## _wp12130 - 3. Selected Asia: Impact of Macroprudential Measures

## Source details

**Canonical URL:** [_wp12130 - 3. Selected Asia: Impact of Macroprudential Measures](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12130.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12130.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12130.pdf.json)

---

### I. Introduction and summary findings
- Net capital flows to emerging Asia rebounded at a record pace following the global financial crisis (GFC), raising concerns about overheating and financial stability.
- Key questions addressed:
  - How does the recent rebound compare with previous surges (pre-Asian crisis and pre-GFC)?
  - What has been the impact so far of the inflows on the broader economy and financial stability?
  - How do recent non-macroeconomic measures compare with those taken in previous episodes and what lessons do they provide?
- Structure of the chapter: Section II (size and composition of recent flows); Section III (broader economy, asset prices, corporate variables across episodes); Section IV (country experiences managing inflows); Section V (conclusion).
- Summary findings:
  - Net inflows in the recent episode have yet to reach previous peaks and there is significant variation across countries.
  - Unlike in previous episodes, there is little evidence of a significant buildup of imbalances and resource misallocation in the recent episode.
  - Measures taken mostly targeted specific types of portfolio flows and leverage of banks’ foreign currency exposure, rather than generalized controls on all flows.
  - Regulatory measures in response to capital inflows are not new to Asia; past surges also saw such measures, but there is no strong evidence of their efficacy without supporting monetary and fiscal policies.

### II. Capital flows across episodes: facts and stylized statistics
- Historical peaks and timing:
  - Global peak in net capital flows for the region: second quarter of 1996, at about 6¾ percent of aggregate regional GDP.
  - Similar peak nearly reached in the first quarter of 2004.
  - Recent episode peak: around 4¼ percent of GDP.
  - Speed of recovery post-GFC: within five quarters net inflows rose from a trough (early 2009) to a peak (mid-2010); pre-Asian crisis and pre-GFC periods saw about 25 quarters between troughs and peaks.
- Episodes and distribution:
  - Total of 32 surge instances over the last 20 years.
  - 2000s episode had the highest number of surges: 13.
  - Nine surges since the global financial crisis (recent episode).
  - NIEs (Hong Kong SAR, Korea, Singapore, Taiwan Province of China) and India experienced surges recently; only China had an ongoing surge as of mid-2011.
  - ASEAN-5: only the Philippines and Vietnam qualify as having experienced a surge during the recent episode.
- Size and composition:
  - Surges during the 1990s: weighted average net capital inflow of 5 percent of GDP.
  - Surges during the pre-global financial crisis period and the recent rebound: weighted average inflow of about 4 percent of GDP.
  - More than half of total inflows to Asia now are bank-related and other private flows (relatively risky), versus about 20 percent during 1987–98 and about half during 1999–08.
  - ASEAN-5: shift toward sizable portfolio debt flows; banking flows and FDI fell since mid-1990s.
- Risk of sudden stops:
  - Over 60 percent of surges in Emerging Asia have ended in sudden stops (percentage varies across subperiods).

### II.B Volatility and financial depth
- Volatility:
  - Volatility of net capital flows has generally increased across time for all types of flows and regional groupings.
  - Growing gross inflows and outflows associated with financial globalization are likely drivers.
- Foreign holdings of local government bonds:
  - Foreign ownership rose considerably since early 2000s for four countries most associated with the Asia crisis.
  - Foreign holdings reached around 30 percent (Indonesia) and around 20 percent (Malaysia).
- Inflows relative to financial depth:
  - Indonesia received the largest amount of inflows relative to the size of its financial markets; Korea the least.
  - Large inflows in shallow markets raise concerns about rapid bubble formation and destabilizing sudden withdrawals.

### III. Broader economic and asset valuation patterns across episodes
- Macroeconomic dynamics:
  - Growth did not collapse after the GFC as it did after the Asian crisis.
  - Increase in growth rate was more marked during pre-Asian crisis surges than subsequent ones.
  - Current account balances were healthier in the run up to the GFC (surpluses rather than sizable deficits before the Asian crisis).
  - Reserve levels significantly higher and REERs more depreciated on average before the GFC.
  - During recent surges, both short-term and long-term interest rates fell; pre-Asian crisis surges saw rates tend to increase.
  - External buffers larger now in terms of current account surpluses and reserves levels.
- Asset markets and valuations:
  - Exchange market pressure (EMP): region historically intervened and accumulated reserves; recent surge EMP generally did not reach pre-GFC peaks except in Hong Kong SAR and Korea.
  - Real effective exchange rates: despite appreciations in 2010, REERs are significantly below historical peaks in most economies.
  - Equities: forward-looking price-earnings ratios reached significantly higher peaks during previous episodes, particularly pre-Asian crisis.
  - Bonds: 10-year sovereign bond spreads were wider in 2011:Q3 than low points reached before the GFC and at the eve of the crisis (2007:Q4).
  - Real estate: strong signs of overheating pre-Asian crisis by house price-to-rent and price-to-income indicators (possible exception: Indonesia); fewer signs before the GFC except in Taiwan Province of China (price-to-rent) and China, India, Indonesia, Taiwan Province of China (price-to-income). As of 2011:Q2, price-to-rent ratios appear relatively strong only in China, Hong Kong SAR, and Malaysia.
  - Credit growth: excessive credit expansion in most countries during 1990s inflow episodes; fewer signs before the GFC; more recently credit dynamics particularly strong in China, Hong Kong SAR, Malaysia, and Vietnam.
- Corporate sector:
  - Firms deleveraged markedly since the Asian crisis; corporate debt-to-equity ratios improved.
  - Little duration risk in accumulated debt; only Korea in orange zone for duration risk.
  - Firm-level disaggregation (large vs small, export- vs domestic-oriented) does not materially change the aggregate picture.
- Overall assessment:
  - External buffers larger and asset valuations more in line with fundamentals now than during previous surges.
  - Asian economies are at an early stage of the capital flow cycle; equity booms in the 1990s took at least ten quarters to peak from the start of a surge.
  - Few signs of recent overheating more likely reflect domestic imbalances than capital inflows.
  - Volatility concerns remain; shallow markets can allow asset price bubbles to form quickly and sudden stops remain a real possibility.

### IV. Macroprudential and other capital flow measures: objectives, tools, and country experiences
- Objectives of recent measures (five broad objectives observed):
  - (i) mitigate complications for central bank market operations stemming from inflows to short-term instruments;
  - (ii) limit inflows to local bond markets;
  - (iii) reduce risks in banking system and real economy;
  - (iv) limit vulnerabilities from private sector external borrowing;
  - (v) reduce currency speculation.
- Policy toolkit and guidance:
  - Macroprudential, structural, and capital flow management measures are part of the toolkit.
  - Macroprudential and structural measures to strengthen capacity to absorb inflows are encouraged.
  - Appropriate macroeconomic policies should be implemented when faced with surging inflows.
  - Capital flow management measures (administrative, tax, prudential) can address specific macroeconomic and financial stability risks under certain circumstances but are not substitutes for macroeconomic policy adjustment.
  - Measures that do not discriminate by residency are generally preferable (IMF, 2011c).
- Selected country cases and measures:
  - Indonesia:
    - Problem: foreign holdings of central bank bills reached almost a third of outstanding stock by mid-2010, complicating sterilization and attracting more inflows.
    - Measures: 2010 introduced a one month holding period for central bank bill purchases and phased out shorter maturity bills; May 2011 extended minimum holding period for SBIs from one month to six months and introduced longer tenors for SBIs.
    - Effects: foreign holdings of central bank bills and government bonds increased in absolute terms and as proportion of total outstanding; foreign ownership of SBIs as proportion of outstanding liquidity-absorbing instruments declined as non-tradable term deposits became a more important sterilization tool.
  - Thailand:
    - Faced historically high portfolio flows in 2010:Q3, mostly into sovereign debt.
    - Measures: reinstated a 15 percent withholding tax on nonresidents’ interest earnings and capital gains on new purchases of state bonds in October 2010.
    - Effects: inflows fell immediately but resumed by December 2010; limited effect attributed to double taxation agreements and uncertainty over operational aspects of the tax.
  - Korea:
    - Re-introduced a withholding tax on foreign purchases of state bonds in January 2011.
  - Taiwan Province of China:
    - Measures: limited local banks’ capacity to provide liquidity to NDF markets; discouraged nonresident deposits by imposing punitive reserve requirements.
    - Outcome: volatility rose in late 2010 and early 2011; FX market turnover and inbound remittances for investment purposes remained high.
  - Real estate measures:
    - Jurisdictions that introduced or tightened real estate lending criteria in 2009–11: China, Hong Kong SAR, India, Korea, Malaysia, Singapore, Taiwan Province of China, Thailand.
    - Hong Kong SAR and Singapore raised sellers’ stamp duty substantially for short-held properties; Singapore imposed an additional buyer’s stamp duty targeted at foreign and corporate buyers.

### IV.B Comparison with past measures and evidence on effectiveness
- Comparison with past measures:
  - Recent measures generally less heavy-handed compared with past (no repetition of Thailand’s unremunerated reserve requirement of 30 percent on non-FDI inflows introduced in late 2006).
  - Objectives remain similar, with more focus this time on managing flows into sovereign debt.
  - Past measures in the 1990s targeted offshore commercial borrowing and bank activities (Indonesia 1991, Malaysia 1992 and 1994, Thailand 1995); short-term inflows (Malaysia 1994, Thailand 1995); and currency speculation in offshore markets (Thailand 1997, Malaysia 1998).
- Event-study evidence and quantitative impacts:
  - Macroprudential policies are associated with a reduction in net private capital inflows of 1.9 percentage points of GDP (relative to the preexisting capital flow volumes during the four quarters before the policy).
  - Impact on flow components:
    - Portfolio investment: reduction of 1.2 percentage points of GDP.
    - Bank loans/other investment: reduction of 1.7 percentage points of GDP.
  - FX-related macroprudential policies tend to lower net inflows by 1.3 percentage points of GDP; large variation implies FX-related measures are as likely to decrease inflows as to increase them.
  - Full sample 1996–2011 (impacts reported with most-effective quarter in brackets):
    - Net private capital inflows: -1.3(4) then -1.9(5)
    - Portfolio investment: -0.6(1) then -1.2(4)
    - Bank loans/other investment: -1.8(4) then -1.7(5)
    - Exchange Market Pressure Index: -1.4(2)
    - Residential price/rent ratios: -11.5(4)
    - Residential price/income ratios: -2.0(5)
  - 2000–2011 subsample (where reported):
    - Net private capital inflows: -1.4(1) and 0.2(6)
    - Portfolio investment: 0.3(1) and -0.1(5)
    - Bank loans/other investment: -1.6(1) and 0.2(6)
    - Exchange Market Pressure Index: -1.4(2)
    - Residential price/rent ratios: 0.2(6)
    - Residential price/income ratios: -2.3(2)
  - Housing-market prudential measures:
    - Residential price/rent ratios: associated with an 11.5-point decline (statistically significant) in the full sample.
    - Residential price/income ratios: associated with a 2.0-point decline (statistically significant) in the full sample.
    - In the 2000–2011 subsample, price-to-income ratios response is similar magnitude and statistically significant (response taking 2 quarters); price-to-rent ratios did not move much in response to housing measures introduced after the 1990s.
    - Significance of housing results appears driven by property cooling measures introduced in Singapore in the late 1990s and early 2000s.
  - Methodology notes:
    - Event study window: four quarters prior and up to six quarters post policy announcement.
    - Measures classified into FX-related measures targeting foreign-currency liabilities; housing-market prudential measures; other prudential measures not targeting foreign liabilities.
    - Significance tested by averaging post-minus-pre announcement differences across country events and applying a one-sided t-test. Results reported in bold in source are significant at the 10 percent level.
  - Caveats:
    - Among policies introduced during 2000–2011, only FX-related measures had some significant impact on bank and other inflows; effects of other prudential measures on net flows mostly insignificant due to continued strong flows to India and Philippines in the run-up to the GFC.
    - The analysis notes these countries could have experienced an even larger surge without the policy response.

### V. Policy conclusions and recommendations
- Vigilance:
  - Policymakers should remain focused on potential risks to the real economy and financial stability from capital inflow surges, given significant variation across countries in magnitudes and types of inflows and isolated pockets of concern (credit dynamics in some countries and certain segments of property markets).
- Mixed toolkit:
  - A mixed policy toolkit will likely be required to manage large capital inflows. Macroprudential measures taken in the most recent surge have been appropriately narrowly targeted.
- Limitations of prudential measures:
  - Experience with past surges suggests prudential measures may not always suffice by themselves in preventing the buildup of risks.
- Continued role for traditional policy tools:
  - Monetary policy and fiscal policy remain essential in managing demand conditions to prevent risks to the real economy and financial stability from building up.
  - Given the size of external buffers, allowing the exchange rate to take more of the burden of exchange market pressure will help manage inflows and contribute to rebalancing global demand.

*Source: IMF staff analysis in _wp12130 (Selected excerpts).*

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

### _wp12130 - References

### Figures
- 1.   EmergingAsia: Net Private Capital Flows .......................................................................... 4
- 2.   Drivers of Inflows ............................................................................................................... 5
- 3.   Distribution of Capital Inflows Episodes ............................................................................ 6
- 4.   Emerging Asia: Net Private Capital Flows ......................................................................... 8
- 5.   Emerging Asia: Average Net Private Capital Flows during Episodes ............................... 9
- 6.   Emerging Asia: Volatility of Net Ptivate Capital Flows by Flow Type and Recipient 
            Economies      ................................................................................................................... 10
- 7.   Foreign Holdings of Local Government Bonds ................................................................ 10
- 8.   Cumulative Gross Capital Inflows in 2010:Q1–2011:Q2 to Financial Depth .................. 11
- 9.   Emerging Asia: Macroeconmics Dynamics around Surges .............................................. 12
- 10. Emerging Asia: Exchange Market Pressure Index ........................................................... 14
- 11. Real Effective Exchange Rate .......................................................................................... 15
- 12. Selected Asia: 10-Year Bond Spreads .............................................................................. 13
- 13. Indonesia and Thailand: Capital Flows ............................................................................. 21

### Tables
- 1.   Episodes of Large Net Private Capital Inflows––Summary Statistics ................................ 7
- 2a. Financial Indicators across Episodes of Large Net Capital Flows to Emerging Asia ...... 16
- 2b. Corporate Indicators across Episodes of Large Net Capital Flows to Emerging Asia ..... 17

*Source: _wp12130 - References.*

### 3.   Selected Asia: Impact of Macroprudential Measures ....................................................... 21

### _wp12130 - 3.   Selected Asia: Impact of Macroprudential Measures ....................................................... 21

### I. INTRODUCTION
- Net capital flows to emerging Asia rebounded at a record pace following the global financial crisis (GFC), raising concerns about overheating and financial stability.
- Key questions posed:
  - How does the recent rebound compare with previous surges, in particular the episodes before the Asian crisis and the GFC?
  - What has been the impact so far of the inflows on the broader economy and financial stability?
  - How do recent non-macroeconomic measures compare with those taken in previous episodes and what lessons do they provide?
- Structure of the paper:
  - Section II: size and composition of recent flows to Asian emerging markets from a historical perspective.
  - Section III: compares developments in the broader economy, asset prices, and corporate variables across episodes of strong inflows.
  - Section IV: reviews country experiences in managing risks associated with inflows.
  - Section V: concludes.
- Summary findings:
  - Net inflows in the recent episode have yet to reach previous peaks and there is significant variation across countries.
  - Unlike in previous episodes, there is little evidence of a significant buildup of imbalances and resource misallocation in the recent episode.
  - Measures taken mostly targeted specific types of portfolio flows and leverage of banks’ foreign currency exposure, rather than generalized controls on all flows.
  - Regulatory measures in response to capital inflows are not new to Asia; past surges also saw such measures, but there is no strong evidence of their efficacy without supporting monetary and fiscal policies.

### II. CAPITAL FLOWS ACROSS EPISODES: THE FACTS
- Historical peaks and recent episode:
  - Global peak in net capital flows for the region: second quarter of 1996, at about 6¾ percent of aggregate regional GDP.
  - Similar peak nearly reached in the first quarter of 2004.
  - Recent episode peak: around 4¼ percent of GDP.
  - Speed of recovery post-GFC: within five quarters net inflows rose from a trough (early 2009) to a peak (mid-2010); pre-Asian crisis and pre-GFC periods saw about 25 quarters between troughs and peaks.
- Drivers:
  - Growth and interest rate differentials do not appear to explain much of the recent inflows.
  - Some countries show stronger relationships between bond flows and interest rate differentials (Indonesia, Korea, the Philippines, and Thailand).
  - Global factors—including U.S. interest rates and risk aversion—are important determinants of surges (cited support: Ghosh and others (2012)).
- Surge identification methodology (summary):
  - Focus on net private capital flows excluding official flows to general government and monetary authorities (other investment component).
  - A “surge” defined broadly following IMF (2007): period of two or more quarters where flows (share of GDP) are significantly larger (1 standard deviation) than historical trend or above the 75th percentile of distribution.
  - Trend measured by eight quarter moving average. Supplementary rules applied to combine adjacent surges and adjust for errors and omissions (Vietnam adjustment noted).
- Main findings on episodes:
  - Distribution:
    - Total of 32 surge instances over the last 20 years.
    - 2000s episode had the highest number of surges: 13.
    - Nine surges since the global financial crisis (recent episode).
    - NIEs (Hong Kong SAR, Korea, Singapore, and Taiwan Province of China) and India experienced surges recently, most ending before end-2010. Only China had an ongoing surge as of mid-2011.
    - For ASEAN-5, only the Philippines and Vietnam qualify as having experienced a surge during the recent episode.
  - Size:
    - Surges during the 1990s were the biggest for emerging Asia: weighted average net capital inflow of 5 percent of GDP.
    - Surges during the pre-global financial crisis period and the recent rebound: weighted average inflow of about 4 percent of GDP.
  - Composition and country variation:
    - More than half of total inflows to Asia now are bank-related and other private flows (relatively risky), versus about 20 percent during 1987-98 and about half during 1999-08.
    - ASEAN-5: banking flows and FDI fell since mid-1990s in favor of sizable portfolio debt flows.
    - Pre-Asian crisis surges in ASEAN economies were far bigger than subsequent surges.
    - India: net capital flows on secular rise since 1991 liberalization, but declined in 2011; recent surge driven by portfolio equity inflows.
    - China: surges gradually decreased over time, composition shifted from FDI to banking flows.
  - NIEs and India specifics:
    - NIEs average inflow in recent surge: almost 8 percent of GDP (versus about 5½ percent during 1990s and 2000s).
    - India recent surge average: nearly 5 percent of GDP (portfolio flows rose dramatically).
    - Korea: unprecedented surge faded as large portfolio inflows were offset by FDI and banking-related outflows.
  - Risk of sudden stops:
    - Over 60 percent of surges in Emerging Asia have ended in sudden stops (Table 1: percentage of episodes that ended in a sudden stop varies across subperiods).

### II.B Volatility and Financial Depth
- Volatility:
  - Volatility of net capital flows has generally increased across time for all types of flows and regional groupings.
  - This likely reflects increasing gross inflows and outflows associated with financial globalization.
  - Broner and others (2011) show volatility of net flows increased less than gross flows due to positive correlation between gross inflows and outflows.
- Foreign holdings of local government bonds:
  - Foreign ownership has risen considerably since early 2000s for four countries most associated with the Asia crisis.
  - Foreign holdings reached around 30 percent (Indonesia) and around 20 percent (Malaysia).
  - Combined with potential U.S. interest rate increases, this raises vulnerability to sudden withdrawal of foreign capital in some countries.
- Inflows relative to financial depth:
  - Indonesia received the largest amount of inflows relative to the size of its financial markets; Korea the least.
  - Presence of large inflows in relatively shallow markets raises concerns about rapid bubble formation and destabilizing sudden withdrawals.

### III. BROADER ECONOMIC AND ASSET VALUATION PATTERNS ACROSS TIME
- Macroeconomic dynamics around surges:
  - Growth did not collapse after the GFC as it did after the Asian crisis.
  - Increase in growth rate was more marked during pre-Asian crisis surges than subsequent ones.
  - Current account balances were healthier in the run up to the GFC (surpluses rather than sizable deficits before the Asian crisis).
  - Reserve levels significantly higher and REERs more depreciated on average before the GFC.
  - During recent surges, both short-term and long-term interest rates fell; pre-Asian crisis surges saw rates tend to increase.
  - External buffers larger now in terms of current account surpluses and reserves levels.
- Asset market indicators and imbalances:
  - Exchange markets:
    - Region historically intervened and accumulated reserves in response to exchange market pressure (EMP).
    - During the recent surge, EMP generally did not reach pre-GFC peaks except in Hong Kong SAR and Korea.
    - Despite appreciations in 2010, real effective exchange rates are significantly below historical peaks in most economies.
  - Equities and bonds:
    - Equity valuations (forward-looking price-earnings ratios) reached significantly higher peaks during previous episodes, particularly pre-Asian crisis.
    - Bond spreads: 10-year sovereign bond spreads were wider in 2011:Q3 than low points reached before the GFC and at the eve of the crisis (2007:Q4).
  - Real estate:
    - Strong signs of overheating pre-Asian crisis by house price-to-rent and price-to-income indicators (possible exception: Indonesia).
    - Fewer signs before the GFC except: price-to-rent in Taiwan Province of China; price-to-income in China, India, Indonesia, Taiwan Province of China.
    - As of 2011:Q2, price-to-rent ratios appear relatively strong only in China, Hong Kong SAR, and Malaysia.
  - Credit growth:
    - Excessive credit expansion in most countries during 1990s inflow episodes.
    - Fewer signs before the GFC; more recently credit dynamics particularly strong in China, Hong Kong SAR, Malaysia, and Vietnam.
    - Elekdag and Wu (2011): credit booms tightly interconnected with large capital inflows; bank and other flows increase during build-up phase, particularly for booms followed by credit busts.
  - Corporate sector:
    - Firms deleveraged markedly since the Asian crisis; corporate debt-to-equity ratios were high in the past but improved.
    - Little duration risk in accumulated debt; only Korea in orange zone for duration risk.
    - Disaggregated firm-level indicators (large vs small, export-oriented vs domestic-oriented) do not change the overall picture significantly.
- Overall assessment:
  - External buffers larger and asset valuations more in line with fundamentals now than during previous surges.
  - Asian economies are at an early stage of the capital flow cycle; equity booms in the 1990s took at least ten quarters to peak from the start of a surge.
  - Few signs of recent overheating more likely reflect domestic imbalances than capital inflows.
  - Volatility concerns remain; shallow markets can allow asset price bubbles to form quickly and sudden stops remain a real possibility.

### IV. MACROPRUDENTIAL AND OTHER CAPITAL FLOW MEASURES
- Objectives of recent measures (five broad objectives observed):
  - (i) mitigate complications for central bank market operations stemming from inflows to short-term instruments;
  - (ii) limit inflows to local bond markets;
  - (iii) reduce risks in banking system and real economy;
  - (iv) limit vulnerabilities from private sector external borrowing;
  - (v) reduce currency speculation.
- Policy toolkit:
  - Macroprudential, structural, and capital flow management measures are all part of the toolkit.
  - Macroprudential and structural measures to strengthen capacity to absorb inflows are encouraged.
  - Appropriate macroeconomic policies should be implemented when faced with surging inflows.
  - Capital flow management measures include administrative, tax, and prudential measures to influence capital flows; they can address specific macroeconomic and financial stability risks under certain circumstances but are not substitutes for macroeconomic policy adjustment.
  - Measures that do not discriminate by residency are generally preferable (IMF, 2011c).
- Selected country cases and measures:
  - Indonesia:
    - Problem: foreign holdings of central bank bills reached almost a third of outstanding stock by mid-2010, complicating sterilization and attracting more inflows.
    - Causes: large interest rate differentials; arbitrage opportunities between offshore and onshore short-term local currency interest rates.
    - Measures: 2010 introduced a one month holding period for central bank bill purchases and phased out shorter maturity bills; May 2011 extended minimum holding period for SBIs from one month to six months and introduced longer tenors for SBIs.
    - Effects: Since measures, foreign holdings of central bank bills and government bonds have increased in absolute terms and as proportion of total outstanding; foreign ownership of SBIs as proportion of outstanding liquidity-absorbing instruments declined as non-tradable term deposits became a more important sterilization tool.
  - Thailand:
    - Faced historically high portfolio flows in 2010:Q3, mostly into sovereign debt.
    - Measures: reinstated a 15 percent withholding tax on nonresidents’ interest earnings and capital gains on new purchases of state bonds in October 2010.
    - Effects: inflows fell immediately but resumed by December 2010. Limited effect attributed to double taxation agreements and uncertainty over operational aspects of the tax which dampened both debt and equity inflows.
  - Korea:
    - Re-introduced a withholding tax on foreign purchases of state bonds in January 2011 (similar to Thailand).
  - Taiwan Province of China:
    - Concerned that onshore and NDF trading driven by speculators caused excessive exchange rate volatility.
    - Measures: limited local banks’ capacity to provide liquidity to NDF markets; discouraged nonresident deposits by imposing punitive reserve requirements.
    - Outcome: volatility rose in late 2010 and early 2011; FX market turnover and inbound remittances for investment purposes remained high.
  - Real estate measures:
    - Jurisdictions that introduced or tightened real estate lending criteria in 2009–11: China, Hong Kong SAR, India, Korea, Malaysia, Singapore, Taiwan Province of China, Thailand.
    - Hong Kong SAR and Singapore raised sellers’ stamp duty substantially for short-held properties; Singapore imposed an additional buyer’s stamp duty targeted at foreign and corporate buyers.
- Comparison with past measures:
  - Recent measures generally less heavy-handed compared with past (e.g., no repetition of Thailand’s unremunerated reserve requirement of 30 percent on non-FDI inflows introduced in late 2006).
  - Objectives of interventions remain similar, with more focus this time on managing flows into sovereign debt.
  - Past measures in the 1990s targeted: offshore commercial borrowing and bank activities (Indonesia 1991, Malaysia 1992 and 1994, Thailand 1995); short-term inflows (Malaysia 1994, Thailand 1995); and currency speculation in offshore markets (Thailand 1997, Malaysia 1998).
  - Different approach reflects earlier stage of recent inflow cycle and policymakers’ learning from past negative repercussions (e.g., 2006 Thai equity market impact).
- Evidence on effectiveness (event study summary and findings):
  - Dataset and classification:
    - Measures identified using IMF AREAER for China, Hong Kong SAR, India, Indonesia, Korea, Malaysia, Philippines, Singapore, and Thailand during 1999-2009; 1995-1998 measures collected from printed AREAER reports for epicenter countries.
    - Focus on measures similar to those taken recently (not outright residency-discriminating capital controls).
    - Measures classified into: FX-related measures targeting foreign-currency liabilities; housing-market prudential measures; other prudential measures not targeting foreign liabilities.
  - Methodology:
    - Event study over a short period: four quarters prior to and up to six quarters post policy announcement.
    - Impact tested on net private capital inflows, portfolio investment, bank loans/other investment, exchange market pressure index, residential price-to-rent ratios, and price-to-income ratios.
    - Significance tested by averaging post-minus-pre announcement differences across country events and applying a one-sided t-test.
  - Result summary:
    - FX-related and other prudential measures generally associated with some moderation in net capital inflows (top panel of Table 3 in source).
    - Full-sample event study results suggest macroprudential policies intended for limiting buildup of systematic risk in the financial sector without targeting foreign liabilities are most likely to stem net private capital inflows in a statistically significant manner.

*Source: _wp12130 - 3.   Selected Asia: Impact of Macroprudential Measures (IMF working paper chapter content provided).*

### 1.9 percentage points of GDP (relative to the preexisting capital flow volumes during the four

### _wp12130 - 1.9 percentage points of GDP (relative to the preexisting capital flow volumes during the four

### Impact of macroprudential measures on capital inflows
- Macroprudential policies are associated with a reduction in net private capital inflows of 1.9 percentage points of GDP (relative to the preexisting capital flow volumes during the four quarters before the policy was implemented).
- Impact on portfolio and banking flows:
  - Portfolio investment: reduction of 1.2 percentage points of GDP.
  - Bank loans/other investment: reduction of 1.7 percentage points of GDP.
- FX-related macroprudential policies tend to lower net inflows by only 1.3 percentage points of GDP.
- Large variations in responses imply FX-related measures are as likely to decrease inflows as they are to increase them.
- In the 2000-2011 subsample:
  - Among policies introduced during 2000-2011, only FX-related measures had some significant impact on bank and other inflows.
  - The overall impact of other prudential measures on net overall flows as well as portfolio and banking flows was mostly insignificant, largely due to strong flows to India and Philippines in the run-up to the global financial crisis where flows continued to increase regardless of implemented measures.
  - The analysis notes these countries could have experienced an even larger surge in capital inflows in the absence of the policy response (IMF, 2010a).

### Impact on asset and exchange market measures
- Exchange Market Pressure Index:
  - Reported impact: -1.4 (number in brackets indicates the quarter in which the policy is most effective: (2)).
- Housing market prudential measures and asset valuation:
  - Residential price/rent ratios: associated with an 11.5-point decline (statistically significant).
  - Residential price/income ratios: associated with a 2.0-point decline (statistically significant).
  - In the 2000-2011 subsample:
    - Price-to-income ratios response is similar magnitude and statistically significant, with the response taking 2 quarters (as opposed to 5 quarters in the full sample).
    - Price-to-rent ratios did not seem to move much in response to housing measures introduced after the 1990s.
- Note on drivers of housing results:
  - Further analysis suggests significance of these results is driven by property cooling measures introduced in Singapore in the late 1990s and early 2000s (lowering LTV ceilings, hiking stamp duties and capital gains taxes, tightening access to public housing). Given Singapore’s unique nature (a city state with a large share of public housing), applicability to other countries is uncertain.

### Selected quantitative impacts reported (table excerpts)
- Full sample 1996-2011 (impact reported during first quarter policy is effective, or most effective quarter in brackets):
  - Net private capital inflows: -1.3(4) then -1.9(5)
  - Portfolio investment: -0.6(1) then -1.2(4)
  - Bank loans/other investment: -1.8(4) then -1.7(5)
  - Exchange Market Pressure Index: -1.4(2)
  - Residential price/rent ratios: -11.5(4)
  - Residential price/income ratios: -2.0(5)
- 2000-2011 subsample (impacts shown where reported):
  - Net private capital inflows: -1.4(1) and 0.2(6)
  - Portfolio investment: 0.3(1) and -0.1(5)
  - Bank loans/other investment: -1.6(1) and 0.2(6)
  - Exchange Market Pressure Index: -1.4(2)
  - Residential price/rent ratios: 0.2(6)
  - Residential price/income ratios: -2.3(2)
- Note: The table reports impacts assessed within the six quarters following the introduction of the measure, controlling for volumes/levels during the year before the measure was introduced. Significance tested using the standard one-sided t-test. Results reported in bold are significant at the 10 percent level.

### Policy conclusions and recommendations
- Vigilance: Policymakers should remain focused on potential risks to the real economy and financial stability from capital inflow surges, given significant variation across countries in magnitudes and types of inflows and isolated pockets of concern (credit dynamics in some countries and certain segments of property markets).
- Mixed toolkit: A mixed policy toolkit will likely be required to manage large capital inflows. Macroprudential measures taken in the most recent surge have been appropriately narrowly targeted.
- Limitations of prudential measures: Experience with past surges suggests prudential measures may not always suffice by themselves in preventing the buildup of risks.
- Continued role for traditional policy tools:
  - Monetary policy and fiscal policy remain essential in managing demand conditions to prevent risks to the real economy and financial stability from building up.
  - Given the size of external buffers, allowing the exchange rate to take more of the burden of exchange market pressure will help manage inflows and contribute to rebalancing global demand.

*Source: IMF staff analysis in _wp12130 (selected excerpts).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12130.pdf_