalities) could lead in originating and trading electronically-registered claims to avoid double assignment risks.
  - Scale down and better target current support to SMEs by redirecting GFIs’ lending away from reinsuring SME guarantees toward risk-based lending; studies show startups obtaining GFIs’ financing tend to outperform (Fukanuma, Tadanobu, and Watanabe, 2006).
  - Revise investment restrictions for institutional investors (e.g., pension funds)—once supervision and regulatory framework are strengthened—to encourage alternative investments such as securitized loans and venture capital.
  - Development of capital markets for securitized loans could improve long-term returns and channel funds toward private investment.
- Streamline regulatory measures to reduce time and cost of starting businesses:
  - Simplify business registration to promote business creation and transfers.
  - Broaden coverage of credit registries through a more centralized database and include a consumer data bureau for personal credit information.
  - Facilitate by linking the proposed taxpayer identification system to identification of credit information of firms and individuals.

### Quantified macroeconomic impact (modeling)
- IMF GIMF model simulation:
  - Addressing SME structural weaknesses could raise aggregate productivity by about ¼ percentage points.
  - This increase in aggregate productivity would lift long-term growth by 0.1–0.2 percentage points from the baseline.
  - Modeling assumption: productivity of smaller firms would increase to about 80 percent of that of large firms, producing the ¼ percentage point aggregate productivity gain.

### Conclusion: policy rationale and expected outcomes
- To encourage supply of risk-based capital, costly government support measures should be phased out and SME restructuring accelerated.
- Deepening capital markets and addressing regulatory barriers to starting businesses are necessary to enhance risk capital availability.
- Addressing SME weaknesses is expected to improve private investment, enhance firm productivity, and lift growth.

*Source: _wp12291 - References (extracted content).*

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

### _wp12291 - References

### Introduction: role and recent performance of SMEs
- SMEs account for nearly 70 percent of total employment and over half of manufacturing value-added.
- Over the last decade, SMEs "have no longer been a thriving source of growth": profitability and investment of SMEs have declined significantly and business registration of SMEs (in net terms) has been stagnant since the global financial crisis, despite very accommodative financing conditions.
- Paper's three questions:
  - What are the principal reasons of weak credit growth since the Lehman shock, especially among SMEs?
  - Why has SME restructuring been so slow and risk capital so limited?
  - What financial sector policies are needed to boost growth and what gains can be expected?

### Findings on weak credit growth
- Aggregate indicators of financial conditions:
  - Loan-to-deposit ratios at historic lows (near 70 percent) and the lowest among advanced countries.
  - Bond spreads returned to pre-crisis levels (less than ½ percent for AA-rated debt).
  - Bank of Japan Tankan diffusion index recovered to positive territory by late 2011, but bank attitudes towards larger firms recovered more than for small firms.
  - Credit flows: credit barely grew in early 2012 for large corporations and fell by 1 percent y/y for SMEs (decline narrowed later).
- Principal contributors to weak credit growth:
  - Macroeconomic conditions:
    - Weak growth prospects, slowing consumer demand, stagnant land prices (particularly outside metropolitan areas).
    - Lingering deflation and interest rates at the zero lower bound limited the decline of real interest rates relative to other countries (Figure 2).
    - Structural factors: population aging and shrinking rural economy served mainly by regional banks.
  - Firms deleveraging:
    - Firms scaled back capital investment and shifted toward internal financing.
    - Internal financing (retained earnings and depreciation) increased from 60 percent of external financing to 150 percent during 2008–10.
  - Low SME profitability and legacy effects:
    - SMEs have a lower net profit margin (adjusted by capital ratio) at about 1½ percent versus an average of 8 percent for other advanced economies.
    - SMEs still have high debt at 250–300 percent of equity; about 2½ times higher than large firms.
    - Loans to SMEs still account for about half of bank credit.
    - SMEs with credit guarantees tend to take longer to repay current debt and are more likely to incur losses.

### Evidence on SME financial structure and crisis impact
- Summary statistics (averages from 2000 by sector and size of capital, Table 1):
  - Capital / Total assets: All 6.7; Large 9.6; Medium 4.0; Small 3.8 (Nonmanufacturing: All 7.9; Large 10.4; Medium 3.9; Small 3.5).
  - Return on assets: All 0.8; Large 0.9; Medium 0.9; Small 0.7.
  - Debt equity ratio: All 222.1; Large 168.2; Medium 278.9; Small 307.3 (Manufacturing: All 138.6; Large 114.9; Medium 177.2; Small 208.1).
  - Interest rate coverage (in times): All 1.0; Large 0.9; Medium 1.0; Small 1.0.
  - Adjusted net profit margin: All 3.3; Large 6.2; Medium 1.6; Small 1.5 (Nonmanufacturing: All 4.4; Large 6.6; Medium 1.9; Small 1.4).
  - Labor productivity (value-added per employee): All 0.3; Large 0.8; Medium 0.3; Small 0.2.
- SME heterogeneity by risk rating (Table 2 highlights):
  - Full sample: Debt equity ratio 1.75; Return on assets 0.30; Interest coverage ratio 0.82; Borrowing cost 2.21; Repayment capacity 0.09.
  - After 2008: Debt equity ratio 1.28; Return on assets 0.14; Interest coverage ratio 0.44; Borrowing cost 2.06; Repayment capacity 0.08.
  - Risk-rating gradients (lower rating = higher creditworthiness): e.g., Risk rating 1: Debt equity ratio 11.24; Return on assets 3.57; Interest coverage 8.39; Borrowing cost 1.79; Repayment capacity 0.39; Risk rating 10: Debt equity ratio -1.74; Return on assets -1.85; Interest coverage -0.30; Borrowing cost 2.99; Repayment capacity 0.02.
- Regression and time-series evidence on profitability (Table 3, Figure 3):
  - Lagged debt-equity ratio coefficient negative and statistically significant (e.g., -0.26 in specification (1) full sample).
  - Lagged value-added per employee positive (e.g., 1.60 in specification (1)).
  - Dummy variables indicate manufacturing and larger firms have higher profit margins than nonmanufacturing and smaller firms after controlling for yearly cycles.
  - Two empirical results from interactive-year dummies:
    - Firm size: Profitability premium between large and small/medium firms widened steadily to about 2–3 percentage points over the last 30 years. Current profitability levels for small enterprises are at historic lows; large enterprises show an increasing trend since the early 1970s except for a temporary fall during the peak of the global financial crisis.
    - Sector: Manufacturing firms tended to have higher profitability than nonmanufacturing firms, while nonmanufacturing recorded steady improvement in profitability in the 2000s. The decline in manufacturing profit margin could be due to yen appreciation and weaker external demand.

### Reasons for slow restructuring and limited risk-based capital
- Limited restructuring or exits of nonviable SMEs despite low profitability and high leverage:
  - Creditors and nonviable SMEs have little incentive to restructure loans because refinancing is easy and several credit support measures (e.g., full-value credit guarantees) reduce incentives for voluntary workouts or business transfers.
  - Recognizing losses would hit banks’ profit and capital, especially smaller regional/shinkin banks that generally lack expertise in large-scale business restructuring and face higher nonperforming loans.
  - Reliance on collaterals and personal guarantees has limited the availability of risk-based capital.

### Policy implications and recommendations
- To encourage supply of risk-based capital and support growth, financial policies should:
  - Accelerate SME restructuring.
  - Phase out costly government support measures (e.g., full-value credit guarantees).
  - Strengthen the capital base of smaller banks to foster consolidation.
  - Deepen capital markets to improve availability of risk capital.
  - Address regulatory barriers to starting businesses.

### Paper organization and empirical approach
- The paper:
  - Discusses reasons behind weak credit growth and presents firm-level and sectoral balance-sheet evidence.
  - Illustrates reasons and implications of slow SME restructuring and limited risk capital.
  - Discusses policy options to resolve problems and quantifies possible growth gains.
- Empirical approach:
  - Sectoral panel analysis from 1960Q1 to 2011Q3.
  - Adjusted net profit margin used as dependent variable, with lagged debt-equity ratio, lagged value-added per employee, lagged interest cost, and dummy variables for sector and firm-size.
  - Yearly dummy variables included to capture macroeconomic/time effects; interactions with firm-size and sector used to identify structural differences over time.

*Source: _wp12291 - References (extracted content).*

### 11.    The global financial crisis further weakened the financial performance across

### _wp12291 - 11.    The global financial crisis further weakened the financial performance across

### Financial performance of SMEs by risk groups
- Risk groups divided into ten deciles according to SME credit worthiness (with “1” indicating the best credit worthiness and “10” the least). Groups "8" to "10" indicate nonperforming SMEs. Data span from 2000-11 annually and are separated into two sample periods: prior to the global financial crisis (2000-07) and after the global financial crisis (2008-11). Figures indicate median values unless specified.
- Key findings:
  - The median return on assets has fallen across all risk classes after the global financial crisis.
    - The decline for SMEs with good credit has been marginal.
    - The decline for SMEs in risk groups 7th-10th deciles was substantial and continued after 2008.
    - The overall median return on assets has been low at about 1 percent.
  - The median debt-equity ratio deteriorated significantly for high-risk SMEs.
    - The median debt-equity ratio has turned negative for SMEs in risk groups 7th to 10th—an indication of negative equity.
    - For SMEs with good credit worthiness, the debt-equity ratio has slightly declined to about 2 times on average.
  - Interest rate coverage declined after the global financial crisis, reducing repayment capacity.
    - SMEs in risk groups 8th-10th had interest coverage close to zero on average before the global financial crisis and saw coverage ratios turn negative after 2008.

### Borrowing costs, breakeven rates, and role of credit support
- Observations on interest rates and credit risks:
  - Median borrowing costs for SMEs declined by 20–30 basis points after the global financial crisis across risk groups by similar magnitudes.
  - The difference in median borrowing costs between the lowest and highest risk groups is about 100 basis points.
  - Borrowing costs eased across the board, and interest rates do not fully reflect underlying credit risks among weak SMEs.
- Comparison of actual borrowing costs, firms’ loanable (payable) rates, and banks’ breakeven rates:
  - Firms’ loanable rates are derived from interest coverage data (interest rates firms can afford).
  - Actual borrowing costs are largely flat across risk groups.
  - Bank’s breakeven rates (estimated by Shirota, Imakubo, and Nishioka (2011)) implicitly account for funding costs and credit risks based on nonperforming loans and CRD data.
- Specific findings:
  - SME borrowing rates slightly exceeded banks’ breakeven rates for strong SMEs — indicating a thin but positive net interest margin.
  - Borrowing costs of weak SMEs are far below banks’ breakeven rate once accounting for credit risks; some credit risks were borne elsewhere, such as through credit guarantees.
  - Firms’ loanable rates for risk groups 8th-10th are on average lower than actual borrowing costs, implying these SMEs are, on average, not able to afford borrowings at current interest rates and could have faced credit rationing if banks extended loans solely on market perspectives.
  - SMEs in the 8th risk group could afford higher payable interest rates in the early- and mid-2000s than in 2009-10, but in both periods their borrowing costs were much lower.

### Impact of credit support measures and implications
- Credit support measures (safety-net lending, special credit guarantees, SME Financing Facilitation Act revisions) likely eased financing conditions but may have unintended consequences:
  - After FY2009-10, firms’ payable rates shifted down, suggesting general weakening of SME financial soundness while borrowing costs and banks’ breakeven rates remained largely unchanged.
  - Credit support measures likely prevented a systemic tightening of financing conditions among SMEs across all risk groups but tended to protect SMEs in higher risk groups that could be nonviable and should have exited or been restructured.
  - Part of the risks was borne by various credit support measures beyond banks and SMEs.
- Magnitudes and program design details noted:
  - Public guarantees increased from about 4 to 6 percent of GDP over the past four years.
  - Guarantee design highlights:
    - Guarantee limit: ¥280 million (of which: unsecured loans ¥80 million).
    - Guarantee percentage: 80 percent of loans under accountability-sharing system; 100 percent (in some programs).
    - Guarantee fees: 1.15 percent on average; about 0.8-1.0 percent on average (for special/emergency).
- Indicators of public support usage:
  - Cumulative use of safety-net guarantees shown in trillions of yen; number of guarantees (RHS) tracked cumulatively.

### Economic costs and behavioral effects of public credit support
- Notable economic costs and behavioral effects:
  - Weakening credit risk assessment:
    - Guarantees covering a large share of loans (in some cases up to 100 percent) reduce banks’ incentives to assess and take on credit risk.
  - Creating negative stigma:
    - Over half of SMEs that received some support (e.g., loan modification) reported the support would have a negative impact on their ability to obtain new borrowing.
  - Understating credit risk:
    - Relaxation of classification requirements for restructured loans could understate true credit risk.
    - Bank of Japan estimates banks’ NPL ratio would rise by about 1 percentage point if firms with restructured loans failed to recover.
    - Higher proportion of loans reclassified under the Act: about 3-6 percent of total credit in regional banks and credit cooperatives relative to 1.7 percent in major banks.
  - Fiscal cost precedent:
    - A previous special guarantee program in the late 1990s and early 2000s cost the government about ¥2 trillion (½ percent of GDP), relative to total credit guaranteed of ¥40 trillion.
  - Crowding out and reduced restructuring incentives:
    - Keeping nonviable SMEs afloat through credit support can reduce credit supply for startups and limit incentives for creditors and borrowers to restructure.
    - Creditors can roll over guaranteed loans without requiring business improvements that restructuring would necessitate.

### Slow restructuring dynamics and bank incentives
- Evidence of slow SME restructuring:
  - Entry and exit rates of SMEs are on average at about one-third of those in other advanced countries.
  - About 7,000 SMEs (less than 1 percent of total SMEs) applied for business rehabilitation during 2000–09 under new procedures; only 40 percent of those achieved rehabilitation.
- Factors limiting restructuring:
  - Small size of individual SME loans makes case-by-case restructuring costly.
  - Recognizing losses reduces profit and capital, especially for smaller regional/shinkin banks that:
    - Generally lack expertise in business restructuring.
    - Face higher nonperforming loan ratios: 4–6 percent relative to 2 percent among major banks.
    - More than half the larger (tier 1) regional banks and most small (tier 2) regional banks have capital adequacy ratios below that of the major banks.
  - Multiple creditors and widespread use of personal guarantees complicate voluntary workouts or business transfers.
  - Bankruptcy carries a high social stigma in Japan.
- Aggregate assessment:
  - The slow restructuring of SMEs is attributable to limited incentives and some credit support measures.
  - Continued survival of nonviable SMEs could impede credit intermediation and growth potentials by limiting risk capital and reducing credit availability to startups and viable firms when demand recovers.

*Sources: Credit Risk Database, Bank of Japan, and authors' estimates.*

### 22.    Availability of risk-based capital is very limited despite plenty of liquidity. Start-

### 22. Availability of risk-based capital is very limited despite plenty of liquidity

### Limited availability of risk-based capital: findings
- Start-ups rely predominantly on self-finance; available funds (median value) raised from this source are small relative to other financing sources (Figure 6).
- Although there are no legal restrictions on raising venture capital funding, availability is limited: in 2009, Japan ranked second to last in venture capital investment as a share of GDP among OECD countries.
- Business transfer rates, initial public offerings, and securitization of SME loans are low by international standards.
- Several market- and institution-level practices constrain risk-based financing:
  - Banks’ preference for fixed-asset collateral and personal guarantees limits risk-based lending. Personal guarantees (historically unlimited) have been legally constrained by the Act for Partial Revision of the Civil Code (2005) to require maximum guarantee amounts and to limit guarantee obligations to within five years (within three years when no period is specified).
  - Claims with personal guarantees are not easily transferable due to legal provisions; more than a quarter of SMEs consider personal guarantees an obstacle to business transfers (JSBRI 2011).
- Institutional investor behavior limits risk funding: public pension funds hold a much higher share of assets in government bonds and have no exposure to property or alternative investments (text figure), contributing to low provision of risk capital.
- Risk appetite in the financial system is tepid: household financial assets are about five times disposable income, the highest among the OECD, but are concentrated in cash and deposits; financial institutions allocate most assets to government securities.
- Asset-based lending (ABL) and other risk-based methods:
  - GFIs and BOJ have promoted ABL, but many financial institutions—particularly smaller regional/shinkin banks—have limited expertise in ABL (Figure 7).
  - Receivable-backed financing is well established; inventory-based and current-asset based financing are limited by difficulty in assessing fair value and double assignment risks.
  - The Act on Electronically Recorded Monetary Claims (ERMC) introduced in FY2007 may help stipulate rights on claims and avoid double assignment, but banks face sizeable investment and operating costs to use it.

### Factors contributing to limited risk-based capital
- Lack of timely credit information:
  - Over half of surveyed financial institutions consider assessing SME credit risks challenging due to limited information and untimely disclosure.
  - Credit registries in Japan include the Credit Risk Database (CRD), Risk Data Bank of Japan (RDB), Credit Risk Information Total Services (CRITS), and shinkin data bank (SDB).
  - Financial institutions can access credit risk databases from designated credit bureaus under the Money Lending Business Act, such as Japan Credit Information Reference Center Corp. (JICC) and Credit Information Center (CIC).
  - There appears to be no centralized credit bureau for consumer loans (late payment, bankruptcy and others), with scope to improve efficiency through information sharing.
  - Legal constraints and lack of an identification system (absence of a tax identification number) hinder credit information sharing and likely create a financing gap for the “middle-risk” group (Schaede, 2005).
- Market practices:
  - Heavy use of fixed-asset collateral and personal guarantees.
  - Limited use of asset-based lending and difficulty valuing inventory and intangible assets.
- Institutional investor restrictions:
  - Public pension funds restricted from investing in riskier assets, limiting demand for securitized SME loans and venture capital.

### Policy recommendations (financial sector measures)
- Gradually phase out credit support measures:
  - Phase out special credit guarantees with full coverage of loan value as the recovery takes hold.
  - Reduce the normal guarantee coverage ratio from 80 to 60 percent over time (in line with international averages (IMF, 2006)) and scrutinize the rollover of guarantees to ensure market discipline.
  - Phase out the temporary SME Financing Facilitation Act by March 2013 so banks would not have an obligation (but retain the option) to restructure loans at SMEs’ request, reducing incentives to classify restructured loans as ‘normal.’
- Encourage consolidation of regional and shinkin banks by raising capital requirements:
  - Increase the capital requirement for domestic-oriented banks from 4 percent at present.
  - Example: less than 5 percent of smaller banks would have a capital shortfall if the capital requirement for those banks was raised to 6 percent of risk-weighted assets (still below the minimum requirement of internationally-active banks at 8 percent).
  - Near-term risk: possible contraction in credit as regional banks deleverage; mitigation options include equity issuance, lower dividend payouts, or temporary public capital injection.
  - Medium-term benefit: higher capital would strengthen institutions’ capacity to take on risk.
- Accelerate SME restructuring:
  - Consider refocusing the public asset management company to advance SME restructuring.
  - Use debt-equity swaps to incentivize banks and SMEs for out-of-court voluntary workouts (Laryea, 2010).
- Promote risk-based financing and deepen capital markets:
  - Authorities (government and municipalities) could lead in originating and trading electronically-registered claims to avoid double assignment risks.
  - Scale down and better target current support to SMEs by redirecting GFIs’ lending away from reinsuring SME guarantees toward risk-based lending; studies show startups obtaining GFIs’ financing tend to outperform (Fukanuma, Tadanobu, and Watanabe, 2006).
  - Revise investment restrictions for institutional investors (e.g., pension funds)—once supervision and regulatory framework are strengthened—to encourage alternative investments such as securitized loans and venture capital.
  - Development of capital markets for securitized loans could improve long-term returns and channel funds toward private investment.
- Streamline regulatory measures to reduce time and cost of starting businesses:
  - Simplify business registration to promote business creation and transfers.
  - Broaden coverage of credit registries through a more centralized database and include a consumer data bureau for personal credit information.
  - Facilitate by linking the proposed taxpayer identification system to identification of credit information of firms and individuals.

### Quantified macroeconomic impact (modeling)
- IMF GIMF model simulation:
  - Addressing SME structural weaknesses could raise aggregate productivity by about ¼ percentage points.
  - This increase in aggregate productivity would lift long-term growth by 0.1–0.2 percentage points from the baseline.
  - Modeling assumption: productivity of smaller firms would increase to about 80 percent of that of large firms, producing the ¼ percentage point aggregate productivity gain.

### Conclusion: policy rationale and expected outcomes
- To encourage supply of risk-based capital, costly government support measures should be phased out and SME restructuring accelerated.
- Deepening capital markets and addressing regulatory barriers to starting businesses are necessary to enhance risk capital availability.
- Addressing SME weaknesses is expected to improve private investment, enhance firm productivity, and lift growth.

*Source: IMF working paper content unit: “22. Availability of risk-based capital is very limited despite plenty of liquidity.”*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12291.pdf_