## _cr11289

## Source details

**Canonical URL:** [_cr11289](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2011/_cr11289.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2011/_cr11289.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2011/_cr11289.pdf.json)

---

### Introduction and key observations
- Household debt to personal disposable income rose steadily since the mid-1990s, reaching a record 163 percent in 2010.
- Banks’ exposure to the household sector increased markedly: household loans comprised 35 percent of total loans outstanding in 2010, compared to 18 percent in 1995.
- Mortgage debt is the primary driver of the increase in total household debt:
  - Mortgage debt as a share of household income rose from 73 percent in 1996 to 145 percent in 2010.
  - Other personal debt rose from 15 percent to some 20 percent over the same period.
- The increase in household debt and rising house prices are highly correlated and have feedback effects that fuel further borrowing.

### Macroeconomic context and drivers
- Sweden experienced structural reforms after the 1991–93 crisis, including a shift to a freely-floating exchange rate, a rules-based fiscal framework, and inflation targeting.
- Macroeconomic performance from 1995–2007:
  - Growth averaged close to 3½ percent (1995-2007).
  - European average growth was 2½ percent during the same period.
- Factors attributed to house price increases (RUTH analysis):
  - Disposable income explains roughly ½ of the increase in house prices since 1995.
  - Lower interest rates explain roughly 35 percent.
  - Household financial wealth explains some 10 percent.
  - The residual suggests prices are roughly 20 percent above their long term trend.
- An alternative study (Swedish National Housing Finance Board, February 2011) concluded house prices were overvalued by some 20–25 percent.

### Demand, supply and financial sector factors (Box 1 summary)
- Demand-side:
  - Low interest rates and rising household income lowered user costs and increased borrower demand.
  - Tax regime changes—abolition of property, wealth, inheritance and gift taxes—and the 30 percent mortgage interest deduction eased access to housing.
  - Demographic shift toward metropolitan areas raised pressure on limited housing stock; price appreciation concentrated in Stockholm, Malmo and Gothenburg.
- Supply-side:
  - Structural under-supply of housing due to rent regulation in main cities, land scarcity and onerous procedures for developers; housing investment remained low.
- Financial sector:
  - Financial innovations expanded credit access (e.g., “interest only” or “amortization-free” loans).
  - Introduction of covered bonds legislation in 2004 transformed the covered bonds market and bolstered mortgage lending.

### Loan-to-value (LTV) developments and borrower profiles
- Average LTV ratio rose to 55 percent at end-2010 from 10 percent in 1995.
- Prior to the 85 percent LTV cap imposed in October 2010:
  - Average LTV on new lending had risen to 75 percent for dwellings (FI’s 2009 mortgage survey).
  - 12 percent of new borrowers had an LTV above 90 percent.
  - One third of new borrowers had an LTV above 85 percent.
- Bank practices:
  - Banks’ own rules varied LTV ratios between 75 and 95 per cent of market value (or estimated market value).
  - Most banks offered a first mortgage limited to 75–85 per cent of market value and a second mortgage for the amount exceeding this limit.
- Younger buyers:
  - LTV ratios increasingly concentrated among younger buyers; their LTV rose to 70 percent in 2009.
  - First-time homebuyers account for a very small percentage of homeowners.
  - More than 95 percent of households who own their own homes have done so for at least two years and have seen a drop in their LTV ratio.

### Recent crisis period dynamics and policy responses
- The 2008 global financial crisis produced only a brief slowdown in the increase in household debt and house prices; house prices did not adjust downwards in Sweden during the crisis.
- Policy responses helped sustain the housing market:
  - Monetary policy: expansionary stance and later Riksbank tightening beginning July 2010.
  - Fiscal policy: fiscal stimulus supported employment and household incomes; a tax deduction introduced end-2008 for housing renovation and rebuilding (raised to SEK 50,000 (EUR 5,000) per house owner and year) supported mortgage lending and the construction sector.
  - Financial sector measures: Riksbank expanded its collateral policy by accepting covered bonds as collateral; the facility remains active and helped restore covered bonds market functioning and mortgage financing.
- An LTV cap of 85 percent was imposed in October 2010.

### Household balance sheets and wealth buffers
- Aggregate wealth and leverage:
  - Household net wealth buffer exceeds 200 percent of disposable income.
  - Households’ non-equity financial buffer was roughly 75 percent of disposable income at end-2010.
- Composition risks:
  - More than ½ of households’ financial wealth is in equities, increasing sensitivity to equity market declines.
  - Real assets account for less than ¼ of total household net wealth in Sweden (2007 data).
- Net debt trends:
  - Net debt has been rising since the mid-1990s as gross debt grew faster than liquid assets, but net debt declined during the recent financial crisis as households shifted to safer, more liquid assets.

### Interest rate sensitivity and mortgage structure
- Variable-interest rate mortgages became predominant:
  - By the end of the 1990s, share of loans with variable interest rate rose to roughly 40 percent of new mortgage loans and remained constant until 2007.
  - In 2008, the share of variable-rate loans on new lending increased to over 60 percent.
  - Variable rate loans peaked at 62 percent of the total outstanding stock in early 2010.
  - Fixed rates (1–5 years) accounted for 28 percent of the outstanding stock.
  - Fixed rates (>5 years) accounted for 10 percent of the outstanding stock.
- Increased exposure to variable rates raises household vulnerability to a rapid increase in interest rates.

### Assessment and implications for financial stability (summarized conclusions)
- Sweden exhibits several Sweden-specific institutional and structural features that provide safety buffers and lower direct vulnerability of the financial sector despite household debt trends.
- Key risks that could heighten banking sector credit risks:
  - Negative shocks to household income and rising unemployment.
  - A rapid increase in interest rates.
  - A decline in real estate prices.
- Aggregate buffers (large net wealth, continued mortgage market functioning through covered bonds) mitigate immediate aggregate insolvency risk, but distributional aspects—higher LTVs among recent and younger borrowers and large equity exposures—pose potential channels for stress and feedback effects between real and financial sectors.

### Interest rates, affordability, and recent trends
- Long term decline in interest rates has reduced household interest cover or debt service ratios despite increases in household debt relative to income and assets (Figure 9).
- Increase in household indebtedness has been offset by the decline in borrowing rates so that on average households have not devoted a greater share of income to debt service than in the past.
- With interest rates at historical lows and set to rise further as the Riksbank withdraws monetary stimulus, debt service will rise as mortgage rates increase with the policy rate. Interest ratios have been on the rise since mid-2010, when the current monetary tightening cycle began.
- Swedbank’s affordability index: 100 percent when household uses 15 percent of its income before taxes (roughly 25 percent of disposable income) to service debt (interest and amortization). Since 2008 the index fell below 100, reaching 96 in 2010Q4.

### Distributional heterogeneity and S-curve evidence
- Estimates of the S-curve, derived from household survey data, confirm aggregate trends:
  - Household interest ratios follow official policy rates closely: interest ratio declined for all households from 1994–2008, but rose from 2006–08 with monetary tightening.
  - Younger borrowers are more sensitive to interest rate changes than the average household: when interest rates rose in 2008, younger borrowers paid more interest as a share of income compared to the average household (consistent with lower incomes, early-stage wealth formation, and higher LTV ratios among younger borrowers).
  - Vulnerable borrowers (prone to illness and unemployment) are similarly more sensitive to interest rate fluctuations and experienced a sharper decline in interest payments from 1994–2008; they likely have fewer assets and higher LTV ratios than the average household.
  - Older borrowers are less sensitive to interest rate changes, pay lower interest than both the average household and younger borrowers, and experienced the smallest change in their interest ratio from 1994–2008.
- The bulk of the debt stock is held by households with high net worth; the majority of housing debt is held by high income households who also hold the bulk of real and financial assets.

### Implications from stress tests and credit-risk assessment
- Overall household balance sheets remain in good shape: substantial rise in real and financial assets has offset the increase in debt; most debt is held by households that can afford to service it.
- Riksbank stress tests on micro survey data find low credit risks from household indebtedness:
  - Low income households present highest financial risks but represent roughly 6 percent of total indebted households in the survey and hold about 6 percent of total household debts.
  - Projected loan losses associated with a default by vulnerable households amount to 0.9 percent.
- A one-time mortgage survey carried out in 2009 (covering banks representing 90 percent of Swedish mortgage markets) found negligible loan losses in three single-factor shock scenarios:
  - Interest rate shock scenario: credit losses rise to 1.9 percent.
  - Household income shock scenario: credit losses rise to 1.3 percent.
  - Unemployment shock scenario: credit losses rise to 0.8 percent.
- FSAP Update banking stress tests confirm households pose the lowest credit risks relative to aggregate household, financial and non-financial corporate sector exposures, conditional on conservative loss-given-default (LGD) assumptions based on historically low mortgage losses in Sweden.

### Structural and institutional mitigants to mortgage losses
- Swedish-specific factors that historically limited household-lending losses:
  - Highly regulated covered bonds market provides incentives for selective mortgage origination.
  - Direct and life-long personal liability: by law a borrower is personally liable for life even after default and foreclosure procedures have been initiated.
  - Generous social security: a well-developed social welfare system (including unemployment insurance for up to three years and support for the long-term unemployed) cushions households’ ability to service debt during unemployment.
  - No buy-to-let market: a highly regulated rental market and tenant owner subletting restrictions prevent development of a speculative buy-to-let market.
- Covered Bonds market specifics and safeguards:
  - Covered Bond Act (effective 1 July 2004) introduced double recourse and contributed to covered bonds rising from 40 percent of GDP to over 160 percent of GDP.
  - Issuers must comply with property valuation principles, present value calculations, stress tests for interest rate risk (symmetric yield-curve shifts of 100bps), currency risk hedging (present value of covered assets must exceed covered bonds by an amount equivalent to a 10 percent shift in the currency), and minimum swap counterpart ratings.
  - LTV requirements for collateral: 75 percent for residential property, 70 percent for agricultural property, and 60 percent for commercial real estate.
  - Mortgage loans in arrears for more than 60 days do not count as collateral for Swedish covered bonds.
  - The amount of covered bond issuance as share of total mortgage lending doubled to 80 percent since 2006; covered bond funding costs are cheaper than alternatives, contributing to lower mortgage rates.
  - Cover pools are dynamic and frequently monitored by FI through an independently appointed inspector; issuers must ensure cover pool always exceeds nominal and net present value claims of covered bond holders.

### Funding and systemic vulnerability
- Despite lower direct credit risk from households, growth in covered bond financing raises banks’ exposure to funding risk:
  - A housing market correction could reduce investor appetite for covered bonds, raising refinancing risks for banks and increasing mortgage interest rates.
  - The covered bonds market froze during late-2008 to early-2009; policy responses included the Riksbank accepting a wider range of covered bonds as collateral, NDO purchasing covered bonds in exchange for treasury securities, and a government bank debt guarantee program.
  - Covered bonds’ asset-liability matching lessens liquidity risk for issuers but shifts liquidity/refinancing risk to final mortgage borrowers; banks remain indirectly exposed to market funding shocks.

### Conclusions and financial stability assessment
- House prices and household indebtedness have risen to record levels but do not appear to present significant credit risks to the financial sector given:
  - Robust household balance sheets and concentration of debt within high net worth segments.
  - The Swedish Covered Bonds Act (2004) helped structure a high-quality covered bonds market, supporting mortgage lending growth and mortgage loan quality.
- Swedish policy and institutional features—full recourse household debt, generous social benefits, and the public sector’s implicit backing—reduce household default incentives and support capacity to pay, shifting a portion of financial risks toward the public sector.
- Nonetheless, a disorderly correction of house prices could destabilize the financial system through the funding channel due to potential retrenchment in covered bond investor demand and resultant refinancing risks.

*Source: _cr11289 - 1. Household Debt as a Share of Disposable Income; 18. The structural long term decline in interest rates has been key to affordability*

### 1. Household Debt as a Share of Disposable Income ................................................................5

### 1. Household Debt as a Share of Disposable Income

### Introduction and key observations
- Household debt to personal disposable income rose steadily since the mid-1990s, reaching a record 163 percent in 2010.
- Banks’ exposure to the household sector increased markedly: household loans comprised 35 percent of total loans outstanding in 2010, compared to 18 percent in 1995.
- Mortgage debt is the primary driver of the increase in total household debt:
  - Mortgage debt as a share of household income rose from 73 percent in 1996 to 145 percent in 2010.
  - Other personal debt rose from 15 percent to some 20 percent over the same period.
- The increase in household debt and rising house prices are highly correlated and have feedback effects that fuel further borrowing.

### Macroeconomic context and drivers
- Sweden experienced structural reforms after the 1991–93 crisis, including a shift to a freely-floating exchange rate, a rules-based fiscal framework, and inflation targeting.
- Macroeconomic performance from 1995–2007:
  - Growth averaged close to 3½ percent (1995-2007).
  - European average growth was 2½ percent during the same period.
- Factors attributed to house price increases (RUTH analysis):
  - Disposable income explains roughly ½ of the increase in house prices since 1995.
  - Lower interest rates explain roughly 35 percent.
  - Household financial wealth explains some 10 percent.
  - The residual suggests prices are roughly 20 percent above their long term trend.
- An alternative study (Swedish National Housing Finance Board, February 2011) concluded house prices were overvalued by some 20–25 percent.

### Demand, supply and financial sector factors (Box 1 summary)
- Demand-side:
  - Low interest rates and rising household income lowered user costs and increased borrower demand.
  - Tax regime changes—abolition of property, wealth, inheritance and gift taxes—and the 30 percent mortgage interest deduction eased access to housing.
  - Demographic shift toward metropolitan areas raised pressure on limited housing stock; price appreciation concentrated in Stockholm, Malmo and Gothenburg.
- Supply-side:
  - Structural under-supply of housing due to rent regulation in main cities, land scarcity and onerous procedures for developers; housing investment remained low.
- Financial sector:
  - Financial innovations expanded credit access (e.g., “interest only” or “amortization-free” loans).
  - Introduction of covered bonds legislation in 2004 transformed the covered bonds market and bolstered mortgage lending.

### Loan-to-value (LTV) developments and borrower profiles
- Average LTV ratio rose to 55 percent at end-2010 from 10 percent in 1995.
- Prior to the 85 percent LTV cap imposed in October 2010:
  - Average LTV on new lending had risen to 75 percent for dwellings (FI’s 2009 mortgage survey).
  - 12 percent of new borrowers had an LTV above 90 percent.
  - One third of new borrowers had an LTV above 85 percent.
- Bank practices:
  - Banks’ own rules varied LTV ratios between 75 and 95 per cent of market value (or estimated market value).
  - Most banks offered a first mortgage limited to 75–85 per cent of market value and a second mortgage for the amount exceeding this limit.
- Younger buyers:
  - LTV ratios increasingly concentrated among younger buyers; their LTV rose to 70 percent in 2009.
  - First-time homebuyers account for a very small percentage of homeowners.
  - More than 95 percent of households who own their own homes have done so for at least two years and have seen a drop in their LTV ratio.

### Recent crisis period dynamics and policy responses
- The 2008 global financial crisis produced only a brief slowdown in the increase in household debt and house prices; house prices did not adjust downwards in Sweden during the crisis.
- Policy responses helped sustain the housing market:
  - Monetary policy: expansionary stance and later Riksbank tightening beginning July 2010.
  - Fiscal policy: fiscal stimulus supported employment and household incomes; a tax deduction introduced end-2008 for housing renovation and rebuilding (raised to SEK 50,000 (EUR 5,000) per house owner and year) supported mortgage lending and the construction sector.
  - Financial sector measures: Riksbank expanded its collateral policy by accepting covered bonds as collateral; the facility remains active and helped restore covered bonds market functioning and mortgage financing.
- An LTV cap of 85 percent was imposed in October 2010.

### Household balance sheets and wealth buffers
- Aggregate wealth and leverage:
  - Household net wealth buffer exceeds 200 percent of disposable income.
  - Households’ non-equity financial buffer was roughly 75 percent of disposable income at end-2010.
- Composition risks:
  - More than ½ of households’ financial wealth is in equities, increasing sensitivity to equity market declines.
  - Real assets account for less than ¼ of total household net wealth in Sweden (2007 data).
- Net debt trends:
  - Net debt has been rising since the mid-1990s as gross debt grew faster than liquid assets, but net debt declined during the recent financial crisis as households shifted to safer, more liquid assets.

### Interest rate sensitivity and mortgage structure
- Variable-interest rate mortgages became predominant:
  - By the end of the 1990s, share of loans with variable interest rate rose to roughly 40 percent of new mortgage loans and remained constant until 2007.
  - In 2008, the share of variable-rate loans on new lending increased to over 60 percent.
  - Variable rate loans peaked at 62 percent of the total outstanding stock in early 2010.
  - Fixed rates (1–5 years) accounted for 28 percent of the outstanding stock.
  - Fixed rates (>5 years) accounted for 10 percent of the outstanding stock.
- Increased exposure to variable rates raises household vulnerability to a rapid increase in interest rates.

### Assessment and implications for financial stability (summarized conclusions)
- Sweden exhibits several Sweden-specific institutional and structural features that provide safety buffers and lower direct vulnerability of the financial sector despite household debt trends.
- Key risks that could heighten banking sector credit risks:
  - Negative shocks to household income and rising unemployment.
  - A rapid increase in interest rates.
  - A decline in real estate prices.
- Aggregate buffers (large net wealth, continued mortgage market functioning through covered bonds) mitigate immediate aggregate insolvency risk, but distributional aspects—higher LTVs among recent and younger borrowers and large equity exposures—pose potential channels for stress and feedback effects between real and financial sectors.

*Source: _cr11289 - 1. Household Debt as a Share of Disposable Income*

### 18.      The structural long term decline in interest rates has been key to affordability

### 18.      The structural long term decline in interest rates has been key to affordability

### Interest rates, affordability, and recent trends
- Long term decline in interest rates has reduced household interest cover or debt service ratios despite increases in household debt relative to income and assets (Figure 9).  
- Increase in household indebtedness has been offset by the decline in borrowing rates so that on average households have not devoted a greater share of income to debt service than in the past.  
- With interest rates at historical lows and set to rise further as the Riksbank withdraws monetary stimulus, debt service will rise as mortgage rates increase with the policy rate. Interest ratios have been on the rise since mid-2010, when the current monetary tightening cycle began.  
- Swedbank’s affordability index: 100 percent when household uses 15 percent of its income before taxes (roughly 25 percent of disposable income) to service debt (interest and amortization). Since 2008 the index fell below 100, reaching 96 in 2010Q4.

### Distributional heterogeneity and S-curve evidence
- Estimates of the S-curve, derived from household survey data, confirm aggregate trends:
  - Household interest ratios follow official policy rates closely: interest ratio declined for all households from 1994–2008, but rose from 2006–08 with monetary tightening.
  - Younger borrowers are more sensitive to interest rate changes than the average household: when interest rates rose in 2008, younger borrowers paid more interest as a share of income compared to the average household (consistent with lower incomes, early-stage wealth formation, and higher LTV ratios among younger borrowers).
  - Vulnerable borrowers (prone to illness and unemployment) are similarly more sensitive to interest rate fluctuations and experienced a sharper decline in interest payments from 1994–2008; they likely have fewer assets and higher LTV ratios than the average household.
  - Older borrowers are less sensitive to interest rate changes, pay lower interest than both the average household and younger borrowers, and experienced the smallest change in their interest ratio from 1994–2008.
- The bulk of the debt stock is held by households with high net worth; the majority of housing debt is held by high income households who also hold the bulk of real and financial assets.

### Implications from stress tests and credit-risk assessment
- Overall household balance sheets remain in good shape: substantial rise in real and financial assets has offset the increase in debt; most debt is held by households that can afford to service it.
- Riksbank stress tests on micro survey data find low credit risks from household indebtedness:
  - Low income households present highest financial risks but represent roughly 6 percent of total indebted households in the survey and hold about 6 percent of total household debts.
  - Projected loan losses associated with a default by vulnerable households amount to 0.9 percent.
- A one-time mortgage survey carried out in 2009 (covering banks representing 90 percent of Swedish mortgage markets) found negligible loan losses in three single-factor shock scenarios:
  - Interest rate shock scenario: credit losses rise to 1.9 percent.
  - Household income shock scenario: credit losses rise to 1.3 percent.
  - Unemployment shock scenario: credit losses rise to 0.8 percent.
- FSAP Update banking stress tests confirm households pose the lowest credit risks relative to aggregate household, financial and non-financial corporate sector exposures, conditional on conservative loss-given-default (LGD) assumptions based on historically low mortgage losses in Sweden.

### Structural and institutional mitigants to mortgage losses
- Swedish-specific factors that historically limited household-lending losses:
  - Highly regulated covered bonds market provides incentives for selective mortgage origination.
  - Direct and life-long personal liability: by law a borrower is personally liable for life even after default and foreclosure procedures have been initiated.
  - Generous social security: a well-developed social welfare system (including unemployment insurance for up to three years and support for the long-term unemployed) cushions households’ ability to service debt during unemployment.
  - No buy-to-let market: a highly regulated rental market and tenant owner subletting restrictions prevent development of a speculative buy-to-let market.
- Covered Bonds market specifics and safeguards:
  - Covered Bond Act (effective 1 July 2004) introduced double recourse and contributed to covered bonds rising from 40 percent of GDP to over 160 percent of GDP.
  - Issuers must comply with property valuation principles, present value calculations, stress tests for interest rate risk (symmetric yield-curve shifts of 100bps), currency risk hedging (present value of covered assets must exceed covered bonds by an amount equivalent to a 10 percent shift in the currency), and minimum swap counterpart ratings.
  - LTV requirements for collateral: 75 percent for residential property, 70 percent for agricultural property, and 60 percent for commercial real estate.
  - Mortgage loans in arrears for more than 60 days do not count as collateral for Swedish covered bonds.
  - The amount of covered bond issuance as share of total mortgage lending doubled to 80 percent since 2006; covered bond funding costs are cheaper than alternatives, contributing to lower mortgage rates.
  - Cover pools are dynamic and frequently monitored by FI through an independently appointed inspector; issuers must ensure cover pool always exceeds nominal and net present value claims of covered bond holders.

### Funding and systemic vulnerability
- Despite lower direct credit risk from households, growth in covered bond financing raises banks’ exposure to funding risk:
  - A housing market correction could reduce investor appetite for covered bonds, raising refinancing risks for banks and increasing mortgage interest rates.
  - The covered bonds market froze during late-2008 to early-2009; policy responses included the Riksbank accepting a wider range of covered bonds as collateral, NDO purchasing covered bonds in exchange for treasury securities, and a government bank debt guarantee program.
  - Covered bonds’ asset-liability matching lessens liquidity risk for issuers but shifts liquidity/refinancing risk to final mortgage borrowers; banks remain indirectly exposed to market funding shocks.

### Conclusions and financial stability assessment
- House prices and household indebtedness have risen to record levels but do not appear to present significant credit risks to the financial sector given:
  - Robust household balance sheets and concentration of debt within high net worth segments.
  - The Swedish Covered Bonds Act (2004) helped structure a high-quality covered bonds market, supporting mortgage lending growth and mortgage loan quality.
- Swedish policy and institutional features—full recourse household debt, generous social benefits, and the public sector’s implicit backing—reduce household default incentives and support capacity to pay, shifting a portion of financial risks toward the public sector.
- Nonetheless, a disorderly correction of house prices could destabilize the financial system through the funding channel due to potential retrenchment in covered bond investor demand and resultant refinancing risks.

*Source: _cr11289 - 18.      The structural long term decline in interest rates has been key to affordability*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2011/_cr11289.pdf_