## _sdn1512 - Executive Summary

## Source details

**Canonical URL:** [_sdn1512 - Executive Summary](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1512.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1512.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1512.pdf.json)

---

### Key conclusions and policy messages
- Housing finance characteristics vary widely across countries; LTV, reliance on wholesale funding, and, to some extent, longer loan maturities are positively correlated with mortgage-market depth and homeownership, alongside institutional quality and macro stability (associated benefits: higher school attainment, higher social capital, lower crime).
- Some characteristics associated with deeper mortgage markets increase crisis risk: higher LTVs associate with excessively rapid house-price and credit growth during booms; wholesale funding associates with worse outcomes after housing booms.
- Advanced and emerging markets should avoid relaxing housing-financing standards to deepen mortgage markets; priority should be on improving institutions (for example, legal rights) and the macroeconomic environment.
- Macroprudential policies, particularly housing finance regulation, should be first line of defense for mortgage market booms due to their narrow focus; their effectiveness beyond the short run remains unproven.
- Monetary policy can be a necessary complement to macroprudential measures: about 60 percent of identified past real-estate booms occurred as a result of, or at the same time as, rapid economic growth and broad high credit growth.
- A broad policy mix is required: macroprudential and monetary policies, fiscal incentives, and house supply considerations are structural, country-specific elements that affect boom probability and bust costs.

### Mortgage market heterogeneity and empirical regularities
- Median share of mortgages in household credit (2011): about 70 percent; only six countries below a 40 percent share.
- Mortgage-to-GDP ratios (2001–05 averages) range from below 1 percent in Russia and Turkey to about 80 percent in the Netherlands, New Zealand, and Switzerland.
- In most countries, mortgage-to-GDP ratios as of 2011 remained above their 2001–05 average; cross-country ranking and dispersion did not significantly change recently.
- Univariate regressions: differences in GDP per capita and the credit-to-GDP ratio explain more than 50 and more than 60 percent, respectively, of variation in mortgage-to-GDP ratio.
- Institutional factors (strength of legal rights, ease of registering property) and per-capita GDP are statistically significantly correlated with mortgage market size.
- Adding house-finance characteristics (LTV, term-to-maturity, funding model) to institutional and GDP per capita improves model fit by about 10 percent; institutional variables and GDP per capita remain statistically significant.

### Cross-country housing finance characteristics (sample summary)
- Maximum observed LTV:
  - Range: 70 percent (Colombia, Hong Kong, Hungary) to 125 percent (the Netherlands).
  - Median maximum observed LTV: 83 percent.
  - Most countries fall in 70–80 and 90–100 LTV buckets.
- Term to maturity:
  - Range: 7 years (Turkey) to 45 years (Sweden).
  - Median: 25 years.
- Interest type:
  - Variable-rate standard in 30 countries, fixed in 12 countries, both in 14 countries.
  - Variable rates more common in emerging economies.
- Funding model:
  - In 44 countries, banks funded mostly by retail deposits play a major role in mortgage lending.
  - Nonbanks and wholesale/cross-border funding important in about 22 countries (including the United States, Sweden, Denmark, Australia, South Africa).
- Degree of lender recourse:
  - Full recourse on mortgages in about 44 of the 53 countries.
- Mortgage interest tax deductibility:
  - Allowed in 33 countries; more common in advanced economies (~two-thirds) than emerging markets (~half).
  - Deductions often capped; United States and Norway allow nearly full deductibility without taxing imputed rents.

### Benefits and costs of deeper mortgage markets and homeownership
- Lower down-payment requirements favor homeownership: a decrease in down payment from 40 to 20 percent associated with a 5 to 8 percentage point increase in share of young households (age 26–35) living in owner-occupied houses.
- Homeownership associations:
  - Positive: promotes community investment, higher school attainment, lower crime, higher life satisfaction.
  - Economic: encourages saving and investment for younger households; mortgages can act as committed savings (though equity extraction weakens this).
  - Housing as investment: U.S. evidence shows housing risk-return compares favorably with stock market, with regional variation.
  - Negative: higher homeownership associated with reduced labor mobility; benefits may be smaller in non-U.S. contexts (example countries: Germany, Switzerland).
- Policy implication: policies increasing or subsidizing homeownership should be based on country-specific cost-benefit analyses.

### Housing finance, house-price booms, and systemic risks
- Mortgages are systemically important because real-estate lending often accounts for large shares of household credit and financial sector activity; mortgages introduce high household leverage and affect multiple sectors; major mortgage lenders are leveraged commercial banks.
- Real-estate booms financed through fast credit growth pose macroeconomic risks: booms often end in busts causing household debt overhang, tighter lending standards, financial distress, recessions, and large increases in public debt.
- Data choices:
  - Long time series on mortgage credit are lacking; household credit used as approximation because median mortgage-to-household-credit ratio is about 70 percent and correlation between mortgage and household credit ranges from 0.46 to 0.99 with median 0.99 (only six countries with correlation below 0.9).
- Boom identification:
  - Booms defined as large and persistent deviations of real (year-over-year) growth of house prices and credit from country-specific historical norms at quarterly frequency.

### Definitions, sample, and occurrence of booms
- Boom conditions require both:
  - (i) real growth rate of credit (house prices) > 10 (5) percent, or > two standard deviations of country-specific distribution in a quarter; and
  - (ii) real growth rate of credit (house prices) > 10 (5) percent or > one standard deviation of the country-specific distribution for at least two years.
- Sample: 53 countries with quarterly house-price, household-credit, and corporate-credit data; coverage as far back as 1970s up to 2012.
- Detected booms (1970–2012):
  - 83 household-credit booms
  - 68 corporate-credit booms
  - 67 private-credit booms
- Most episodes (about 60 to 65 percent) occur in advanced countries; after controlling for sample composition, emerging markets are in a boom state roughly double the portion of advanced economies.
- Household-credit booms more frequent than other credit-boom types.
- Higher occurrence of credit booms after 1980s linked to deregulation and financial innovation (securitization).
- Booms cluster across countries, suggesting global factors (e.g., Federal Funds rate and VIX) play roles.
- Identified house-price booms: 85 episodes; peak relative occurrence just before the recent global financial crisis (in 2005, booms in more than half the sampled countries).
- By 2012:Q4, less than 10 percent of countries were experiencing a credit boom (Brazil, Hong Kong, India, Israel, Malaysia, Norway).
- Regression evidence: a 1 percent increase in household credit associated with about 0.2 percent increase in house prices in the following year.

### Interaction between real-estate booms and credit booms; predictability
- Presence of a household-credit boom raises probability of real-estate boom to 57 percent from unconditional 29 percent.
- Household-credit booms are better predictors of house-price booms than private-credit booms.
- Household debt-to-GDP level negatively correlated with occurrence of real-estate booms.
- Lagged GDP growth positively associated with probability of a real-estate boom.
- Higher maximum observed LTV increases probability of real-estate boom; LTVs can be a targeted tool to limit real-estate price fluctuations.

### Classification and statistics of real-estate booms
- Of 85 real-estate booms:
  - 49 coincide with private-credit booms (twin household and corporate)
  - 16 accompany household-credit booms only
  - 2 associate with corporate-credit booms only (Hong Kong 2004:Q1–05:Q4; Singapore 2006:Q3–08:Q2)
  - 18 occur without any type of credit booms
- Duration and magnitude:
  - Average duration: with private-credit booms ≈ 18 quarters; other cases ≈ 14 quarters.
  - Average real increase in house prices during booms:
    - With private-credit boom: 14 percent
    - With household-credit only: 9 percent
    - Without any credit boom: 10 percent
  - Average real household-credit growth during house-price booms:
    - With private-credit booms: about 21 percent
    - With household-credit only: 11 percent
    - Without any credit booms: 5 percent
- Note: In housing booms, private-credit booms almost always reflect simultaneous household and corporate credit booms; four exceptions noted (Canada 1973:Q1–76:Q1; Czech Republic 2006:Q3–08:Q4; Denmark 2004:Q1–07:Q1; Greece 2005–07:Q1).

### Macroeconomic performance during real-estate booms
- Real GDP growth during housing booms higher than during non-boom periods by about 1¼ to 2 ½ percent; differences statistically significant.
- Housing booms coinciding with private-credit booms register higher real-GDP growth than booms with household-credit only or without credit booms.
- Consumption and investment growth higher during house-price booms with private-credit booms.
- Housing booms accompanied by credit booms associated with exchange-rate appreciation and current-account deterioration.
- Inflation typically subdued and similar to tranquil times during real-estate booms.

### Aftermath: declines, recessions, and "bad" booms
- House prices generally decline after real-estate booms; declines largest for booms with private-credit booms.
- Little evidence sharp declines are systematically followed by rebounds.
- Of house-price booms with sufficient post-period data, 49 out of 78 ended up in recessions.
- Recession definition: end of a housing boom followed by two or more consecutive quarters of negative real GDP growth within three years after boom end.
- About two-thirds of booms end badly; proportion similar in advanced economies and emerging markets.
- More booms occurred since 2000, and a larger proportion ended in recessions; 2007–09 global financial crisis influential.
- No clear relationship between boom duration and probability of ending badly.
- Booms accompanied by private-credit booms more likely to end in recessions than those with only household-credit booms or without credit booms.
- Funding model effects:
  - Countries financing housing credit primarily through bank retail deposits have lower probability of booms ending in recessions.
  - Wholesale funding increases leverage and vulnerability.
- About one in five housing booms (about 13 cases) followed by a systemic banking crisis within three years of boom end; no clear link between house-price boom characteristics and systemic banking crisis in this sample.

### Policy implications and recommendations
- Cross-country differences in mortgage-market depth linked to institutional elements (collateral and bankruptcy laws, ease of registering property), macroeconomic variables (GDP per capita), and housing finance characteristics — indicating room for policy action.
- Improve legal frameworks and macroeconomic stability to enhance access to housing finance.
- Deeper and more inclusive mortgage markets correlate with higher homeownership and social benefits (higher school attainment, higher social capital, lower crime, higher life satisfaction).
- Some features that increase access and affordability (higher LTVs, longer maturities) may promote rapid credit growth and raise financial-stability risks.
- House-price booms financed through wholesale markets are more likely to end badly (in recessions).
- Macroprudential policies can target household leverage, indebtedness, and originator/investor risk profiles; useful for containing systemic vulnerabilities.
- Monetary policy role:
  - About 60 percent of real-estate booms occur together with private-credit booms, often alongside rapid, broad economic growth — circumstances that may call for monetary tightening after weighing benefits and risks to financial stability.
  - Monetary tightening can weaken household and firm financial conditions short run but reduce leverage and strengthen stability medium term.
  - Absence of inflation pressure during many real-estate booms suggests careful consideration of including real-estate prices in monetary policy reaction functions.
- Broader measures:
  - Well-paced, country-specific supply-side measures to mitigate demand shocks long run.
  - Avoid abrupt supply-side changes at boom peaks or bust starts as they may exacerbate price corrections.
  - Address distortions linked to special treatment of housing and homeownership.
- Targeted LTV limits:
  - Max observed LTV is a significant predictor of boom probability; reducing LTVs through macroprudential limits could be effective, conditional on enforcement and avoidance of regulatory arbitrage.

### Regression and probit model key statistics (selected exact figures)
- Annex II (mortgage market depth regressions):
  - Legal rights index: 4.175*** ( 0.003)
  - Log of GDP per capita: 12.62*** (0.000) with other columns reporting 12.53***, 12.96***, 14.06***, 12.01***, 12.19***, 13.25***, 14.47***, 14.38*** (all with (0.000))
  - Max observed LTV: 0.319* ( 0.083); alternate specifications: 0.296 ( 0.115); 0.426** ( 0.034)
  - Term to maturity: 0.634* (0.056); other columns 0.457 (0.180); 0.210 (0.554)
  - Retail funding: -9.816* ( 0.073); -10.48* ( 0.053); -10.74** ( 0.050)
  - R-sq range across columns: 0.291 to 0.696; Observations: 53 (most columns), 52 in column (10)
  - Note: Column (10) reports robust regression; ***, **, * indicate significance at 1, 5, 10 percent.
- Annex IV (probit predicting housing booms):
  - HH Credit Boom (lag): Column (1) 0.784*** (0.237); Column (2) 0.734*** (0.241); Column (3) 0.753*** (0.240); Column (4) 0.758*** (0.236); Column (5) 0.571*** (0.202)
  - Private Credit Boom (lag): Column (1) 0.310 (0.224); Column (2) 0.516** (0.225); Column (4) 0.494** (0.233); Column (5) 0.397** (0.189)
  - HH Indebtedness (lag): Column (1) -0.0331*** (0.00947); Column (2) -0.0439*** (0.0109); Column (3) -0.0288*** (0.00953); Column (4) -0.0363*** (0.0114); Column (5) -0.0137*** (0.00397)
  - Max Observed LTV (Column 5): 0.0214*** (0.00568)
  - US Fed Fund Rate (Column 1): -0.0358* (0.0204)
  - VIX Index (Column 2): -0.0210*** (0.00648)
  - Current Account (lag) Column (1): 0.0578*** (0.0161)
  - GDP growth (lag) Column (1): 0.0723*** (0.0163)
  - CPI Inflation (lag) Column (1): -0.0754*** (0.0228)
  - Estimation samples and fit:
    - Country Fixed Effects: YES for Columns (1)-(4); NO for Column (5)
    - Observations: Column (1) 440; Column (2) 630; Column (3) 814; Column (4) 2183; Column (5) 213315 (as printed)
    - R-sq: Column (1) 0.191; Column (2) 0.252; Column (3) 0.254; Column (4) 0.289; Column (5) 0.172
  - Significance notation: *** 1 percent, ** 5 percent, * 10 percent.

### Robustness of boom identification
- Identified episodes robust across alternative methods (backward-looking cubic trend, Hodrick-Prescott filter, different thresholds, minimum durations).
- Incidence of post-bust recessions varies in a relatively close range from 59 percent to 67 percent across methodologies.
- Table A3.2 exact figures:
  - Baseline (8 quarters): Total booms 85; Followed by recession within three years of boom end 64.1%
  - Topthird (6 quarters): Total booms 104; Followed by recession 64.2%
  - Topquarter (8 quarters): Total booms 70; Followed by recession 59.1%
  - Cubic trend (8 quarters): Total booms 87; Followed by recession 67.1%
  - HP trend (8 quarters): Total booms 89; Followed by recession 66.7%
- Correlations across methodologies generally high (examples of simple correlations, Table A3.1):
  - House price: Baseline vs Topthird6 = 0.939; Baseline vs Topquarter = 0.802; Baseline vs Cubic trend = 0.918; Baseline vs HP trend = 0.907
  - Household credit: Baseline vs Topthird6 = 0.966; Baseline vs Topquarter = 0.898; Baseline vs Cubic trend = 0.842; Baseline vs HP trend = 0.819
  - Private credit: Baseline vs Topthird6 = 0.936; Baseline vs Topquarter = 0.955; Baseline vs Cubic trend = 0.834; Baseline vs HP trend = 0.773

*Source: _sdn1512 - Executive Summary (IMF).*

### Executive Summary ......................................................................................................

### _sdn1512 - Executive Summary

### Major sections
- Executive Summary (page 4)
- I. Introduction (page 5)
- II. Mortgage Markets around the World (page 6)
  - A. Benefits of Deep Mortgage Markets
  - B. Factors Associated with Cross-Country Differences in Mortgage Markets
- III. Housing Finance and Real-estate Booms (page 12)
  - A. Defining and Identifying Credit Booms and Real-estate booms
  - B. Interaction between Real-estate booms and Credit Booms
- IV. Policy Implications (page 21)
- References (page 23)

### Key tables (listed)
- Table 1. Characteristics of House-price Booms (page 16)
- Table 2. Macroeconomic   Performance   during House-price Booms (page 18)

### Key figures (listed)
- Figure 1. Share of Mortgages to HH Credit and HH Credit to Total Credits, as on 2011
- Figure 2. The Cross-Section of Outstanding Mortgage Debt/GDP, 2001-2005 Average
- Figure 3. Development Levels and Mortgage Debt/GDP, 2001-05
- Figure 4. Mortgages and Homeownership Across US States
- Figure 5. Cross-Country Differences in LTV and Maturities of Mortgages
- Figure 6. Cross-Country Differences in Mortgage Interest Type and Funding Models
- Figure 7. Occurrence of Credit Booms during 1970-2012
- Figure 8. Occurrence of House-price booms and Credit Booms
- Figure 9. Change in House Prices and HH Credit during House-price Booms
- Figure 10. Average Growth of Real GDP during House-price Booms
- Figure 11. Change in House Prices after House-price Booms
- Figure 12. Lowest Annual Change in Read GDP after House-price Booms
- Figure 13. Bad and Good Booms

### Content scope and focus
- Coverage spans mortgage market characteristics, benefits of deep mortgage markets, cross-country differences in mortgage contract features (LTV, maturities, interest types, funding models), definitions and identification of credit and real-estate booms, interactions between credit booms and house-price booms, macroeconomic performance during booms, and policy implications.
- Empirical materials include cross-country and cross-state comparisons, time-series occurrence of booms (1970-2012), and measures of house-price and household credit dynamics during and after booms.
- Policy implications are addressed in a dedicated section (IV) focused on lessons from housing finance and real-estate boom analysis.

*Document: _sdn1512 - Executive Summary*

### EXECUTIVE SUMMARY

### _sdn1512 - EXECUTIVE SUMMARY

### Key conclusions and policy messages
- First, housing finance characteristics vary widely across countries, and several characteristics are correlated with the relative depth of mortgage markets. Larger loan-to-value ratios (LTVs), larger reliance on wholesale funding, and, to some extent, longer loan maturities are positively correlated, together with institutional quality and macro stability, with the depth of a country’s mortgage markets and homeownership rates (with the associated benefits of higher school attainment, higher social capital, lower crime).
- Second, some of the housing finance characteristics associated with deeper mortgage markets are also associated with increased risks of crisis. For example, higher LTVs are associated with excessively rapid house-price and credit growth during booms, and wholesale funding is associated with worse outcomes in the aftermath of housing booms.
- Third, both advanced and emerging markets should avoid relaxing house financing standards in order to achieve deeper mortgage markets, and focus first on doing so through improving institutions (for example, legal rights) and the macroeconomic environment.
- Fourth, macroprudential policies, and in particular housing finance regulation, should be the first line of defense for handling mortgage market booms, as their narrow focus gives them an advantage over monetary policy. However, their effectiveness beyond the short run has yet to be proven.
- Fifth, the role of monetary policy in addressing house-related credit booms should not always be downplayed. Despite the absence of important inflation pressures, about 60 percent of the identified past real-estate booms occurred as a result of, or at the same time as, rapid economic growth and broad high credit growth in the economy. Monetary policy would be a necessary complement of macroprudential measures in those cases.
- Finally, dealing effectively with real-estate booms requires a broad mix of policies. Macroprudential and monetary policies are key ingredients, but fiscal incentives and house supply considerations are structural country-specific elements that may bear heavily on the probability of booms occurring and the potential costs of a bust.

### Mortgage market heterogeneity and empirical regularities
- Median share of mortgages in household credit in the sample (2011): about 70 percent; only six countries below a 40 percent share.
- Mortgage-to-GDP ratios in levels (based on 2001–05 averages) range from below 1 percent in Russia and Turkey to about 80 percent in the Netherlands, New Zealand, and Switzerland.
- In most countries, mortgage-to-GDP ratios as of 2011 remained above their average for 2001–05; cross-country ranking and dispersion did not significantly change over the past few years.
- Univariate regressions: differences in GDP per capita and the credit-to-GDP ratio explain, respectively, more than 50 and more than 60 percent of the variation in the mortgage-to-GDP ratio.
- Institutional factors (strength of legal rights, ease of registering property) and per-capita GDP are statistically significantly correlated with the size of mortgage markets.
- Adding house-finance characteristics (LTV, term-to-maturity, funding model) to institutional and GDP per capita improves model fit by about 10 percent, with institutional variables and GDP per capita remaining statistically significant.

### Cross-country housing finance characteristics (sample summary)
- Maximum observed LTV:
  - Range in sample: 70 percent (Colombia, Hong Kong, Hungary) to 125 percent (the Netherlands).
  - Median maximum observed LTV: 83 percent.
  - Most countries in 70–80 and 90–100 LTV buckets.
- Term to maturity:
  - Range in sample: 7 years (Turkey) to 45 years (Sweden).
  - Median term to maturity: 25 years.
- Interest type:
  - Standard mortgage rate is variable in 30 countries, fixed in 12 countries, and both contracts are observed in 14 countries.
  - Variable rates are more common in emerging economies.
- Funding model:
  - In 44 countries, banks funded mostly by retail deposits play a major role in mortgage lending.
  - The role of nonbanks and the use of wholesale and cross-border funds is important in about 22 countries (including the United States, Sweden, Denmark, Australia, South Africa).
- Degree of lender recourse:
  - In about 44 of the 53 countries in the sample, there is full recourse on mortgages.
- Mortgage interest tax deductibility:
  - In 33 of the countries in the sample, households are allowed to deduct mortgage interest payments from their taxable income.
  - Interest deductibility is more common in advanced economies than in emerging market countries (about two-thirds versus half of the cases in the sample).
  - In many cases, deductions are capped; the United States and Norway are noted as allowing nearly full deductibility without taxing imputed rents.

### Benefits and costs of deeper mortgage markets and homeownership
- Evidence supports that deepening and innovations (for example, lower down payment requirements) in mortgage markets favor homeownership.
- Microeconomic estimates: a decrease in down payment from 40 to 20 percent is associated with an 5 to 8 percentage point increase in the proportion of young households (age 26–35) living in owner-occupied houses (Chiuri and Jappelli, 2003; Chambers, Garriga, and Schlagenhauf, 2009).
- Homeownership associations:
  - Positive: promotes community investment, higher school attainment, lower crime incidence, higher life satisfaction.
  - Economic: encourages saving and investment for younger households; mortgages commit households to savings that might not otherwise occur, though equity extraction weakens this channel.
  - Housing as investment: U.S. evidence suggests housing risk-return profile compares favorably with stock market, with regional variation.
  - Negative: higher homeownership rates associated with reduced labor mobility; benefits may be smaller outside U.S.-style contexts (examples: Germany, Switzerland).
- Policy implication: policies that increase or subsidize homeownership should be based on country-specific cost-benefit analyses.

### Housing finance, house-price booms, and systemic risks
- Mortgage markets are systemically important because:
  - Real-estate-related lending often accounts for a large share of household credit and financial sector activity.
  - Mortgages introduce high leverage for households; real estate collateral affects multiple sectors; major mortgage lenders are leveraged commercial banks.
- Real-estate booms financed through fast credit growth pose significant macroeconomic risks: booms often end in busts, leading to household debt overhang, tighter lending standards, financial distress, recessions, and large increases in public debt.
- Data and measurement choices:
  - Long time series on mortgage credit are lacking; household credit is used as an approximation since the median mortgage-to-household credit ratio is about 70 percent and the correlation between mortgage and household credit ranges from 0.46 to 0.99 with a median of 0.99 and only six countries with correlation below 0.9.
  - Boom episodes are defined as large and persistent deviations of real (year-over-year) growth of house prices and credit from a country-specific historical norm, identified at quarterly frequency.

*INTERNATIONAL MONETARY FUND*

### Box 1: Cross-Country Institutional Differences Related to Housing Markets

### Box 1: Cross-Country Institutional Differences Related to Housing Markets

### Institutional heterogeneity and housing finance depth
- Legal-right index (World Bank Doing Business) is used as a proxy for the extent to which bankruptcy and collateral laws facilitate lending. The sample shows most emerging economies and a few advanced economies display relatively low legal-right indexes.
- The ease-of-registering-property index shows high heterogeneity across countries.
- Cross-country differences are less sharp for the WBDB credit-information index, which measures lenders’ access to standardized and informative sources of borrowers’ history and creditworthiness.
- Source of indices: World Bank, Doing Business, 2005 data.
- Footnotes:
  - 1/ It measures the extent to which the country’s bankruptcy and collateral laws facilitate lending.
  - 2/ It measures the depth of lenders’ access to standardized and informative sources of credit information on potential borrowers.
  - 3/ It measures the costs of registering a property.

### Definition and sample for credit and house-price booms
- Boom conditions (credit or house prices) require both:
  - (i) the real growth rate of credit (house prices) is greater than 10 (5) percent, or two standard deviations of the country-specific distribution of credit (house prices) real growth rates in a given quarter; and
  - (ii) the real growth rate of credit (house prices) is above 10 (5) percent or one standard deviation of the country-specific distribution of credit (house prices) real growth rates for a period of at least two years.
- The sample covers 53 countries with quarterly house-price, household-credit, and corporate-credit data, with sample start dates as far back as the 1970s and extending to 2012.
- Robustness: results robust to changing thresholds to the upper quarter of distributions and reducing minimum duration of booms to six quarters.

### Occurrence of credit booms: counts and patterns
- Based on definition, detected booms during 1970–2012:
  - 83 household-credit booms
  - 68 corporate-credit booms
  - 67 private-credit booms
- Most episodes (about 60 to 65 percent) occur in advanced countries; after controlling for sample composition, emerging markets are in a boom state roughly double the portion of advanced economies (portion of quarterly observations classified as booms is roughly double that of advanced economies).
- Household-credit booms are more frequent than other credit-boom types.
- Higher occurrence of credit booms after the 1980s linked to banking and mortgage deregulation and financial innovations such as securitization.
- Booms tend to cluster across countries, suggesting global factors (e.g., Federal Funds rate and VIX) play a role.

### Occurrence of house-price booms
- Identified 85 house-price booms.
- Most countries experienced at least one house-price boom.
- All-time peak in relative occurrence was just before the recent global financial crisis (in 2005, there were booms in more than half of the sampled countries).
- By 2012:Q4, less than 10 percent of countries were experiencing a credit boom (these countries were Brazil, Hong Kong, India, Israel, Malaysia, and Norway).
- Household-credit booms and house-price booms tend to co-occur; household credit is a better proxy for understanding house-price fluctuations than private-sector credit.
- Regression evidence: a 1 percent increase in household credit is associated with about 0.2 percent increase in house prices in the following year.

### Interaction between real-estate booms and credit booms; predictability
- Presence of a household-credit boom raises probability of a real-estate boom to 57 percent from an unconditional probability of 29 percent.
- Household-credit booms are better predictors of house-price booms than private-credit booms.
- The level of household debt to GDP is negatively correlated with occurrence of real-estate booms.
- Lagged GDP growth is positively associated with probability of a real-estate boom.
- Higher maximum observed LTV increases probability of a real-estate boom (relaxed lending standards effect); LTVs appear to be a targeted tool for limiting real-estate price fluctuations.

### Classification of real-estate booms by coincident credit episodes and key statistics
- Of 85 real-estate booms:
  - 49 coincide with private-credit booms (twin household and corporate)
  - 16 accompany household-credit booms only
  - 2 associate with corporate-credit booms only (Hong Kong 2004:Q1–05:Q4 and Singapore 2006:Q3–08:Q2)
  - 18 occur without any type of credit booms
- Duration and magnitude differences:
  - Average duration: house-price booms with private-credit booms ≈ 18 quarters; other cases ≈ 14 quarters.
  - Average real increase in house prices during booms:
    - With private-credit boom: 14 percent
    - With household-credit only: 9 percent
    - Without any credit boom: 10 percent
  - Average real household-credit growth during house-price booms:
    - With private-credit booms: about 21 percent
    - With household-credit only: 11 percent
    - Without any credit booms: 5 percent
- Note on private-credit booms: in housing booms, private-credit booms almost always reflect simultaneous household and corporate credit booms; four exceptions when very high household credit growth triggered private-credit booms without corporate credit increases (Canada 1973:Q1–76:Q1, Czech Republic 2006:Q3–08:Q4, Denmark 2004:Q1–07:Q1, and Greece 2005–07:Q1).

### Macroeconomic performance during real-estate booms
- Real GDP growth during housing booms is higher than during non-boom periods by about 1¼ to 2 ½ percent; these differences are statistically significant.
- Housing booms coinciding with private-credit booms register higher (statistically significant) real-GDP growth than episodes with household-credit booms only or without credit booms.
- Consumption and investment growth are higher during house-price booms with private-credit booms than in tranquil times.
- Housing booms accompanied by credit booms are associated with exchange-rate appreciation and current-account deterioration.
- Inflation typically remains subdued and similar to tranquil times during real-estate booms.

### Aftermath: declines, recessions, and prediction of "bad" booms
- House prices generally decline after real-estate booms; declines are largest for booms with private-credit booms.
- Little evidence that sharp declines are systematically followed by rebounds.
- Of house-price booms with sufficient post-period data, 49 out of 78 ended up in recessions.
- Definition: end of a housing boom is followed by a recession if real GDP growth (year-over-year) of two or more consecutive quarters is negative within a three-year interval after the end of the boom.
- About two-thirds of booms end badly; proportion similar in advanced economies and emerging markets.
- More booms occurred since 2000, and a larger proportion ended in recessions; the 2007–09 global financial crisis is influential.
- No clear relationship between boom duration and probability of ending badly.
- Real-estate booms accompanied by private-credit booms are more likely to end in recessions than those accompanied by only household-credit booms or without credit booms.
- Funding model matters: countries financing housing credit primarily through bank retail deposits have a lower probability of booms ending in recessions; wholesale funding increases leverage and vulnerability.
- About one in five housing booms (about 13 cases) are followed by a systemic banking crisis within three years after the end of the boom; no clear link found between house-price boom characteristics and systemic banking crisis in this sample.

### Policy implications and recommendations
- Cross-country differences in mortgage-market depth are associated with institutional elements (collateral and bankruptcy laws, ease of registering property), macroeconomic factors (GDP per capita), and housing finance characteristics — suggesting room for policy action.
- Improving legal frameworks and macroeconomic stability can enhance access to house financing in several emerging and advanced countries.
- Deeper and more inclusive mortgage markets correlate with higher home ownership and associated social benefits (higher school attainment, higher social capital, lower crime, higher life satisfaction and psychological health).
- Some housing finance features that increase access and affordability (higher LVRs, longer mortgage maturities) may promote rapid credit growth and greater financial stability risks.
- House-price booms funded through wholesale markets are more likely to end badly (in recessions).
- Macroprudential policies can target household leverage and indebtedness and the risk profile of mortgage originators and investors; these are useful tools for containing systemic vulnerabilities.
- Monetary policy should not be downplayed:
  - About 60 percent of real-estate booms occur together with private-credit booms, often alongside rapid and broad economic growth — signs that may call for monetary tightening after weighing benefits and risks to financial stability.
  - Monetary tightening can weaken household and firm financial conditions in the short run but can reduce leverage and strengthen financial stability over the medium term.
  - The absence of inflation pressure during many real-estate booms suggests careful consideration of including real-estate prices in monetary policy response functions.
- A broad policy mix is required beyond macroprudential and monetary measures:
  - Well-paced, country-specific supply-side measures to mitigate demand shocks over the long run.
  - Avoid abrupt supply-side modifications at peaks of booms or start of busts, as these could exacerbate price corrections.
  - Address distortions linked to special treatment of housing and homeownership.

*Source: Box 1, _sdn1512 - Cross-Country Institutional Differences Related to Housing Markets (IMF).*

### References

### _sdn1512 - References

### Annex I — Data sources and house finance characteristics
- Table A1.1 reports country-level period coverage for:
  - Household credit, Private credit, Mortgage, House price (average sample periods by country are listed).
- Table A1.2 reports institutional and house finance characteristics (selected variables shown exactly as in the source):
  - Legal Right Index, Credit Info. Index, Cost of Registering Property, MaxLTV, Term to Maturity, Tax Deduction, Full Recourse, Interest Type, Retail Funding.
  - Example entries (preserved exactly as in the source):
    - Argentina: 4 6 7 8 0 20 Yes No Variable Retail Deposit
    - Australia: 9 5 5 1 0 0 25 No Yes Variable Other
    - Canada: 7 6 1.8 9 5 25 No Yes Mixed Retail Deposit
    - USA: 9 6 0.5 1 0 0 30 Yes No Mixed Other
  - (Full country-by-country matrix appears in Table A1.2 of the source.)

### Annex II — Mortgage market depth and institutional, macroeconomic and housing finance factors
- Regression framework:
  - Dependent variable: MCY — ratio of mortgage credit to GDP (average 2001–05).
  - Institutional variables: legal-rights index, credit-information index, ease-of-registering property (2001–05 averages).
  - Macroeconomic variables: average GDP per capita (in log), volatility of inflation (90–07).
  - Housing finance variables: tax-deductibility dummy, maximum observed LTV, full-recourse dummy, interest-type variable (1–3 increasing in popularity of fixed-rate mortgages), term to maturity (years), retail funding dummy.
- Main empirical findings (text summary and exact reported statistics preserved):
  - The legal rights index is a variable "very significantly associated with a deeper mortgage market."
  - Higher GDP per capita is "strongly and significantly associated with deeper mortgage markets."
  - Ease of registering a property is associated with deeper mortgage markets.
  - Maximum observed LTV shows a positive relationship with the size of the mortgage market (text references column 4 and robust regression in column 10).
  - Typical duration of mortgage contracts is positively correlated with mortgage market depth in some specifications, but "this result is not robust."
  - Countries where the main originators are banks that fund themselves primarily with retail deposits have significantly lower mortgage-to-GDP ratios.
  - Housing finance characteristics—LTV, term to maturity, and funding model—"contribute to explaining about an additional 10 percent of the cross-country variation in the depth of mortgage markets relative to a specification based solely on institutional and macro variables."
- Selected regression coefficients and statistics (preserved exactly from Table A2.1):
  - Legal rights index: 4.175*** (column 1) with ( 0.003) reported.
  - Credit info. index: 2.899 (column 1) with ( 0.108).
  - Ease of registering property: 0.0877* (one specification) with ( 0.099).
  - Log of GDP per capita: 12.62*** (column 2) with (0.000); other columns report 12.53***, 12.96***, 14.06***, 12.01***, 12.19***, 13.25***, 14.47***, 14.38*** (all with (0.000)).
  - CPI volatility (90-07): -0.983 (column 1) with (0.206).
  - Tax deduction: 1.269 (one column) with (0.795); other columns report -2.754, -4.159 with (0.564) and (0.394).
  - Max observed LTV: 0.319* with ( 0.083); also reported 0.296 ( 0.115) and 0.426** ( 0.034) in other columns.
  - Full recourse: -7.764 with ( 0.206); other columns -7.719 ( 0.188) and -9.742 ( 0.104).
  - Interest type: 3.368 (0.253), 2.605 (0.355), 2.030 (0.476) across specifications.
  - Term to maturity: 0.634* with (0.056); other columns 0.457 (0.180) and 0.210 (0.554).
  - Retail funding: -9.816* with ( 0.073); -10.48* ( 0.053); -10.74** ( 0.050).
  - Constant terms (examples): -9.578 ( 0.380); -105.8*** ( 0.000); -106.2*** ( 0.000); -138.5*** ( 0.000); -113.7*** ( 0.000); -107.2*** ( 0.000); -115.7*** ( 0.000); -101.3*** ( 0.000); -146.4*** ( 0.000); -144.6*** ( 0.000).
  - Observations: 53 in most columns; 52 in column (10).
  - R-sq: 0.291 (col 1), 0.600 (col 2), 0.600 (col 3), 0.625 (col 4), 0.613 (col 5), 0.611 (col 6), 0.630 (col 7), 0.627 (col 8), 0.696 (col 9), 0.692 (col 10).
- Note (as in source): The last regression in column (10) reports the result from a robust regression. ***, **, * indicates statistical significance at the 1, 5, and 10 percent, respectively.

### Annex III — Robustness analysis of booms definitions
- Purpose: Test sensitivity of identified boom episodes to alternative filters and thresholds (backward-looking cubic trend, Hodrick-Prescott filter, different thresholds separating one-quarter or one-third of the real growth rate distribution, and minimum boom durations of six vs. eight quarters).
- Key qualitative conclusion from the text:
  - "The list of episodes we identify is not very sensitive to the methodology used. The major booms are captured under all methodologies. Differences appear in small- and medium-sized booms where different thresholds matter more."
  - The incidence of post-bust recessions remains similar across methodologies, "varying in a relatively close range from 59 percent to 67 percent."
- Number of booms and share followed by recession within three years of boom end (Table A3.2 — exact figures preserved):
  - Baseline (8 quarters): Total number of booms 85; Followed by recession within three years of boom end 64.1%
  - Topthird (6 quarters): Total number of booms 104; Followed by recession 64.2%
  - Topquarter (8 quarters): Total number of booms 70; Followed by recession 59.1%
  - Cubic trend (8 quarters): Total number of booms 87; Followed by recession 67.1%
  - HP trend (8 quarters): Total number of booms 89; Followed by recession 66.7%
- Correlations across different methodologies (Table A3.1 — summary of patterns):
  - Simple correlations and tetrachoric correlations between alternative boom-identification methodologies are generally high.
  - Examples (preserved from table headings and some entries):
    - House price simple correlation: Baseline vs Topthird6 = 0.939; Baseline vs Topquarter = 0.802; Baseline vs Cubic trend = 0.918; Baseline vs HP trend = 0.907.
    - Household credit simple correlation: Baseline vs Topthird6 = 0.966; Baseline vs Topquarter = 0.898; Baseline vs Cubic trend = 0.842; Baseline vs HP trend = 0.819.
    - Corporate credit simple correlation: Baseline vs Topthird6 = 0.921; Baseline vs Topquarter = 0.994; Baseline vs Cubic trend = 0.825; Baseline vs HP trend = 0.806.
    - Private credit simple correlation: Baseline vs Topthird6 = 0.936; Baseline vs Topquarter = 0.955; Baseline vs Cubic trend = 0.834; Baseline vs HP trend = 0.773.
  - Corresponding tetrachoric correlations reported alongside the simple correlations in Table A3.1.

*International Monetary Fund — content as provided in the source PDF*

### Annex IV: Credit Booms as Predictors of Housing Booms.

### Annex IV: Credit Booms as Predictors of Housing Booms

### Probit model and data
- Estimated model: Probit where dependent variable "housing boom" = 1 for real-estate boom, 0 otherwise.
- Main predictor: credit_boom = four-quarters lag credit-boom dummy in household or private credit.
- Control variables (lagged by four quarters): log of per-capita real GDP, level of short-term interest rates, household indebtedness, the VIX, GDP growth, CPI inflation, current account balance (percent of GDP), and housing finance characteristics where available.
- Lags and estimation choices:
  - All "slow moving" regressors lagged by four quarters to explore predictive power and to help reduce endogeneity.
  - Time dimension of panel ~100 observations on average; Fernandez-Val (2009) probit bias-corrected estimator deemed unnecessary (no incidental parameters bias problem). (Footnote 19)
  - Housing finance characteristics available only for recent years; estimations with those variables run for period 2000–12 without country fixed effects.

### Main empirical findings
- Predictive power of household-credit booms:
  - Presence of a household-credit boom increases the probability of a real-estate boom to 57 percent against an unconditional probability of 29 percent.
  - Across specifications, household-credit booms are better predictors of house-price booms than private-credit booms.
- Household indebtedness:
  - The level of household debt to GDP (proxy for depth of mortgage markets) is statistically significant with a negative coefficient, indicating that higher initial household leverage levels are associated with a lower occurrence of real-estate booms.
- Global factors:
  - Federal Funds rate and VIX have expected negative signs and are statistically significant across most specifications, reflecting global factors that simultaneously drive house-price booms across countries.
- Real activity and external conditions:
  - Lagged GDP growth is positively associated with the probability of a real-estate boom, indicating booms tend to start during or immediately after periods of buoyant economic growth.
  - Current account balance coefficient is positive and statistically significant across specifications, signaling that, on average, real-estate booms are more likely to start during favorable external conditions; but this relation does not hold in all countries (examples noted: Germany and South Korea associated with current account surpluses; United States associated with current account deficits).

### Housing finance characteristics and LTV
- When including housing finance characteristics for 2000–12 (no country fixed effects):
  - Results remain similar for household-credit booms, household leverage, GDP growth, and current account balance.
  - Max observed LTV is the only house finance variable reported that is statistically significant.
  - Interpretation: Higher maximum observed LTV → higher probability of a real-estate boom, likely capturing relaxed lending standards. Cited supporting studies: Crowe and others, 2011; IMF 2011; Cerutti, Claessens, and Laeven, 2015.
  - Policy implication noted: Reducing LTVs through LTV limits would be a well-targeted objective, with effectiveness depending on whether such limits erode over time (e.g., regulatory arbitrage).

### Table A4.1: Key coefficient estimates (probit, period 1970q1–2012q4 unless noted)
- HH Credit Boom (lag):
  - Column (1): 0.784*** (0.237)
  - Column (2): 0.734*** (0.241)
  - Column (3): 0.753*** (0.240)
  - Column (4): 0.758*** (0.236)
  - Column (5): 0.571*** (0.202)
- Private Credit Boom (lag):
  - Column (1): 0.310 (0.224)
  - Column (2): 0.516** (0.225)
  - Column (3): 0.302 (0.235)
  - Column (4): 0.494** (0.233)
  - Column (5): 0.397** (0.189)
- HH Indebtedness (lag):
  - Column (1): -0.0331*** (0.00947)
  - Column (2): -0.0439*** (0.0109)
  - Column (3): -0.0288*** (0.00953)
  - Column (4): -0.0363*** (0.0114)
  - Column (5): -0.0137*** (0.00397)
- Log of GDP per capita (lag):
  - Column (1): 1.642** (0.676)
  - Column (2): 1.299 (0.987)
  - Column (3): 0.672 (0.637)
  - Column (4): 0.713 (1.033)
  - Column (5): 0.248** (0.111)
- US Fed Fund Rate:
  - Column (1): -0.0358* (0.0204)
  - Column (2): -0.0289 (0.0270)
- VIX Index:
  - Column (2): -0.0210*** (0.00648)
  - Column (3): -0.0201*** (0.00662)
  - Column (4): -0.0190*** (0.00545)
- Current Account (lag):
  - Column (1): 0.0578*** (0.0161)
  - Column (2): 0.0716*** (0.0217)
  - Column (3): 0.0367*** (0.0142)
- GDP growth (lag):
  - Column (1): 0.0723*** (0.0163)
  - Column (2): 0.0668*** (0.0179)
  - Column (3): 0.0552*** (0.0128)
- CPI Inflation (lag):
  - Column (1): -0.0754*** (0.0228)
  - Column (2): -0.0331 (0.0284)
  - Column (3): -0.00302 (0.0216)
- Max Observed LTV (included in Column 5 only):
  - Column (5): 0.0214*** (0.00568)
- Estimation details per column:
  - Country Fixed Effects: YES for Columns (1)-(4); NO for Column (5).
  - Observations: Column (1): 440; Column (2): 630; Column (3): 814; Column (4): 2183; Column (5): 213315 (as printed).
  - R-sq: Column (1): 0.191; Column (2): 0.252; Column (3): 0.254; Column (4): 0.289; Column (5): 0.172
- Notes from table:
  - Estimates over period 1970q1–2012q4, robust standard errors clustered at country level.
  - Lagged variables correspond to 4 quarters lags.
  - Significance: *** 1 percent, ** 5 percent, * 10 percent.
  - Max observed LTV is the only house finance variable reported because others were not statistically significant.
  - Results robust to selection of number of lags. In regression without house finance characteristics, country fixed effects included in addition to time-varying variables.

### Policy-relevant conclusions
- Household-credit booms are strong predictors of house-price booms; monitoring household credit cycles provides early warning signal for housing booms.
- LTV (maximum observed) is a significant indicator: higher observed LTVs raise boom probability, suggesting macroprudential policies that limit LTVs could be well-targeted to reduce house-price boom risk, subject to enforcement and regulatory arbitrage considerations.
- Global financial conditions (lower Fed Funds rate, lower VIX) and favorable external positions (current account surpluses) are associated with higher likelihood of house-price booms.
- Booms are more likely to start amid or after periods of positive GDP growth.

*Source: Annex IV: Credit Booms as Predictors of Housing Booms (extracted from the supplied IMF content).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/sdn/2015/_sdn1512.pdf_