## hmsaapea

## Source details

**Canonical URL:** [hmsaapea](https://www.imf.org/-/media/files/publications/dp/2022/english/hmsaapea.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2022/english/hmsaapea.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2022/english/hmsaapea.pdf.json)

---

### Executive summary — recent housing dynamics and key findings
- Housing markets in the Asia-Pacific region "are at an important juncture." Housing prices rose over the past decade and during the pandemic and "now appear slated for a decline in many countries."
- Pandemic price surges were driven by low mortgage rates and country-specific demand and supply factors, particularly in the region’s advanced economies (AEs).
- Price misalignment and downside risks quantified as "in the order of 5-20 percent in some countries."
- Pandemic-era demand shifts included a structural move toward larger dwellings and locations farther from job centers (telework and movement restrictions).
- Post-pandemic withdrawal of monetary stimulus and rising mortgage rates are already cooling demand; market corrections are underway in some economies.

### Selected quantitative indicators and notable episodes
- New Zealand: "real housing prices had increased most in Asia-Pacific during 2020–21 (nearly 35 percent)"; the central bank "has raised the policy rate by 400 basis points since September 2021," and housing prices "have already declined for two consecutive quarters, with the decline in 2022:Q2 having been the most pronounced since 2009."
- Historical peak-to-trough real housing price amplitudes (1970–2021) examples (Amplitude, percent):
  - China: −7.4
  - Australia: −13.7
  - Malaysia: −21.4
  - Singapore: −26.6
  - Indonesia: −27.2
  - Thailand: −32.0
  - Korea: −33.7
  - New Zealand: −38.2
  - United States: −39.1
  - Japan: −45.2
  - Hong Kong SAR: −51.0
- Downside risk magnitudes reiterated as "in the order of 5-20 percent in some countries."

### Risks to financial stability and households
- Pronounced housing cycles frequently reoccur and "come at a cost"; vulnerabilities from upswings manifest during downturns, affecting the broader economy.
- Financial sectors in major AEs and EMs "appear sound and would be expected to remain resilient under such shocks," but close supervision is warranted to identify pockets of risk.
- Housing affordability deterioration:
  - Housing-related items account for 20 percent or more of the CPI basket in many Asia-Pacific economies.
  - Housing cost overburden: in some countries the share of population spending more than 40 percent of income on housing exceeds OECD averages.
- In regional EMs, supply often fails to keep pace with rapidly growing populations; poorer households are disproportionately affected.

### Effectiveness of macroprudential policies (MPMs)
- MPMs "should be the first line of defense" to address the housing cycle and safeguard financial stability.
- Aggregate effect on household credit:
  - A one-standard-deviation tightening of the aggregate macroprudential policy index is associated with 0.5–1 percentage point lower household credit growth.
  - Most impact materializes after about five quarters.
- Asymmetry:
  - Policy tightening effects are generally stronger than loosening effects; loosening has smaller, often statistically insignificant impacts.
- Relative effectiveness:
  - Demand-side measures (LTV, DSTI) "have been more effective than capital-based tools" in curbing household credit.
  - LTV/DSTI directly curb lending to higher-risk borrowers; capital requirements show weaker direct effects on lending.
- Impact on housing prices:
  - MPMs have a limited average effect on real housing prices across Asia-Pacific; statistically significant but small and short-lived effects exist in regional EMs.
- Leakages and limits:
  - Leakages via nonbanks, regulatory arbitrage, and cross-border channels reduce MPM effectiveness; policymakers "should widen the regulatory perimeter where needed."

### House price-at-risk (HaR) quantitative findings
- Method: panel quantile regression focusing on the 5th percentile for a sample of 12 regional economies; regressors include misalignment, short-term interest rates, output gap, household credit deviations, and lagged price growth.
- Misalignment effects:
  - In EMs: a 1 percent increase in misalignment is associated with a 0.6 percent decline in four-quarter-ahead housing price growth at the 5th percentile.
  - In AEs: the effect is almost twice as large (roughly 1.2 percent decline for a 1 percent increase in misalignment at the 5th percentile over four quarters).
- Short-term interest rate effects:
  - A 100 basis point increase in short-term interest rates is associated with four-quarter-ahead housing price growth being about 0.7 percent lower in both AEs and EMs.
  - The negative effect more than doubles in AEs at the eight-quarter horizon compared to four quarters ahead.
- Momentum and other drivers:
  - Higher lagged housing price growth is associated with higher future price growth (persistence).
  - Positive output gaps are associated with a decline in future housing prices as the economy reverts toward potential.
  - Household credit deviations have mixed significance (no statistically significant impact in EMs at short horizons; impact in AEs mainly at longer horizons).
- Evolving downside risk since the pandemic (fitted quantile-regression results as of 2021:Q4):
  - New Zealand: four-quarter-ahead house price growth predicted about 20 percentage points lower than at the start of the pandemic.
  - Australia and Korea: four-quarter-ahead house price growth predicted about 10 percentage points lower than at the start of the pandemic.
  - Rising interest rates in 2022 and expected further hikes will add to downside risks beyond 2021:Q4 levels.

### Housing affordability and quality metrics
- Homeownership affordability proxy (borrowing capacity, DSTI benchmark 30 percent):
  - 2022 estimates indicate housing prices may be as much as 70 percent above what a median household can afford in New Zealand and 50 percent above in Australia based on a DSTI ratio of 30 percent.
  - The average household would need to spend more than 40 percent of disposable income to afford the median-priced dwelling in those countries.
- Housing quality in EMDEs:
  - On average, close to 30 percent of the urban population in regional EMs live in slums; in Asia-Pacific developing countries the share is 40 percent.
  - The Behr and others (2021) housing inadequacy index shows regional EMs do significantly worse than non-Asian peers.
- Household debt trends:
  - Household debt climbed to historical highs in some regional economies (credit to households, 2007–21, percent of GDP); notable high household debt in Australia, Hong Kong SAR, Korea, New Zealand.

### Policy recommendations — supply-side measures
- Facilitate stronger housing supply response:
  - Review land use regulations (remove zoning restrictions that limit higher-occupancy buildings or greenfield development).
  - Increase focus on urban planning to ensure proximity to jobs and services.
  - Incentivize use of idle land (land value taxation cited).
  - Provide adequate social and affordable housing directly and via incentives for private developers.
- Supply-side complements:
  - Use low-interest loans, subsidies, and tax credits for private developers to stimulate affordable housing construction.
  - Examples cited: New Zealand Resource Management Act amendment (2021); Thailand Land and Building Tax (2020) with maximum tax rate ranges from 0.15 percent to 1.2 percent; Malaysia’s NAHP (2019).

### Policy recommendations — demand-side, fiscal, and financing measures
- Demand-side (short-term, targeted):
  - Targeted government support: vouchers, housing subsidies, grants, tax relief, accommodative financing terms (examples across Australia, Malaysia, New Zealand, Thailand).
  - Progressive taxation on property purchases (stamp duties) and transformation of stamp duties to recurrent property taxes to lower up-front costs (example: New South Wales, Australia).
  - Target macroprudential policy to contain systemic risks while protecting lower-income households and owner-occupiers (e.g., tighten LTV only for higher-risk borrower types).
- Financing and insurance mechanisms:
  - Government housing finance agencies, mortgage insurance, and guarantee schemes (examples: Cagamas Berhad in Malaysia; HDB in Singapore; Government Housing Bank in Thailand).
  - Mandatory national saving vehicles: Singapore CPF (employee contribution 20 percent of monthly salary; employer contribution 17 percent), Indonesia Tapera (employees contribute 2.5 percent; employers contribute 0.5 percent).
  - Community-led finance examples: Philippines’ Community Mortgage Program (CMP).

### Macroprudential design and coordination
- Primary role: MPMs to build resilience during booms and limit systemic risk during downturns.
- Design guidance:
  - Use targeted LTV and DSTI restrictions for high-risk borrowers.
  - Apply measures addressing speculative demand (capital gains taxes on quick resales, property holding taxes).
  - Refine tools to limit leakages (expand scope to non-banks and foreign credit providers; pursue jurisdictional reciprocity and host control for foreign affiliates).
  - Maintain reasonable minimum standards during downturns; recognize loosening is not symmetric in effect relative to tightening.
  - Coordinate macroprudential policy with monetary and fiscal policy given high household debt, tightening conditions, and pandemic scarring.

### Conclusion and policy priorities
- Housing prices appear misaligned especially among the region’s AEs, increasing risk of price corrections.
- Key drivers of housing price-at-risk are the magnitude of price misalignment and borrowing costs; combined high misalignment and high policy rates can lead to sizable price corrections "nearly comparable to past housing bust episodes."
- High household debt can amplify effects of housing corrections on the real economy and financial stability.
- Multipronged approach emphasized:
  - Macroprudential policies as the first line to safeguard financial stability.
  - Structural supply-side reforms for long-term affordability restoration.
  - Targeted demand-side and financing measures to protect vulnerable households while avoiding measures that exacerbate imbalances.

*Italic: Source: Executive Summary and selected chapters from "Housing Stability and Affordability in Asia-Pacific" (IMF Departmental Paper) as provided in the content unit.*

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

### Executive Summary

### Key findings on recent housing dynamics
- Housing markets in the Asia-Pacific region "are at an important juncture." Housing prices rose over the past decade and during the pandemic and "now appear slated for a decline in many countries."
- Price surges during the pandemic were driven by low mortgage rates and country-specific demand and supply factors, particularly in the region’s advanced economies (AEs).
- Price misalignment and downside risks: housing prices in many countries "now appear misaligned, with significant downside risk to prices in the period ahead," quantified as "in the order of 5-20 percent in some countries."
- Pandemic-era shifts in housing demand included a structural shift toward larger dwellings and locations farther from job centers due to telework and movement restrictions.
- The post-pandemic withdrawal of monetary stimulus and rising mortgage rates are already cooling demand; market corrections are underway in some economies.

### Selected quantitative indicators and episodes
- New Zealand: "real housing prices had increased most in Asia-Pacific during 2020–21 (nearly 35 percent)"; the central bank "has raised the policy rate by 400 basis points since September 2021," and housing prices "have already declined for two consecutive quarters, with the decline in 2022:Q2 having been the most pronounced since 2009."
- Historical housing cycle amplitudes (1970–2021): peak-to-trough real housing price adjustments vary "from 4 percent in China to 18 percent in Hong Kong SAR."
- Downside risk magnitudes: "in the order of 5-20 percent in some countries" driven by pandemic-era misalignment and rising interest rates.

### Risks to financial stability and households
- Pronounced housing cycles frequently reoccur in Asia-Pacific and "come at a cost." Vulnerabilities built up during upswings tend to manifest during downturns with marked impacts on the broader economy.
- While financial sectors in major AEs and emerging market economies (EMs) "appear sound and would be expected to remain resilient under such shocks," close supervision is warranted to identify pockets of risk.
- Housing affordability has deteriorated, especially in AEs: households "have to stretch their wallets" and an increasing share are "overburdened" by housing costs.
- In many EMDEs, adequate housing is less affordable due to supply failing to keep pace with rapidly growing populations; poorer households are disproportionately affected.

### Effectiveness of macroprudential policies (MPMs)
- Macroprudential policies "should be the first line of defense" to address the housing cycle and safeguard financial stability.
- Asia-Pacific countries "have a broad toolkit at their disposal," and MPMs "have proven effective at mitigating housing credit growth during upswings."
- Within the toolkit:
  - Demand-side measures, such as loan to value (LTV) and debt service to total income (DSTI) limits, "have been more effective than capital-based tools."
  - The effect of policy tightening during upswings has been "generally stronger than stimulatory effects of policy loosening during downturns."
  - MPMs are effective at targeting household credit growth but "only have a limited effect on housing prices on average, and only significantly so in regional EMs."
  - Leakages of macroprudential measures are of concern; "policymakers should widen the regulatory perimeter where needed."

### Policy recommendations to improve affordability
- A multi-faceted approach is required; facilitating a stronger supply response is "key in many countries" and includes:
  - "reviewing land use regulations,"
  - "increasing the focus on urban planning,"
  - "incentivizing the use of idle land," and
  - "providing adequate social and affordable housing, both directly by the public sector and indirectly through incentives for private developers."
- Since supply-side measures "often take significant time to produce results," demand-side measures are important for quicker impact. These can include:
  - targeted government support,
  - progressive taxation on property,
  - targeted macroprudential policy to contain systemic risks while being mindful of repercussions on lower-income households and owner-occupiers,
  - targeted financing, insurance, and guarantee mechanisms.

### Conclusion and implications for policymakers
- Housing price deceleration appears underway in several markets; central bank tightening and higher mortgage rates will add to downside risks and further strain affordability.
- Policymakers should rely on macroprudential tools as the first line of defense to safeguard financial stability, while pursuing structural supply-side reforms and targeted measures to protect affordability for vulnerable households.
- Close supervision and timely identification of pockets of risk are warranted given the potential for pronounced spillovers from housing downturns.

*Source: Executive Summary, Housing Stability and Affordability in Asia-Pacific (IMF Departmental Paper).*

### 1. Many Regional Economies Had Large Real Housing

### 1. Many Regional Economies Had Large Real Housing Price Increases Prior to the Pandemic

### Pre-pandemic and pandemic-era price dynamics
- Real housing prices rose substantially across many Asia-Pacific economies from 2011:Q1 to 2019:Q4 and then surged further from 2019:Q4 to 2021:Q4.
- Housing cycles tend to be larger in advanced economies (AEs), with Hong Kong SAR, Japan, Korea, and Singapore recording larger average peak-to-trough housing price declines than the United States (used here as a comparison).
- Developing economies in Asia-Pacific tended to have smaller, but more frequent, housing cycles.
- Many economies have experienced contractions of 30 percent or larger in peak-to-trough real housing prices.

### Largest observed peak-to-trough housing market adjustments (Real housing prices, largest peak-to-trough adjustments by economy, percent)
- China — Peak: 2014:Q1, Trough: 2014:Q4, Duration (quarters): 4, Amplitude (percent): −7.4
- Australia — Peak: 1981:Q2, Trough: 1983:Q3, Duration (quarters): 10, Amplitude (percent): −13.7
- Malaysia — Peak: 1997:Q2, Trough: 1999:Q1, Duration (quarters): 8, Amplitude (percent): −21.4
- Singapore — Peak: 2008:Q1, Trough: 2009:Q2, Duration (quarters): 6, Amplitude (percent): −26.6
- Indonesia — Peak: 2003:Q3, Trough: 2008:Q4, Duration (quarters): 22, Amplitude (percent): −27.2
- Thailand — Peak: 1997:Q2, Trough: 1999:Q2, Duration (quarters): 9, Amplitude (percent): −32.0
- Korea — Peak: 1991:Q2, Trough: 1996:Q3, Duration (quarters): 22, Amplitude (percent): −33.7
- New Zealand — Peak: 1974:Q3, Trough: 1980:Q4, Duration (quarters): 26, Amplitude (percent): −38.2
- United States — Peak: 2006:Q1, Trough: 2011:Q2, Duration (quarters): 22, Amplitude (percent): −39.1
- Japan — Peak: 1991:Q1, Trough: 2006:Q3, Duration (quarters): 63, Amplitude (percent): −45.2
- Hong Kong SAR — Peak: 1981:Q1, Trough: 1984:Q3, Duration (quarters): 15, Amplitude (percent): −51.0

### Economic spillovers and importance for households
- Pronounced housing cycles—especially “housing busts”—are associated with severe recessions in most cases.
- Housing is typically the main source of households’ wealth and debt; large declines in housing prices can severely damage household balance sheets and cashflows.
- Housing busts often trigger severe contractions in construction and related sectors after boom-period growth.
- Housing, acting as collateral, can amplify business cycles as a financial accelerator.
- Large declines in housing net worth tend to be associated with sizable declines in employment and consumption.
- Housing busts can have adverse impacts on financial stability.

### CPI and distributional impacts
- Housing-related items account for 20 percent or more of the CPI basket in many Asia-Pacific economies, and this share is larger than the OECD average.
- Housing costs tend to be a larger share of total living costs for lower-income households (examples: New Zealand and Singapore).

---

### 2. Drivers of the Pandemic-Era Housing Price Surge in Advanced Asia-Pacific

### Structural decomposition findings (AEs only)
- A four-variable structural vector autoregression with sign restrictions was used to disentangle drivers of real housing price changes (applied to AEs due to EMDE data limits).
- Low mortgage rates associated with accommodative monetary policy stances played an important role in recent housing price increases.
  - Example: In Australia and New Zealand, mortgage rate shocks linked to conventional and unconventional monetary policy tools during the acute pandemic phase boosted housing prices; strong demand also contributed.
- Japan and Korea: supply shortages and stronger housing price expectations contributed to price increases.
- Singapore: contemporaneous demand remained subdued; low mortgage rates and stronger expectations contributed to the increase.
- Common factors across regional AEs and EMs (easing of international financial conditions during the pandemic among them) also drove prices.

---

### 3. Misalignment, Measurement Approaches, and Extent of Overvaluation

### Methods to measure price misalignment (applied to AEs)
I. Long-term average: compute average over entire sample period and treat deviations as misalignment.  
II. Hodrick-Prescott (HP) filter: removes cyclical component and gives more weight to recent observations; generally yields lower measured misalignment.  
III. Regression-based trend: controls for fundamentals (interest rates, lagged dwelling investment, lagged population growth, GDP growth) and captures misalignment unexplained by fundamentals.

### Findings on misalignment
- All three methods indicate significant price misalignment in a number of regional AEs (and some international comparators including Canada, the United Kingdom, and the United States), though magnitudes vary by method.
- Deviations from long-term average tend to be much larger due to the upward trend in price ratios over the past two decades.
- The HP filter and regression-based approach are directionally similar; the regression-based approach focuses on misalignment not explained by fundamentals and also shows significant overvaluation for most AEs in Asia-Pacific.
- Valuations are also stretched in some EMs, including China (analysis for EMs limited by data).

---

### 4. Downside Risks to Housing Markets: Housing Price-at-Risk Analysis

### Methodology summary
- Housing price-at-risk estimated using panel quantile regression (focus on 5th percentile to highlight downside risks) for a sample of 12 regional economies; analyses done separately for AEs and EMs.
- Regress future housing price changes on past housing price growth and various fundamentals (Annex 2 contains methodological details).

### Key quantitative regression results and interpretations
- Housing price misalignment:
  - A 1 percent increase in misalignment (deviation from HP-filter trend in house-price-to-income ratio) is associated with a 0.6 percent decline in four-quarter-ahead housing price growth at the 5th percentile in EMs.
  - The effect of misalignment on downside risk is almost twice as large in AEs (implying roughly 1.2 percent decline for a 1 percent increase in misalignment at the 5th percentile over four quarters).
- Short-term interest rates:
  - A 100 basis point increase in short-term interest rates is associated with four-quarter-ahead housing price growth being about 0.7 percent lower in both AEs and EMs.
  - The negative effect of interest rates is larger at longer horizons: the effect more than doubles in AEs at the eight-quarter horizon compared to four quarters ahead.
  - The effect in Asia-Pacific is bigger than in other regions.
- Output gap:
  - Positive output gaps are associated with a decline in future housing prices as the economy reverts towards potential.
- Household credit:
  - Deviations of household credit-to-GDP ratios from trend do not have a statistically significant impact on housing price-at-risk in EMs.
  - Household credit deviations impact AEs only at longer horizons.
- Housing price momentum:
  - Higher lagged housing price growth is associated with higher future price growth (persistence).

### Evolving downside risk since the pandemic
- Since the start of the pandemic, downside risks to housing prices have risen significantly in many AEs.
- As of 2021:Q4, fitted quantile-regression results indicate:
  - Four-quarter-ahead house price growth is predicted to be about 20 percentage points lower in New Zealand than at the start of the pandemic.
  - Four-quarter-ahead house price growth is predicted to be about 10 percentage points lower in Australia and Korea than at the start of the pandemic.
- The deterioration in housing price-at-risk mainly reflects greater misalignment; momentum from lagged prices mitigates some of the misalignment effect.
- By contrast, housing price-at-risk changed much less in most EMs in Asia-Pacific, possibly reflecting lower mortgage credit and household debt and differences in mortgage market structures.
- Rising interest rates in 2022 and expected further hikes will add to downside risks beyond levels captured through 2021:Q4 data; higher rates will further reduce future housing price growth.

---

*Source: IMF staff estimates and analysis (from "Housing Stability and Affordability in Asia-Pacific").*

### 1. Advanced Economies2. Emerging Market Economies

### 1. Advanced Economies2. Emerging Market Economies

### Housing price downside risks and drivers
- Quantile regressions (5th percentile) show housing price misalignment, short-term interest rate, output gap, and credit deviation from trend affect four-quarter-ahead real house price growth. Coefficients on housing price misalignment, output gap and credit show the impact of a 1 percent increase in each variable, while the coefficient on short-term interest rate shows the impact of a 100 basis point increase.
- Greater misalignment and higher interest rates add to downside risks to housing prices; light-shaded coefficients are insignificant at 10 percent confidence level.
- Higher interest rates have a bigger effect on housing prices at the 5th percentile compared to the median, indicating higher rates add more to downside risks.
- A 300 basis points increase in short-term interest rates:
  - Can further reduce eight-quarter-ahead housing price growth by more than 5 percent in Advanced Economies.
  - Leads to significantly lower housing price growth, with effects larger at longer horizons and in AEs.
- Rising mortgage rates and debt service obligations increase the importance of monitoring financial sector vulnerabilities, given some economies in the region face high and rising household debt as a share of GDP and a sizable share linked to floating interest rates (Australia, Korea, New Zealand).

### Evolution of Housing Price-at-Risk (HaR) in Asia-Pacific
- Housing Price-at-Risk has deteriorated significantly in Asia-Pacific, especially in Advanced Economies: negative changes in the 5th percentile house price growth between 2019:Q4 and 2021:Q4 imply an increase in downside risk to house prices.
- Figure-based country groupings used International Organization for Standardization (ISO) country codes; panels show changes in fitted values from panel quantile regressions (Annex 2).

### Housing affordability and quality
- Broad affordability measure: share of household expenditure spent on housing and basic utilities (water, electricity, gas and other fuels).
  - Compared to the OECD average, housing-related expenditures form a higher share of total spending in Australia, Japan, and New Zealand.
  - In both Australia and Japan, the housing cost overburden rate (share of the population spending more than 40 percent of its income on housing) is higher than the OECD average.
- Homeownership affordability proxy (borrowing capacity approach):
  - Uses maximum mortgage loan size attainable given income, prevailing mortgage rate, and leverage requirements.
  - Debt-service-to-income (DSTI) benchmark of 30 percent for the median-income household used to calculate price-consistent benchmarks.
  - In 2022, housing prices may be as much as 70 and 50 percent above what a median household can afford in New Zealand and Australia, respectively, based on a DSTI ratio of 30 percent.
  - The average household would need to spend more than 40 percent of its disposable income to afford housing priced at the median in those countries.
- Methodological assumptions for attainable housing prices:
  - Average annual household disposable income by fiscal year-end (end-June).
  - Affordable housing cost assumed in the conventional range of a DSTI ratio of 30 to 40 percent.
  - Mortgage rate: one-year fixed average residential mortgage rate (monthly), applying to a principal and interest loan of 30-year maturity.
  - Loan-to-value ratio up to 80 percent with a down payment of 20 percent.
  - Actual housing prices are the median value of dwellings in June of each year. For 2022, estimates based on projected housing prices, average household disposable income, and interest rates.
  - Note: DSTI 30 percent benchmark does not account for income distribution or individual household liquidity buffers.
- Housing quality in EMDEs:
  - On average, close to 30 percent of the urban population in regional EMs live in slums; in Asia-Pacific developing countries the share is 40 percent.
  - The Behr and others (2021) housing inadequacy index (combining access to water, sanitation, electricity, living space adequacy, durability, and security of tenure) shows regional EMs do significantly worse than non-Asian peers.
  - Significant within-country heterogeneity: poorer households score significantly worse on the housing inadequacy index than richer households; inequality in housing quality is particularly large in Bangladesh, India, Nepal, and the Philippines.
  - UN-HABITAT slum household definition components are listed and were used in panel notes.

### Household debt dynamics
- Household debt has climbed to historical highs in some regional economies (credit to households, 2007–21, percent of GDP). Country series illustrated include AUSHKGJPNKORNZLSGPCHNIDNINDMLYTHA (as reported).

### Macroprudential policies and financial stability
- Since the GFC, AEs and EMs in Asia-Pacific have gained institutional knowledge and built macroprudential toolkits, including:
  - Instruments for the household sector: loan-to-value ratio (LTV) restrictions, DSTI limits, loan restrictions, and capital-based tools.
- Deployment patterns:
  - Macroprudential tools have been increasingly deployed over the years to address risks and build resilience.
  - The COVID-19 pandemic ended a decade of macroprudential tightening: authorities eased restrictions during the pandemic to temporarily reduce liquidity requirements and release capital buffers to support financial system functioning and encourage bank lending to the private sector.
  - Some EMs relaxed restrictions on bank lending to households during this phase.
  - The macroprudential dataset ends in 2020; subsequent evidence indicates many pandemic-related support measures were withdrawn (at least partly) in 2021, and macroprudential policy tightened again to tackle emerging risks, including strong housing cycles (Australia, Korea, New Zealand).
  - With housing cycles turning amid higher interest rates, there has been less macroprudential policy action so far in 2022.
- Effectiveness of Macroprudential Policy Measures (MPMs):
  - Analysis using the iMaPP database and local projection techniques for Asia-Pacific finds:
    - Macroprudential tools can be effective at addressing financial stability risks from the housing sector by curtailing credit growth in general and to risky borrowers in particular.
    - The impact of macroprudential tools on housing prices is limited.
  - Local projection method used to estimate dynamic effects and impulse response functions, controlling for initial macroeconomic conditions and subsequent policy actions (Annex 3).

_Italic: Source: IMF staff calculations; IMF DEPARTMENTAL PAPERS • Housing Stability and Affordability in Asia-Pacific (figures and notes as provided in the source content)._

### 1. AEs in Asia2. EMs in Asia

### 1. AEs in Asia2. EMs in Asia

### Aggregate effect of macroprudential tightening on household credit
- Sample and method:
  - Quarterly data over the period 2000–20.
  - Focus on 12 large regional AEs and EMs (aggregate results for a wider global sample of 42 countries reported in Annex 3).
  - Impulse response functions estimated; graphs show response and 90 and 75 percent confidence bands.
- Key finding:
  - A one-standard-deviation tightening of the aggregate macroprudential policy index is associated with 0.5–1 percentage point lower household credit growth.
  - Measures tend to affect credit growth gradually, with most of the impact having materialized after about five quarters.
- Composition of the aggregate macroprudential policy index (included tools):
  - Demand-side instruments: LTV and DSTI.
  - Tax measures.
  - Bank-capital-based measures: leverage ratios, countercyclical buffers, conservation buffers, and capital requirements.
  - Supply-affecting measures: reserve requirements, liquidity requirements, limits to FX positions, limits to the loan-to-deposit ratio, and limits to foreign currency loans.

### Asymmetry: tightening versus loosening
- Tightening:
  - MPM tightening has a statistically and economically significant effect on household credit (see panel 1, Figure 20).
- Loosening:
  - The impact of MPM loosening on household credit is smaller, in the right direction, but not statistically significant (see panel 2, Figure 20).
- Interpretation:
  - Results should be interpreted as the effect of the initial tightening at date 0; the specification controls for additional measures and macroeconomic developments in the interim.

### Relative effectiveness of tools: demand-side vs capital-based
- Finding:
  - Demand-side tools such as LTV or DSTI restrictions appear to be more effective in Asia-Pacific than capital-based tools.
- Evidence:
  - Strengthening LTV restrictions has a significant impact on housing credit growth, whereas tightening sectoral capital requirements does not show a similar significant effect (Figure 21).
- Mechanism:
  - Loan-level analysis suggests LTV and DSTI limits directly curb lending to higher-risk borrowers, while the link between capital requirements and lending decisions is less direct and hence weaker.
- Trade-offs:
  - These demand-side policies tend to worsen housing affordability by disproportionately limiting access to credit for lower income households, highlighting the need for designated policies to strengthen affordability.

### Impact on housing prices
- Overall effect:
  - The impact of MPMs on real housing prices is limited for the overall Asia-Pacific sample; no significant average effect detected (Figure 22).
  - In regional EMs there is a statistically significant effect, but it is not large and is relatively short-lived.
- Consistency with literature:
  - Findings align with literature showing macroprudential policy is effective at building resilience and reducing banking-sector risks, but has a more limited role in affecting the housing price cycle.

### Potential leakages and limits to effectiveness
- Sources of leakages:
  - Imperfect regulation and enforcement, regulatory arbitrage, and other circumvention schemes.
  - Growing presence of shadow banking and interconnectedness of banks and other financial intermediaries.
- Evidence on leakages:
  - Based on evidence from more than 6,000 estimates, Araujo and others (2020) find leakages and spillovers that reduce transmission of micro-level effects of macroprudential policy on bank lending.
- Policy implication:
  - Policymakers need to assess and address potential leakages ex ante to preserve MPM effectiveness.

*Source: IMF staff calculations and IMF Macroprudential Policy Survey (2020) as presented in the provided chapter excerpt.*

### 1. Asia: Housing Price Response to Macroprudential

### 1. Asia: Housing Price Response to Macroprudential Tightening

### Impact of Macroprudential Tightening on Housing Prices
- Impulse response functions estimated using a sample of 12 economies in Asia-Pacific using quarterly data over the period 2000–20.
- Graphical analysis shows responses with 90 and 75 percent confidence bands; x-axis shows quarters after the event, with t = 0 the quarter when the macroprudential tool is tightened.
- Finding: Limited impact of one-standard-deviation tightening of MPM on housing prices (percentage points), with only small and delayed effects detected.
- Leakages and spillovers:
  - Households can use consumer loans from nonbank financial institutions (NBFIs) to finance down payments to meet bank LTV limits (Cizel and others 2019).
  - Nonbanks and foreign providers can circumvent domestic macroprudential perimeters, especially in more developed or open financial systems (Cerutti, Dagher, and Dell’Ariccia 2015).
  - Tightening capital requirements may encourage nonbanks to borrow from branches of foreign banks where local capital requirements do not apply (Buch and Goldberg 2016).
  - Tightening of lending standards (for example, LTV ratios for home purchase) tends not to precipitate the same nonbank behaviors because similar lending standards are applied across loan products in a country.
  - Tightening of reserve requirements yields inconclusive results due to identification challenges versus monetary policy changes (Reinhardt and Sowerbutts 2015).

- Policy implications for improving effectiveness:
  - Expand scope of application to include nonbanks and foreign providers of credit in domestic financial markets.
  - Pursue jurisdictional reciprocity agreements (BCBS 2010) and greater host control over foreign affiliates (CGFS 2012, IMF 2014) to address international spillovers.

### Policies to Promote Affordability — Overview
- Housing affordability and access to adequate housing in Asia-Pacific have been deteriorating over time due to disproportionate increases in housing prices (Chapter 3).
- High mortgage rates from monetary policy tightening worsen affordability and negatively affect housing construction.
- Urbanization trends (Figure 23) indicate continued concentration of urban population, reinforcing the need for multifaceted policy approaches.
- Policy categories: structural supply-side measures; demand-side policies to directly support households; financing measures.

### Structural Supply-Side Policies
- Rationale: Supply constraints, especially in urban areas where jobs are concentrated, push up housing prices and exacerbate inequality and intergenerational persistence of disadvantage.
- Key measures:
  - Reform land use regulations:
    - Remove zoning restrictions that limit higher-occupancy buildings or greenfield development.
    - Example: New Zealand amended its Resource Management Act in 2021 to allow higher-density housing construction without local government resource consents and established a Housing Acceleration Fund to finance enabling infrastructure.
  - Promote productive and efficient use of idle land:
    - Introduce taxation such as land value taxation (tax on underlying land value only) to encourage development and discourage sprawl (Dye and England 2010).
    - Regional examples: Philippines and Thailand.
    - Thailand’s Land and Building Tax (2020): maximum tax rate ranges from 0.15 percent for agricultural use to 1.2 percent for empty or unused land.
  - Provide adequate social and affordable housing:
    - Public sector should step up provision in cooperation with private sector and civil society.
    - Example: Australia’s National Housing and Homeless Agreement (NHHA) improved funding agreements in 2018, with Commonwealth funding to subnational governments to improve access for people experiencing or at risk of homelessness.
    - Singapore: about 80 percent of households reside in public housing built and managed by the Housing and Development Board (HDB). HDB receives subsidies of about 0.25 percent of GDP (2017).
    - Policy tools in other countries: mandated shares of affordable/social housing in new private projects (Malaysia, Vietnam).

### Tax and Financing Incentives (Supply-side complements)
- Incentives to private developers: low-interest loans, subsidies, tax credits to reduce up-front costs and stimulate affordable housing construction (ADB 2019).
- Example: Malaysia’s National Affordable Housing Policy (NAHP) implemented in 2019 provides incentives and tax exemption for affordable housing construction.

### Urban Planning
- Map well-planned urbanization to ensure proximity to jobs, safe water and sanitation, access to transportation, education, and health services to support integrated supply-side policies.

### Demand-Side Policies
- Temporary, targeted measures can improve affordability for low-income households or first-time home buyers while supply-side measures take effect.
- Options:
  - Targeted government support: vouchers, housing subsidies, grants, tax relief, or accommodative financing terms (examples: Australia, Malaysia, New Zealand, Thailand).
  - Use progressive taxation on property purchases (stamp duties) to slow property price inflation; examples in Asia: Hong Kong SAR, Malaysia, Singapore.
  - Transform stamp duties to recurrent property taxes in some contexts to lower up-front purchase costs (example: New South Wales, Australia offering option to pay recurrent annual land taxation in lieu of stamp duty).
  - Apply progressive property tax rate structures so higher-income households bear more of the tax burden.

### Macroprudential Policy Design and Targeting
- Macroprudential tools (LTV, DSTI) can address overheating but may hurt affordability through tighter maximum loan sizes.
- Recommendations:
  - Target macroprudential measures to limit negative affordability impacts on vulnerable groups, e.g., tighten LTV only for borrower types posing significant financial risks.
  - Use measures addressing investment/speculative purchases: high capital gains taxes on properties not occupied and resold quickly (ADB 2019); property holding taxes (Korea).
  - Recognize limited effectiveness of macroprudential measures in dampening housing price growth overall; measures are more effective at constraining household credit growth and are particularly effective for borrowers with high-risk profiles.
  - Refine tools to limit leakages and cross-border spillovers; expand application to non-banks and foreign credit providers.

### Financing Policy and Supportive Mechanisms
- Limited availability of formal housing finance exacerbates unaffordability for low- and middle-income households.
- Supportive financing mechanisms:
  - Government housing finance agencies provide affordable loans and/or mortgage insurance and develop domestic mortgage bond markets; agencies also finance developers (examples: Cagamas Berhad in Malaysia; Housing Development Board in Singapore; Government Housing Bank in Thailand).
  - Community-led housing finance example: Philippines’ Community Mortgage Program (CMP) via the National Home Mortgage Finance Corporation.
  - Mandatory national saving vehicles:
    - Singapore’s Central Provident Fund (CPF) provides tax exemption on interest earnings and can be used for home purchases.
    - CPF contribution rules: employees contribute 20 percent of monthly salary, while employers contribute a further 17 percent. CPF has helped nearly 95 percent of employees aged 21 and older to own public housing.
    - Similar mandatory schemes exist in China, Indonesia, and the Philippines.
    - Indonesia’s Tapera (launched 2020): employees contribute 2.5 percent of monthly salaries, matched by a 0.5 percent contribution by employers.
  - Mortgage insurance or guarantees provided in China, Hong Kong SAR, India, Malaysia, Philippines, Thailand.
- Example of targeted macroprudential change affecting affordability: 2019 Bank of Thailand (BOT) lowered maximum LTV ratio to 80 percent for borrowers seeking residences valued above THB10 million (USD315,000); BOT also set maximum LTV ratios of 80-90 percent for second mortgages and 70 percent for subsequent mortgages.

### Conclusion and Policy Priorities
- Regional context:
  - Housing markets in Asia-Pacific have been volatile with long, sustained price cycles; post-pandemic surges may be at risk of reversing amid slowing growth and rising interest rates.
  - Housing prices appear misaligned especially among the region’s AEs, increasing risk of price corrections.
  - Key drivers of housing price-at-risk: magnitude of price misalignment and borrowing costs.
  - Model-based analysis: high price misalignment combined with high policy rates can lead to sizable price corrections, nearly comparable to past housing bust episodes.
  - High household debt can amplify effects of housing corrections on the real economy and financial stability.
- Multipronged policy approach emphasized:
  - Macroprudential policies are foremost to maintain financial stability:
    - Well-designed measures build resilience during booms and limit systemic risk during downturns.
    - Macroprudential policies limit household credit growth, though full effects can take more than a year to materialize.
    - Targeted LTV and DSTI restrictions are particularly effective for high-risk borrowers.
    - Limited effectiveness in dampening housing price growth overall; more significant (yet small) effects found in regional EM sample.
    - Need to refine tools to limit leakages and cross-border spillovers; expand scope to non-banks and foreign credit providers.
  - Implementation must be well calibrated:
    - Primary role is to build resilience during booms; relaxation during downturns may not be symmetric in effect.
    - Maintain reasonable minimum standards during downturns to guard against systemic risk build-up.
    - Coordinate macroprudential policy with monetary and fiscal policy, considering high household debt, tightening financial conditions, deteriorating real wage growth, and pandemic scarring.
  - Address housing affordability:
    - Structural supply-side policies are central for long-term restoration of affordability: reform land use regulations, promote productive use of idle land, provide tax incentives to builders, and expand social and affordable housing provision.
    - Short- to medium-term demand-side and financing measures can alleviate pressures: vouchers, grants, tax relief, low-interest loans, mortgage insurance or guarantees.
    - Demand-side interventions must be well-targeted (low-income households, aspiring first-time buyers) to avoid exacerbating imbalances.

*IMF Departmental Papers • Housing Stability and Affordability in Asia-Pacific*

### Box 1. Construction, Real Estate, and Housing Markets in Asia-Pacific

### Box 1. Construction, Real Estate, and Housing Markets in Asia-Pacific

### Role of construction and real estate in regional growth
- Construction and real estate are among the main drivers of economic growth in Asia-Pacific.
- Their significance has grown rapidly in both AEs (Australia, Hong Kong SAR, and New Zealand) and EMs (China, Indonesia, and Malaysia).
- EMs in the region, starting from a low base, have experienced rapid growth in the value added of construction and real estate services over the past decades.
- In some regional economies, construction and real estate have been growing broadly in line with the overall economy; exceptions noted include:
  - India: housing-related activities slowed significantly during 2010s.
  - Vietnam: manufacturing has been growing more rapidly than housing.

### Home ownership and affordability
- Activities in construction and real estate have further room to grow to improve home ownership in Asia-Pacific, which remains relatively low in some countries compared to the home-ownership rate of 70 percent in the EU.
- Young generations and low- and middle-income households are increasingly priced out of housing markets.
- Example: Korea has experienced a sharp fall in home ownership among people aged 40 or younger.
- Generational differences in home ownership are increasingly apparent (Japan, Korea, New Zealand), directly contributing to a generational wealth gap.
- The COVID-19 pandemic may have widened the generational divide to the disadvantage of younger generations that have not yet built a career with income security needed to finance a house purchase (Thailand).

### Household balance sheets and mortgage trends
- Housing loans constitute the largest component of household liabilities.
- Households in Australia, Hong Kong SAR, Korea, Malaysia, and Thailand have seen rapid growth in mortgage debt over the past decade.

### Emerging trends and drivers shaping housing markets
- Major trends influencing housing markets include:
  - Urbanization of middle- and low-income countries across Asia-Pacific.
  - Provision of greater incentives for homeownership since the Asian financial crisis.
  - Urban expansion and renewed availability of credit reshaping housing markets, particularly in capitals and other large cities of China, Indonesia, and elsewhere.
- Demographics influence long-term macroeconomic trends (such as labor supply and potential growth) and alter supply and demand for housing.
- Legal and institutional factors, such as property rights, financial deepening, and prevalence of corruption, have significant effects on transaction volumes and housing prices in a set of nine Asia-Pacific economies (Australia, China, Hong Kong SAR, Korea, Malaysia, New Zealand, Philippines, Singapore, and Thailand).
- Other policy settings have affected market structure:
  - Preferential capital gains tax treatments or differentiated rental laws have encouraged large rental markets by homeowners (example: New Zealand).
  - Foreign demand for housing has mattered in economies such as Australia, Hong Kong SAR, New Zealand, and Singapore, sometimes prompting policy measures differentiating by residency of homeowners.
- Cyclical factors (financial conditions) and structural factors (financial deepening and access to financing) have shaped housing markets and at times fueled housing booms, with greater differentiation between AEs and EMs.

### Policy implications and considerations
- Policies to increase affordable home ownership could leverage room for growth in construction and real estate activities.
- Addressing affordability and intergenerational disparities requires attention to supply-side expansions, credit access, and targeted measures for young and low- and middle-income households.
- Legal, institutional, and macrofinancial settings (including taxes, rental laws, and foreign buyer measures) materially influence housing market outcomes and should be considered in policy design.

*IMF DEPARTMENTAL PAPERS • Housing Stability and Affordability in Asia-Pacific — Box 1*

### Box 5. Housing and Household Debt in Emerging Market Economies

### Box 5. Housing and Household Debt in Emerging Market Economies

### Recent trends in household debt
- Household debt has increased in emerging market (EM) economies such as China, Malaysia, and Thailand.
- Household debt has also risen in a number of advanced economies (AEs), including Hong Kong SAR and Korea.
- While Asia-Pacific AEs traditionally had higher household debt due to more developed housing and financial markets, Asia-Pacific EMs have experienced growing household debt as efforts to improve households’ access to financial services for housing purchases pay off.

### Composition of household debt and mortgage uptake
- Mortgages are generally the bulk of a household’s debt in most AEs; in EMs, including in Asia-Pacific, mortgages are typically a smaller share of household liabilities.
- EM households also finance housing purchases via non-mortgage borrowing (for example, unsecured credit), and borrow for durable goods (such as autos) or to fund SME activities.
- Using survey-based data for Bangladesh, China, India, the Philippines, Thailand, and South Africa, Badarinza, Balasubramaniam, and Ramadorai (2019) find that the largest share of household wealth in these economies is stored in physical assets—primarily real estate—but they are not broadly collateralized against mortgage debt (except in China).

### Reasons for relatively smaller mortgage prevalence in EMs
- Level of financial sector development and capacity to intermediate appropriately long-term credit.
- Legal framework and capacity of the legal system to provide clear processes for protection of lenders (limited enforceability) and borrowers (recourse).
- Limited existence of credit bureau information systems (including for collateral verification).
- Fewer government policies including subsidy schemes and incentives (such as mortgage interest deductibility).

### Benefits and risks of deeper mortgage markets
- Benefits:
  - Improved saving and investment opportunities for households.
  - Welfare creation that can extend to greater community investment, higher school attainment, lower crime, and higher social capital (see Cerutti, Dagher, and Dell’Ariccia 2015).
- Risks and downsides:
  - Unproductive use of housing debt (investment or speculation) and an unsustainable pace of credit expansion can create macroeconomic and macro-financial vulnerabilities.
  - Reduced labor mobility as a consequence of higher homeownership.
  - Fiscal and structural inefficiencies from tax preferences such as mortgage interest deductibility, which is not well targeted and tends to amplify wealth inequality.
- Policy implication: Financial deepening is desirable but should be pursued in ways that avoid potentially costly pitfalls.

### Empirical analysis and methods referenced
- Housing price decomposition uses a structural vector autoregression with sign restrictions to identify four structural shocks: housing supply shock, housing demand shock, mortgage rate shock, and expectations shock. The model uses quarterly data from:
  - 2005:Q1–2021:Q2 for Australia, Japan, New Zealand, and Singapore.
  - 2009:Q1–2021:Q2 for Korea.
- House-price-at-risk is estimated using panel quantile regression for 4- and 8-quarter-ahead real housing price percent changes. Main independent variables include:
  - Housing price misalignment measured using deviation from HP-filter trend in house-price-to-income ratio (HPI data from IMD SPR Risk Tracker).
  - Short-term interest rates from CEIC.
  - Output gap based on a linear interpolation from annual data from the World Economic Outlook.
  - Household credit to GDP from BIS, detrended using country-specific linear trends.
  - Lagged housing price growth from BIS.
- Main analysis samples:
  - Six AEs: Australia, Hong Kong SAR, Japan, Korea, New Zealand, Singapore.
  - Six EMs: India, China, Malaysia, Indonesia, Thailand, Philippines.
  - Broader world sample: 27 AEs and 15 EMs (for comparison).
- Key regression result patterns (from Annex Tables 2.1 and 2.2):
  - House price misalignment is negatively and significantly associated with future downside house-price risk (5th percentile) and median outcomes across samples and horizons (coefficients negative and often significant at *** p<0.01 or ** p<0.05).
  - Short-term interest rates and the output gap generally show negative and significant relationships with future downside house-price risk and median outcomes in many specifications.
  - Household credit deviation from trend has mixed significance across samples and horizons; results vary by group (AEs vs EMs) and horizon.
  - Current house price growth is positively and significantly associated with future house price outcomes across samples and horizons.

### Policy relevance and implications
- Improving access to mortgages in EMs can yield welfare benefits but requires strengthening financial intermediation, legal frameworks (enforceability and borrower/lender protections), credit information systems, and carefully designed fiscal incentives.
- Policy tools to limit speculative and investment-driven housing demand are relevant to contain financial stability risks.
- Macroprudential frameworks and targeted fiscal measures (for example, limits on LTV and DSTI, stamp duties, buyer’s stamp duties) can help manage housing cycles and credit dynamics.

### Illustrative policy case (Singapore)
- Singapore’s experience illustrates that soft landings are possible through well-targeted macroprudential and fiscal measures combined with supply-side policies.
- Timeline of measures and outcomes described:
  - 2010: Introduction of a holding period for the imposition of seller’s stamp duty (SSD).
  - 2011–12: SSD increased to limit speculative demand; complemented by lower LTV limits and an additional buyer’s stamp duty (ABSD).
  - 2012: Loan tenure cap introduced; long tenure loans became subject to lower LTV limits.
  - 2013: MAS lowered LTV limits further, raised ABSD rates, and introduced a total debt servicing ratio (TDSR) framework.
  - 2015–16: Large supply of new homes came on stream, helping to balance the market.
  - 2021:Q4: MAS tightened TDSR and LTV limits and raised ABSD rates; authorities continue to calibrate a comprehensive set of macroprudential measures (MPMs) and ramp up efforts to increase housing supply.
- Institutional and market context note: In Singapore, about 4 in 5 people live in public housing (“HDB flats”) developed by the Housing & Development Board (HDB). HDB flats are sold to individuals who meet eligibility criteria; occupancy is bound by a 99-year lease, and the flat can be sold on the private resale market under certain restrictions.
- Evidence cited indicates that the combination of credit and fiscal-based measures along with land use policies to increase housing supply proved effective in controlling transaction volumes, property prices, and mortgage loan quality.

*Italic: Source — IMF staff analysis from "Box 5. Housing and Household Debt in Emerging Market Economies" in the provided content.*

### Annex Figure 3.1. Impact of Macroprudential Policies—Global Sample

### Annex Figure 3.1. Impact of Macroprudential Policies—Global Sample

### Figure overview
- Subject: Household Credit Response to Macroprudential Tightening across four policy types for the global sample.
- Panels:
  - 1. All: Household Credit Response to Macroprudential Tightening
  - 2. All: Household Credit Response to LTV Tightening
  - 3. All: Household Credit Response to Capital Requirements
  - 4. All: Household Credit Response to Tax Measures for Macroprudential Purposes
- Note: LTV = loan-to-value.
- Source: IMF staff estimates.

### Panel 1 — All: Household Credit Response to Macroprudential Tightening
- Y-axis tick values and plotted numeric labels as presented:
  - –1.5
  - 1.0
  - –1.0
  - –0.5
  - 0.0
  - 0.5
- X-axis tick labels as presented:
  - 0 2 4 6 8 10 12 14 16

### Panel 2 — All: Household Credit Response to LTV Tightening
- Y-axis tick values and plotted numeric labels as presented:
  - –1.0
  - 0.4
  - –0.8
  - –0.6
  - –0.4
  - 0.0
  - –0.2
  - 0.2
- X-axis tick labels as presented:
  - 0 2 4 6 8 10 12 14 16

### Panel 3 — All: Household Credit Response to Capital Requirements
- Y-axis tick values and plotted numeric labels as presented:
  - –0.6
  - 0.8
  - –0.4
  - –0.2
  - 0.0
  - 0.4
  - 0.2
  - 0.6
- X-axis tick labels as presented:
  - 0 2 4 6 8 10 12 14 16

### Panel 4 — All: Household Credit Response to Tax Measures for Macroprudential Purposes
- Y-axis tick values and plotted numeric labels as presented:
  - –0.3
  - 0.7
  - –0.2
  - –0.1
  - 0.1
  - 0.5
  - 0.3
  - 0.0
  - 0.2
  - 0.4
  - 0.6
- X-axis tick labels as presented:
  - 0 2 4 6 8 10 12 14 16

*Source: IMF staff estimates. Note: LTV = loan-to-value.*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2022/english/hmsaapea.pdf_