## _wp1318

## Source details

**Canonical URL:** [_wp1318](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1318.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1318.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1318.pdf.json)

---

### I. Introduction — purpose and high-level findings
- Mongolia is resource-rich and undergoing rapid financial deepening; the paper assesses systemic risks and macroprudential responses.
- Key policy concern: rapidly rising government spending raising the risk of macroeconomic instability.
- Fund recommendation summarized:
  - Reorient policies to contain overheating, and
  - Strengthen the economy’s ability to endure prolonged global weakness and downside risks.
- Role of macroprudential instruments:
  - Complement traditional macro-management tools,
  - Maintain financial stability (chief objective),
  - Target a wider set of objectives and respond to shocks affecting financial stability.

### II. The Mongolian context — structural change, recent developments, and financial soundness
- Structural change and resource-windfall effects:
  - Shift from agriculture to mining; challenge to transform finite resource wealth into sustainable assets.
  - Historical warning: many resource-rich low-income countries show lower long-term physical and human capital accumulation.
- Recent macroeconomic developments (selected figures):
  - Real GDP growth reached about 17 percent in 2011, contributing to a positive output gap.
  - Private sector credit increased by 72 percent.
  - Imports of consumer goods increased by over 80 percent.
  - Inflation exceeded the policy target of a single-digit inflation rate.
- Financial system characteristics and indicators:
  - Bank-centric and relatively closed.
  - In 2011, banking system assets accounted for 50 percent of GDP and 96 percent of total financial sector assets.
  - Stock market: more than 300 companies listed on the Mongolia Stock Exchange (MSE); market capitalization surpassed 25 percent of GDP in 2011; index retreated 26 percent during the first 10 months of 2012.
- Financial soundness indicators and fragility:
  - Average Z-score of the banking system in 2011 is still below pre-crisis levels.
  - Ratio of liquid assets to total assets shrunk to 31.4 percent at end-2011 (below end-2009 level).
  - Bank returns on assets show improvement over the previous two years but movement away from center in banking soundness map.
- Policy implication:
  - Macroprudential tools and regulations could complement traditional macroeconomic policies to address systemic risks.

### Box 1 — Two large mining projects: Oyu Tolgoi and Tavan Tolgoi (projected impacts)
- Oyu Tolgoi (OT):
  - Located south near the Chinese border; one of the largest copper and gold mines.
  - Government ownership: 34 percent; rest owned by Turquoise Hill Resources (Canada) with Rio Tinto (UK) holding a 51 percent stake.
  - Production expected to start in early 2013.
- Tavan Tolgoi (TT):
  - Southern Mongolia; estimated reserves of about 6.4 billion tons of coal.
  - Erdenes Tavan Tolgoi LLC (100 percent state-owned enterprise subsidiary) holds mining license for East Tsankhi and West Tsankhi.
  - Operating contract with foreign companies for eastern block concluded in October 2011; modalities for western block under discussion.
- Projected export and fiscal outcomes:
  - Export proceeds from these mines could total US$2 billion in 2013, rising to US$7 billion by 2020.
  - Much of initial OT export proceeds will repay advanced tax payments and a loan used to finance the government’s equity stake.
  - By 2018, fiscal mineral revenue is projected to quintuple compared with 2011.
  - U.S. dollar GDP per capita could triple.

### III. Macro-financial linkages and systemic risk channels
- Financial development and growth:
  - Literature: financial deepening is a key driver of long-term growth; Mongolia may benefit from continued financial deepening as allocation efficiency rises.
- Systemic risk transmission channels (three levels):
  - Cross-market linkages: domestic money, bond, FX, and equity markets.
  - Cross-institution linkages: between banks and between banks and nonbank financial institutions.
  - External linkages: between global and local financial institutions and markets.
- Time-series (procyclical) dimension:
  - Build-up of risk over the macroeconomic cycle via procyclical leverage in upswing and risk aversion in downswing.
  - In Mongolia, commodity-driven credit activity and risk appetite are fueling exuberant asset market behavior.
- Cross-sectional dimension:
  - Growing size and complexity of banks increase interconnectedness and common exposures.
  - Rising exposure to commodity prices and to China increases Mongolia’s economic interdependence with its neighbor.
- Measurement caveats:
  - Market prices and common indicators can be misleading during exuberance or stress; risk measures often behave procyclically.
  - Focus policy on leading indicators derived from credit and asset price behavior; detailed credit-composition analysis is a basic building block.

### IV. Need for macroprudential policy and empirical support
- Constraints on traditional policies:
  - Monetary policy transmission to credit is limited in statistical significance and degree of impact (evidence from recursive VAR and simple regressions).
  - Excerpted empirical regression (Sample period: January 2001 - May 2012) highlights:
    - Constant: 0.030 *** (Std. error 0.007) [Pval][0.000]
    - Lagged bank credit (-1): 0.348 *** (Std. error 0.082) [Pval][0.000]
    - Lagged industrial production (-1): 0.016 (Std. error 0.017) [Pval][0.355]
    - CBB 7 day rate (-1): 0.000 (Std. error 0.001) [Pval][0.934]
    - CBB 7 day rate (-2): -0.002 * or -0.002 ** (Std. error 0.001) [Pval][0.087] [Pval][0.029]
    - Observations (after adj): 137; R-squared: 0.174, 0.174, 0.143 (across specifications).
- Conclusion: given monetary and fiscal constraints, macroprudential policy is especially important to reduce procyclicality, strengthen capital buffers, and reduce systemic risks.

### V. Procyclical channels and fiscal constraints (selected measures and rules)
- Procyclical channels identified:
  - Exports channel: commodity booms lift real GDP growth, spur bank lending, and amplify boom via foreign financing.
  - Fiscal channel: commodity price rises boost revenues; political pressure can increase spending pro-cyclically.
    - 2011: fiscal deficit recorded despite high commodity prices.
    - 2012 budget envisaged a 30 percent increase in spending even as inflation was in double digits.
  - Credit channel: optimism prompts freer lending and high credit growth.
  - Housing channel: government housing programs and highly leveraged mortgages amplify housing and banking cycles.
- Fiscal Stability Law (FSL) design and numerical targets (starting 2013):
  - Structural deficit target of two percent of GDP by 2013 (calculated using smoothed mineral prices, not adjusted for business cycle).
  - From 2013, spending to grow by less than the greater of non-mineral GDP growth in the budget year or the average of the previous 12 years.
  - Ceiling on the NPV of public debt to fall from 60 percent of GDP in 2012, to 50 percent of GDP in 2013, and 40 percent of GDP in 2014.

### VI. Macroprudential institutional framework and governance
- Institutional actors and roles:
  - Bank of Mongolia (BOM): prudential supervisor of banks and specifically charged with maintenance of financial stability; conducts on-site and off-site supervision.
  - Financial Regulatory Council (FRC): regulates non-bank financial institutions.
  - Financial Stability Committee (FSC): steering committee comprising the Governor of BOM, Minister of Finance, and chairman of the FRC; meets quarterly.
- Assessment of Mongolia’s model:
  - De jure coordinating body is the FSC, but it is low profile, focused on specific issues, lacks a permanent secretariat, and its legal basis (joint decree) is weaker than budget and central bank law.
  - BOM is de facto the most important macroprudential institution because banking supervision covers 96 percent of the financial system.
- Criteria for an effective institutional model (preserved):
  - Effective identification, analysis, and monitoring of systemic risk (access to data; resources; strong mandate; ability/willingness to act; accountability).
  - Timely and effective use of macroprudential tools.
  - Effective coordination across policies to reduce gaps and overlaps while preserving autonomy.
- Recommended institutional reforms to raise FSC effectiveness:
  - Modify FSC statutes from decree into law.
  - Enhance FSC accountability: regular updates to parliament and/or publication of assessments.
  - Publicize dissenting views among FSC members and/or publish assessments, respecting confidentiality.
  - Appoint a permanent secretary to the FSC, to be housed at the BOM.
  - Grant FSC the power to make general “recommendations” (not referring to supervision of specific institutions).
  - Establish a macroprudential unit within the BOM (macro economist, researchers, prudential regulators).

### VII. Macroprudential instruments — dampening procyclicality (overview and Mongolia-specific measures)
- General candidate instruments (used in combination):
  - Countercyclical capital requirements (buffers based on cyclical position).
  - Variable risk weights (raise capital for specific lending, e.g., real estate).
  - Forward-looking provisioning (link loss provisions to credit cycle).
  - Collateral requirements (loan-to-value limits, minimum haircuts/margins).
  - Quantitative credit controls and reserve requirements.
- Measures already utilized in Mongolia (selected figures and rules):
  - Liquidity ratio: raised from 18 to 25 percent in 2011; not binding given banks’ current liquidity ratio is 36 percent.
  - Capital Adequacy Ratio: raised from 12 to 14 percent for the 5 systemic banks (SIFI surcharge).
  - Limits on exposure concentration:
    - Single counterparty exposure (loans, loan-equivalent assets, guarantees and warranties) shall not exceed 20 percent of the bank’s capital.
    - For a bank’s related party, exposure shall not exceed five percent.
    - (Revised Banking Law Jan 28, 2010; BOM Governor's order No. 460 on July 30, 2010).
  - Limits on net open currency positions: single foreign exchange open position should not exceed 15 percent of the bank’s equity capital; overall open position limit set at 40 percent of capital.
  - Limits on maturity mismatches: difference between average durations of assets and liabilities shall not exceed 30 percent of total assets’ average duration.
  - Reserve requirement: applied against both domestic and foreign currency deposits.
  - Provisioning: provisioning rate was reduced at the time of the crisis.
  - Countercyclical capital buffers: absent from the current toolkit in Mongolia.

### VIII. Dynamic provisioning — simulation setup, empirical evidence, and implications
- Rationale: backward-looking provisioning contributes to procyclicality; dynamic provisioning builds cushions in good times to be released in downturns.
- Simulation approach:
  - Banks contribute the difference between monthly statistical losses on loans to the non-financial private sector and the realized net loan loss in that month; statistical losses computed from expected rates by loan category.
- Simulation findings:
  - Countercyclical system would have smoothed credit growth; when the 2008/09 crisis hit, small dynamic-provision cushion would have absorbed only a fraction of large loan losses.
  - Preliminary simulations point to a smoother credit growth path with dynamic provisions, implying a safer financial system and less macroeconomic volatility.
- VECM empirical support (quarterly data 2006-2012; selected statistics preserved):
  - Long-run relations: real industrial production positively correlated with real credit and negatively correlated with provisions.
  - Short-run dynamics: changes in provisions play a small role in explaining real industrial production and credit growth.
  - Table 5 summary (preserved elements):
    - Real Credit, 1 lag coefficient -0.83 [-2.00]; Provisions, 1 lag 0.44 [7.38]; Constant -20.32 n/a.
    - Error correction equation coefficients: 0.17, 0.72, -0.126 (t-statistics preserved in source).
    - R-squared: Industrial Production 0.23; Real Credit 0.63; Provisions 0.44.
    - Adj. R-squared: 0.14; 0.59; 0.38.
    - F-statistic: 2.64; 15.47; 7.11.
- Conclusion: there appears to be a case for regulatory dynamic provisions in Mongolia to contain insolvency risk and dampen credit procyclicality.

### IX. Reducing cross-sectional systemic risk and SIFI policies
- SIFI designation and surcharges:
  - Five institutions designated as SIFIs: Khan Bank, Golomt Bank, Trade and Development Bank (TDB), Xac Bank, and Savings Bank.
  - These five SIFIs are subject to capital requirements that are two percentage points higher than for other banks.
- Cross-sectional concerns:
  - SIFIs create systemic risk via size, interconnectedness, and implicit guarantees producing funding advantages and moral hazard.
  - Some smaller banks may be highly interconnected and warrant extra supervision or capital.
  - Common exposures (commodity cycle and China) create simultaneous distress risk.
- Policy options:
  - Broaden SIFI definition to include interconnectedness, common exposure, substitutability, and concentration.
  - Consider disincentives for excessive size (e.g., taxing incremental size).
  - Strengthen the resolution regime to manage SIFI failure in an orderly manner and preserve market discipline.
- International context:
  - Systemic capital surcharges internationally vary from 0.5-2.5 percent of capital (work ongoing on modalities).

### X. Systemic risk from dollarization (yuanization) and FX regulation
- Dollarization metrics and vulnerabilities:
  - Deposits and loans in foreign exchange each account for about one third of total deposits and loans, respectively (around 25-30 percent cited elsewhere).
  - Hedging instruments (forwards, derivatives) are lacking.
- Existing prudential FX measures and gaps:
  - BOM limits on net open foreign exchange positions: single currency limit 15 percent of capital; overall limit 40 percent of capital.
  - Some regulations remain currency-blind despite risks from dollarization and a flexible exchange rate regime.
- Proposed FX-related tools:
  - Variable reserve requirement by currency (lower for togrog deposits than for foreign currency deposits).
  - Currency-matching regulation: calculate open positions by maturities via a maturity ladder and impose limits period-by-period.
  - Stricter capital treatment for FX loans to non-foreign-currency earners (example: 150 percent weight in other jurisdictions).
  - Suggested limit example: cap ratio of USD-denominated loans to USD-denominated deposits at 100 percent to limit liquidity risk in dollar lending.

### XI. Measurement, early warning systems (EWS), and limits of surveillance
- Measurement challenges:
  - Deficiencies include short-sighted risk measures, insufficient recognition of latent endogenous instability, and non-measurable risks.
  - Market prices may be unreliable during calm or stress.
- Recommended EWS focus:
  - Use deviations from long-term trends in credit/GDP, real estate prices, and equity prices as promising EWIs.
  - Strengthen sectoral balance sheet data (household, corporate, public) to track debt levels and debt-servicing capacities.
  - Emphasize credit composition and terms (borrower, amounts, purpose, maturity) to distinguish beneficial financial deepening from imprudent credit growth.
  - Design EWS frameworks that map vulnerabilities and rank threats to support prioritized responses.
- Comparative assessment of tools (preserved conclusions from Borio and Drehmann, 2009):
  - FSIs, credit ratings, and some market-price indicators are limited as leading indicators.
  - EWIs drawing on endogenous cycle perspective are most promising.
  - Macro stress tests and multiple-module measures offer structure and storytelling but may miss endogenous systemic instability.

### XII. Discretionary versus automatic macroprudential measures and policy interactions
- Discretionary vs automatic:
  - Automatic stabilizers reduce need for continual justification and limit political pressure; credible pre-committed rules enforce buildup of buffers in good times.
  - Supervisors should retain discretion to address idiosyncratic risks and require bank-specific buffers when warranted.
- Interaction with monetary policy:
  - Monetary policy should continue to focus on inflation; macroprudential tools should target systemic risk.
  - When credit booms coincide with general overheating, monetary policy should act first and foremost.
  - Complementarities and tensions exist: macroprudential tightening can contract activity and lower price pressures; low interest rates can encourage risk-taking and excessive credit growth.
  - Need for research and statistical analysis to understand complementarities and conflicts.

### XIII. Practical policy recommendations (summary)
- Strengthen macroprudential institutional capacity:
  - Legislate FSC statutes, enhance accountability, appoint a permanent secretary, and create a macroprudential unit within the BOM.
- Enhance surveillance and EWS:
  - Use EWIs focused on credit/GDP and asset prices; improve sectoral balance sheet data; analyze credit composition and terms.
- Broaden and calibrate instruments:
  - Introduce countercyclical capital buffers and/or dynamic provisioning.
  - Use variable risk weights and forward-looking provisioning for targeted sectors (e.g., real estate).
  - Apply collateral/LTV and DTI limits where appropriate.
  - Adjust reserve requirements by currency and strengthen currency-matching rules to address dollarization/yuanization.
- Manage cross-sectional risks:
  - Broaden SIFI assessment beyond size to include interconnectedness and common exposures.
  - Apply SIFI surcharges, consider disincentives for excessive size, and strengthen resolution regimes.
- Preserve complementarity with macro policies:
  - Macroprudential tools are not substitutes for necessary fiscal and monetary adjustments; sound monetary policy and fiscal discipline remain essential.

*Source: Excerpts and analysis drawn from the supplied IMF PDF chapter/section (_wp1318).*

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

### References

### I. INTRODUCTION
- Mongolia is rich in natural resources and undergoing rapid financial deepening.
- The country experienced severe financial distress in 2009 but has made important strides since then.
- Rapidly rising government spending is raising the risk of macroeconomic instability.
- The Fund recommended reorienting policies to:
  - contain overheating, and
  - strengthen the economy’s ability to endure prolonged global weakness and downside risks.
- Macroprudential instruments can complement traditional macro-management tools to:
  - maintain financial stability (chief objective),
  - target a wider set of objectives, and
  - respond to shocks that negatively affect financial stability.
- This paper assesses systemic risks in Mongolia and identifies macroprudential instruments to be used concomitantly with traditional monetary and fiscal policies.

### II. THE MONGOLIAN CONTEXT
- Structural change: shift in economic production from agriculture to mining driven by mineral extraction.
- Resource-windfall effects:
  - Modernization of the economy, investment, and rising urbanization.
  - Challenge: transform finite resource wealth into assets that sustain development.
  - Historical warning: many resource-rich low-income countries show lower long-term physical and human capital accumulation.
- Recent macroeconomic developments:
  - Loose fiscal policy and rising FDI contributed to overheating pressures.
  - Real GDP growth reached about 17 percent in 2011, contributing to a positive output gap.
  - Private sector credit increased by 72 percent.
  - Imports of consumer goods increased by over 80 percent.
  - Inflation exceeded the policy target of a single-digit inflation rate.
- Financial system characteristics:
  - Bank-centric (Figure 2) and relatively closed.
  - In 2011, banking system assets accounted for 50 percent of GDP and 96 percent of total financial sector assets.
  - No presence of foreign banks in Mongolia as of the period discussed; overseas finance corporations may establish a domestic presence.
  - Stock market: more than 300 companies listed on the Mongolia Stock Exchange (MSE); market capitalization surpassed 25 percent of GDP in 2011; index retreated 26 percent during the first 10 months of 2012.
- Capital flows and indicators (as shown in figures):
  - Strong capital inflows including Loans (net), Currency and Deposits (net), Trade Credit (net), Portfolio Investment (net), Direct Investment (net) shown in millions of U.S. dollars.
  - Output gap, GDP, Potential GDP and related series shown (1995–2011, HP filtering).
- Financial soundness and systemic risk:
  - Backward-looking financial soundness indicators suggest the banking system appears to be fine, but a sudden downturn could trigger deterioration (experience in 2008–09).
  - The average Z-score of the system in 2011 is still below pre-crisis levels, pointing to continued fragility.
  - The ratio of liquid assets to total assets shrunk to 31.4 percent at end-2011 (below end-2009 level).
  - Returns on assets show improvement over the previous two years but movement away from the center in the banking soundness map.
- Policy implication:
  - Macroprudential tools and regulations could complement traditional macroeconomic policies to address systemic risks.

### III. MACRO-FINANCIAL LINKAGES IN MONGOLIA
- Financial development and growth:
  - Literature identifies financial sector development as a key driver of long-term growth (Levine, 2005).
  - A small financial system implies higher fixed setup costs, reduced information flows, limited risk diversification, and forgone profitable investments.
  - Rioja and Valev (2004a, 2004b): finance boosts growth by speeding productivity in rich countries and accelerating capital accumulation in poorer countries; effects may be nonlinear and particularly large for middle-income countries.
  - Mongolia may benefit from continued financial deepening; improved resource allocation importance will rise.
- Natural-resource driven growth:
  - Mongolia’s double-digit growth rates reflect tapping previously unused natural resources (Figure 4).
  - Examples:
    - Oyu Tolgoi (OT) copper mine described as the world’s largest underdeveloped copper-gold project.
    - Tavan Tolgoi (TT) described as one of the world’s largest untapped coal deposits.
- Paper structure (what follows in the full text):
  - Section III assesses macro-financial linkages by examining procyclical and cross-sectional systemic risks.
  - Section IV considers available institutional models for macroprudential policy.
  - Section V discusses suitability of potential instruments to dampen procyclicality and control cross-sectional systemic risks.
  - Section VI discusses critical issues complementing macroprudential tools and monetary policy implementation.
  - Section VII concludes.

*Source: _wp1318 - References*

### Box 1. Mongolia: The Effects of Two Large Mining Projects

### Box 1. Mongolia: The Effects of Two Large Mining Projects

### Mining projects and projected export and fiscal impacts
- Oyu Tolgoi:
  - Location: south near the Chinese border.
  - Will be one of the largest copper and gold mines in the world.
  - Government ownership: 34 percent.
  - Rest owned by Turquoise Hill Resources (Canada), in which Rio Tinto (UK) has a 51 percent stake.
  - Production expected to start in early 2013.
- Tavan Tolgoi:
  - Location: southern Mongolia.
  - One of the world’s largest untapped coal deposits with estimated reserves of about 6.4 billion tons.
  - Erdenes Tavan Tolgoi LLC, a subsidiary of a 100 percent state-owned enterprise, holds the mining license for the eastern bloc (East Tsankhi) and the western bloc (West Tsankhi).
  - Operating contract with foreign companies for the eastern block concluded in October 2011; modalities for the western block under discussion.
- Projected export and fiscal outcomes:
  - Export proceeds from these mines could total US$2 billion in 2013, rising to US$7 billion by 2020.
  - Much of initial Oyu Tolgoi export proceeds will be used to repay advanced tax payments and a loan used to finance the government’s equity stake.
  - By 2018, fiscal mineral revenue is projected to quintuple compared with 2011.
  - U.S. dollar GDP per capita could triple.

### Natural resource effects on financial development and stability
- The natural resource curse can:
  - Hold back financial development.
  - Raise risks to the financial system through several channels:
    - Reduced demand for finance:
      - Resource-rich countries absorb shocks and smooth consumption by drawing on resource revenues, retarding financial system development (Gylfason, 2004).
      - Demand for financial services is lower in the natural resource sector than in the non-resource tradable sector.
      - The non-resource tradable sector is held back by Dutch disease effects.
    - Rent-seeking behavior:
      - Commodity booms can shift entrepreneur behavior toward rent-seeking, reducing demand for finance for start-ups and firm expansions (Mehlum and others, 2006).
    - Institutional distortions raising lending risk:
      - Easier short-term profits from natural resources reduce investment in institutional frameworks supporting private property rights, contract enforcement, and transparency (Besley and Persson, 2010).
    - Making lending to some sectors riskier:
      - Resource booms raise macroeconomic volatility; booms typically followed by busts increase riskiness of lending, especially to the tradable sector, and crowd out non-resource activities.

### Overview of risks in the Mongolian financial system and regulatory implications
- Rapid financial expansion increases systemic risk and requires upgrades to regulatory policies.
- Lessons from the global financial crisis:
  - Monetary policy aimed at price stability plus micro-prudential regulation is insufficient to ensure financial stability.
  - Need for a strong set of microprudential (entity-level) regulations complemented by macroprudential (system-wide) regulation and supervision.
  - Macroprudential measures can enhance shock absorbers in terms of capital and liquidity, constrain overall leverage, and extend the regulatory perimeter to systemically important institutions, markets, and instruments.
- Fiscal policy constraints:
  - Significant time-lag in fiscal policy means fiscal tools may not provide timely responses to systemic risk, particularly in a democratic setting with political business cycles.
  - Fiscal Stability Law (FSL), which becomes fully effective in 2013, is designed to make fiscal policy countercyclical.
  - Dell’Ariccia and others (2012) found countercyclical fiscal policy reduces overheating pressures associated with a credit boom and helps reduce systemic risk.

### Monetary policy transmission and empirical findings
- Monetary policy may not on its own be relied upon to reduce system risk in Mongolia:
  - Recursive VAR analysis suggests bank credit is impacted by changes in the policy rate (7-day central bank bill rate) with some time lag, but transmission looks limited in statistical significance and degree of impact (Figure 5).
  - Simple regressions with additional explanatory variables like industrial production give a similar result of limited transmission from the policy rate to bank credit (Table 2).
- Excerpted empirical results (Sample period: January 2001 - May 2012):
  - Constant: 0.030 *** (Std. error 0.007) [Pval][0.000]
  - Lagged bank credit (-1): 0.348 *** (Std. error 0.082) [Pval][0.000]
  - Lagged industrial production (-1): 0.016 (Std. error 0.017) [Pval][0.355]
  - CBB 7 day rate (-1): 0.000 (Std. error 0.001) [Pval][0.934]
  - CBB 7 day rate (-2): -0.002 * or -0.002 ** (Std. error 0.001) [Pval][0.087] [Pval][0.029]
  - Observations (after adj): 137; R-squared: 0.174, 0.174, 0.143 (across specifications).
- These results are consistent with constraints on monetary policy in countries with open capital accounts (Hahm and others, 2012).

### Need for macroprudential policy
- Given constraints on fiscal and monetary policy, macroprudential policy is especially important in Mongolia to:
  - Reduce the pro-cyclical bias of the financial system.
  - Strengthen banks’ capital buffers.
  - Reduce potential systemic risks via targeted macroprudential tools.

### Systemic risk transmission channels and measurement caveats
- Financial linkages transmitting systemic risk exist at three levels:
  - Cross market linkages: domestic money, bond, foreign exchange, and equity markets.
  - Cross institution linkages: between banks and between banks and nonbank financial institutions.
  - External linkages: between global and local financial institutions and markets.
- Systemic risk decomposed into:
  - Time-series dimension:
    - Build-up of risk over the macroeconomic cycle with procyclical leverage in upswing and risk aversion in downswing.
    - In Mongolia, credit activity and risk appetite driven by the commodity boom are fueling exuberant asset market behavior.
  - Cross-sectional dimension:
    - Growing size and complexity of banks increase interconnectedness and common exposures, enabling rapid contagion.
    - Rising exposure to commodity prices and to China increases Mongolia’s economic interdependence with its neighbor.
- Data caveats and limitations:
  - Market prices and financial indicators may not always be good indicators of actual systemic risk, particularly during exuberance or stress.
  - Risk measures often behave procyclically (underestimate risk in upswing, overestimate in downswing).
  - Main measurement challenges relate to latent, endogenous systemic instability (Minsky moment) and non-measurable risks.
  - Policy should focus on leading indicators derived from credit and asset price behavior; understanding credit composition is a basic building block of macroprudential surveillance.

### Pro-cyclical systemic risk channels in Mongolia
- Exports channel:
  - Commodity booms (coal, copper, iron ore) lift real GDP growth, spurring domestic banks to lend; combined with foreign financing, private sector debt upsurges and amplifies the boom.
- Fiscal channel:
  - Rising commodity prices boost royalties and tax revenues through multiplier effects.
  - Political pressure increases spending pro-cyclically; a fiscal deficit was recorded in 2011 despite high commodity prices.
  - The 2012 budget envisaged a 30 percent increase in spending even as inflation was in double digits.
  - Observance of the FSL (fully operational with the 2013 budget) should preclude such procyclical fiscal policy.
  - FSL numerical targets include:
    - A structural deficit of two percent of GDP by 2013 (calculated using smoothed mineral prices to estimate revenue, but not adjusted for the business cycle).
    - Starting in 2013, spending to grow by less than the greater of non-mineral GDP growth in the budget year or the average of the previous 12 years.
    - Ceiling on the NPV of public debt to fall from 60 percent of GDP in 2012, to 50 percent of GDP in 2013, and 40 percent of GDP in 2014.
- Credit channel:
  - Optimism in upswing leads to freer lending and high credit growth; LTV ratios may be raised and lending to riskier segments may increase.
  - Credit is highly procyclical, raising systemic risk when perceived risk is lowest.
- Housing channel:
  - Housing policy and government-supported mortgage programs can amplify the housing and banking cycles.
  - Authorities have a program to build 100,000 apartments, including 75,000 in the capital; a groundbreaking was held in Yarmag for 15,000 apartments; the government decided to allocate 1 million togrogs to citizens who buy an apartment in the project; the 2012 budget allocates 100 billion togrogs.
  - Mortgages are highly leveraged and house price changes exhibit strong autocorrelation and momentum, increasing systemic risk.

### Cross-sectional systemic risks: specifics for Mongolia
- Interconnectedness:
  - Rapid banking system growth, doubling in asset size in 2011, increases interconnectedness and the risk that a shock to one institution spreads system-wide.
- Dollarization:
  - Deposits and loans in foreign exchange each account for about one third of total deposits and loans, respectively, creating potential balance sheet mismatches.
  - 2010 FSAP stress tests confirm large indirect vulnerabilities to FX movements.
- Common exposure:
  - Banks are highly exposed, directly and indirectly, to the commodity cycle and to China; deterioration in commodity prices or weakening in China would immediately impact the banking system.
- Systemically Important Financial Institutions (SIFIs):
  - Five privately-owned banks dominate the banking landscape; problems at any one are likely to create externality effects and contagion risks.

### Macroprudential institutional framework for Mongolia
- Ultimate responsibility for financial stability resides with:
  - Bank of Mongolia (BOM): prudential supervisor of banks and specifically charged with maintenance of financial stability; conducts on-site and off-site supervisory activities.
  - Financial Regulatory Council (FRC): regulates other (non-bank) financial institutions.
  - Financial Stability Committee (FSC): steering committee comprising the Governor of the central bank, the Minister of Finance, and the chairman of the FRC; meets quarterly to discuss potential risks to financial stability.
- Institutional models and assessment:
  - A range of macroprudential institutional frameworks exists; effectiveness of models is untested and varies by country-specific factors.
  - The IMF (2011) classifies macroprudential institutions by five dimensions: (i) degree of integration between central bank and financial regulatory policy functions; (ii) ownership of the macroprudential mandate; (iii) role of government in policy setting; (iv) organizational separation of decision-making and control over instruments; (v) presence of a coordinating committee.
  - Mongolia’s framework:
    - FSC is de jure the main coordinating body for macroprudential coordination.
    - The FSC has a team located at the BOM that conducts analyses and assessments of the overall financial system.
    - Mongolia’s arrangement resembles a “twin-peak” model (Model III in Table 3): close institutional integration between the central bank and prudential supervision of potentially systemic institutions (banks), while conduct regulation in retail and wholesale markets is institutionally separate from the central bank.

*Source: IMF staff analysis as presented in Box 1 of the provided document.*

### 1. Degree of institutional integration of central bank and  supervisory agencies

### 1. Degree of institutional integration of central bank and  supervisory agencies

### Institutional integration matrix (summarized)
- Full (at a central bank)
- Partial
- Partial
- Partial
- No
- No (Partial)
- No
- No

### Ownership of macroprudential policy and financial stability mandate (reported)
- Central bank
- Committee
- “related” to central bank
- Independent committee
- Central bank
- Multiple agencies
- Multiple agencies
- Multiple agencies
- Committee (multinational; regional)

### Role of MOF/treasury/government (reported)
- No (Active)
- Passive
- Active
- No
- Passive
- Active
- No
- Passive
- (European Commission; Economic and Financial Committee)

### Separation of policy decisions and control over instruments (reported)
- No
- In some areas
- Yes
- In some areas
- No
- No
- No
- Yes

### Existence of separate body coordinating across policies (reported)
- No
- No
- No
- No
- Yes
- Yes
- No
- No

### Examples of specific model countries/regions (as listed)
- Singapore
- Czech Republic
- Ireland (new)
- Romania
- United Kingdom (new)
- Brazil
- France
- United States
- Mongolia
- Thailand
- The Netherlands
- Serbia
- Belgium (new)
- Australia
- Canada
- Hong Kong SAR
- Lebanon
- Mexico
- Iceland
- Japan
- Korea
- Peru
- Switzerland
- EU (ESRB)

*Source: IMF (2011); and authors’ additions.*

### Legal foundations and practical effectiveness (Mongolia case)
- The legal document creating the FSC does not clearly identify entities responsible for macroprudential policies.
- The FSC tends to be low profile; discussions are focused on specific issues rather than systemic risk more broadly.
- The FSC lacks a permanent secretariat; only a limited number of persons at the central bank are assigned to FSC-related activities.
- The FSC’s legal basis (established by joint decree) is weaker than the budget laws and the central bank law, weakening accountability.

### De facto institutional power
- The Bank of Mongolia (BOM) is de facto the most important macroprudential/financial stability institution in Mongolia because it is in charge of banking supervision, which in 2011 accounted for 96 percent of the financial system.
- Two reasons for BOM’s relative weight:
  - The FSC lacks adequate legal powers.
  - Mongolia’s financial system is bank-centric; analysis of systemic risk and tool box to tackle it are under the BOM.
- As the non-bank financial system gains sophistication, the FSC’s role should be reinforced.

### Key drivers for effective macroprudential policy arrangements
- Two basic elements:
  - Presence of a well-identified authority (institution or policy committee) with a clear macroprudential mandate.
  - A mechanism promoting consistency across policies to preserve financial stability.
- Model III (as referenced) could be conducive to effective mitigation of systemic risk if:
  - It provides for effective identification, analysis and monitoring of systemic risk.
  - It provides for timely and effective use of macroprudential policy tools by avoiding coordination problems to address systemic risk and reduce gaps and overlaps.

### Criteria for effectiveness of macroprudential institutional model (as presented)
- Effective identification, analysis, and monitoring of systemic risk
  - Access to relevant information
  - Using existing resources and expertise
  - Strong mandate and powers
  - Ability and willingness to act
  - Accountability
- Timely and effective use of macroprudential policy tools
- Effective coordination across policies aiming to address systemic risk
  - Reducing gaps and overlaps
  - Preserving the autonomy of separate policy functions

### Identification and monitoring of risk (observations and weaknesses)
- Main challenge: identify macroeconomic and financial variables that provide insight into risks and vulnerabilities; data availability differs by country.
- For Mongolia, challenge is to identify the data set most likely to provide warnings of systemic risks for preemptive response.
- FSC’s access to BOM’s in-house research and regulatory data from BOM (banking) and FRC (non-banking) should facilitate monitoring, assuming no legal confidentiality restrictions.
- Integration could bring together microprudential expertise and macro-financial linkage assessment, useful for stress-testing.
- The 2010 FSAP indicated stress testing mechanisms need to be further strengthened in Mongolia.
- Risk: current model may limit institutional mechanisms to challenge entrenched “house views” because it relies heavily on BOM.
- Current weakness of BOM’s risk-monitoring function and need to strengthen microprudential regulation imply synergies are currently weak.
- Recommendation: BOM should strengthen expertise in monitoring financial markets and maintain open communications with the FRC to ensure full-spectrum systemic risk monitoring.

### Use of macroprudential tools and coordination (roles and trade-offs)
- Mitigation of system risk is a shared responsibility led by FSC, with central bank and regulatory agency playing important roles.
- Central bank contributions include:
  - Sound monetary policy promoting low inflation.
  - Development of robust payments infrastructure, including reliable settlement system.
  - Support for system liquidity through financial market operations.
- Banking regulator (BOM) and FRC (non-bank regulator) should implement best prudential practices, tailored to national circumstances.
- Progress needed on corporate governance, insolvency, creditor protection, and suitable accounting and auditing standards.
- Assigning the macroprudential mandate to the FSC (no monetary policy role) can prevent reputational cross-over between monetary policy and macroprudential failures.
- If mandate is shared across agencies, incentive and accountability problems increase, potentially delaying action.

### Coordination benefits of integration
- Coordination across macroprudential, monetary, and microprudential functions within one organization can:
  - Increase effectiveness of decision-making when internalizing trade-offs (e.g., price stability vs financial stability).
  - Reduce mismatches between mandates and powers since decision-maker controls most relevant tools.
  - Be achievable without compromising operational autonomy of separate agencies.
- Risks and benefits of treasury involvement:
  - Strong treasury role may risk short-term political considerations prevailing over long-term stability and could cause delays in countercyclical action.
  - Balanced committee structure may lead to unresolved differences and delays.
  - Treasury involvement may improve government understanding of systemic risk.
  - Pressure from commercial banks on BOM may be mitigated in Mongolia’s model.

### MPC versus FSC (comparative observations)
- MPC: Narrow mandate (price stability), clear measurable target (CPI), typically majority decision-making, accountability easier to establish, has own policy instruments, clear punishment mechanism for failing mandate.
- FSC: Broad mandate (financial stability), not clearly measurable (soft target), members include central bank, regulators and Ministry of Finance, decision-making consensual, accountability not clearly measurable, often advisory with moral suasion, lacks own policy instrument, unclear punishment mechanism for failing mandate.
- Implication: FSCs tend to rely more on judgment and less on quantitative frameworks; advice can be ignored, reducing relevance.

### Way forward: recommended steps to raise FSC effectiveness
- Modify statutes of the FSC from decree into law to raise prominence of statutes.
- Enhance accountability of FSC—expand to encompass minimization of buildup of systemic risk through regularly updating parliament and/or publishing its assessment.
- Improve accountability by publicizing dissenting views among FSC members and/or publishing its assessment consistent with confidentiality and privacy requirements.
- Appoint a permanent secretary to the FSC, to be housed at the BOM, to elevate its role and provide necessary institutional setting.
- Grant the FSC the power to make “recommendations” that are general and do not refer to supervision of specific institutions.
- Establish a macroprudential unit within the BOM, comprising a macroeconomist, researchers, and prudential regulators tasked with enhancing macroprudential analysis.

### Macroprudential instruments for Mongolia — dampening procyclicality (overview)
- Choice of instruments depends on country circumstances (stage of development, structure of economy).
- Mature economies tend to use macroprudential instruments less; fixed exchange rate regimes tend to use them more.
- Instruments are empirically used in combination rather than singularly to reduce arbitrage and address risks from multiple angles.
- Candidate instruments include:
  - Countercyclical capital requirements (buffers based on cyclical position).
  - Variable risk weights (raise capital for specific lending, e.g., real estate).
  - Forward-looking provisioning (link loss provisions to credit cycle).
  - Collateral requirements (e.g., loan-to-value limits, minimum haircuts/margins on securities financing).
  - Quantitative credit controls and reserve requirements (limit lending or increase short-term liquidity requirements).

### Macroprudential measures already utilized in Mongolia
- Increase in the liquidity ratio: the ratio was raised from 18 to 25 percent

*Source: IMF (2011); and authors’ additions.*

### 2011. However, this was not a binding increase given that banks’ current liquidity

### _wp1318 - 2011. However, this was not a binding increase given that banks’ current liquidity ratio is 36 percent.

### Prudential measures enacted and calibration
- Increase in the Capital Adequacy Ratio: raised from 12 to 14 percent for the 5 systemic banks (described as akin to a SIFI surcharge).
- Limits on exposure concentration:
  - Total value of loans, loan equivalent assets, guarantees and warranties to one person and/or his/her related party shall not exceed 20 percent of the capital of the bank.
  - In the case of a bank’s related party, it shall not exceed five percent.
  - (Revised Banking Law on Jan 28, 2010, and BOM Governor's order No. 460 on July 30, 2010—Regulation on setting and supervising prudential ratios of banking operations).
- Limits on net open currency positions: amount of a single foreign exchange open position should not exceed 15 percent of the bank’s equity capital.
- Limits on maturity mismatches: difference between average durations of assets and liabilities shall not exceed 30 percent of total assets’ average duration.
- Reserve requirement: all licensed banks must meet the reserve requirement applied against both domestic and foreign currency deposits.
- Provisioning: the provisioning rate was reduced at the time of the crisis.
- Noted absence: countercyclical capital buffers have been absent from the current toolkit in Mongolia.

*BOM Governor's order No. 460 on July 30, 2010—Regulation on setting and supervising prudential ratios of banking operations.*
*BOM Governor’s order No.118 date March 5, 2007.*

---

### Basel Committee buffers and international context
- BCBS agreed two capital buffers to smooth credit through the cycle:
  - Capital conservation buffer: fixed proportion of risk-weighted assets; may be run down during stress.
  - Additional countercyclical buffer: to be imposed in periods of rapid credit growth; could be released in downturns.
- BCBS reference indicators: ratio of credit to GDP as a common reference in the buildup phase; broader set of indicators, including asset prices, also to be considered.
- BCBS press release note: proposals include a conservation capital buffer of 2 ½ percentage points on top of minimum requirements and an additional counter-cyclical buffer of 0–2 ½ percent.
- International examples and lessons:
  - Spain: long-term advocate of dynamic provisioning—helped dampen housing bubble but buffers were not sufficiently elevated to avoid boom-bust.
  - India: used differential weighting in capital regulation to slow growth of bank credit to housing and commercial real estate.
  - Mongolia: policy to encourage housing ownership has led to reluctance to alter Loan-to-Value (LTV) ratio and Debt-to-Income (DTI) ratio, potentially harming affordability over the longer run.

---

### Dynamic provisioning: simulation setup and results
- Rationale: backward-looking loan loss provisioning rules contribute to credit booms and procyclicality; dynamic provisioning builds cushions in good times to be released in downturns.
- Simulation method:
  - Banks contribute the difference between monthly statistical losses on loans to the non-financial private sector and the realized net loan loss in that month.
  - Statistical losses calculated from expected rate of loss for respective loan categories, including specific risk loans weighted by loan volumes.
  - Formal expression presented for contribution to dynamic provisions fund (notation preserved from source).
  - Assumed average of actual provisioning under perfect-foresight; assumed some specific portion of NPL would have been realized as losses.
- Simulation findings:
  - Countercyclical system took effect toward the end of the previous credit cycle; when the 2008/09 crisis hit, the small cushion of dynamic provisions could absorb only a fraction of large loan losses.
  - Preliminary simulations point to a smoother credit growth path had Mongolia adopted dynamic provisions in the past, implying a safer financial system and less macroeconomic volatility.
- Empirical support:
  - A vector error correction model (VECM) based on quarterly data (2006-2012) on real industrial production, domestic credit, and provisions confirms:
    - In the long run: real industrial production exhibits a contemporaneous positive correlation with real credit and a negative correlation with provisions.
    - Short-run dynamics: changes in provisions play a small role in explaining real industrial production and credit growth.
  - Table 5 VECM summary (numbers preserved as in source):
    - Cointegrating Eq: Industrial Production, 1 lag 1 n/a
    - Real Credit, 1 lag -0.83 [-2.00]
    - Provisions, 1 lag 0.44 [7.38]
    - Constant -20.32 n/a
    - Error Correction Equations (Cointegrating Equation coefficients): 0.17, 0.72, -0.126 with associated t-statistics in square brackets as in source.
    - R-squared: Industrial Production 0.23; Real Credit 0.63; Provisions 0.44.
    - Adj. R-squared: 0.14; 0.59; 0.38.
    - F-statistic: 2.64; 15.47; 7.11.
- Conclusion: there appears to be a case for using regulatory dynamic provisions in Mongolia to contain bank insolvency risk and dampen credit procyclicality.

---

### Reducing cross-sectional systemic risk and SIFIs
- SIFI designation and measures in Mongolia:
  - Mongolia has designated five institutions as SIFIs: Khan Bank, Golomt Bank, Trade and Development Bank (TDB), Xac Bank, and Savings Bank.
  - These five SIFIs are subject to capital requirements that are two percentage points higher than for other banks.
  - SIFIs appear identified based on size, but interconnectedness and common exposures are also important considerations.
- Key systemic concerns:
  - SIFIs add disproportionately to systemic risk due to size/interconnectedness and implicit government guarantees leading to funding advantages and moral hazard.
  - Some banks not among largest may be highly interconnected and require extra supervision or capital.
  - Common exposures among banks can cause simultaneous distress from a common shock.
- Policy options to manage cross-sectional systemic risk:
  - Broadening the SIFI definition: monitor interconnectedness, common exposure, substitutability and concentration in addition to size.
  - Introducing policies that further discourage SIFIs: consider disincentives such as taxing the incremental size of already large institutions.
  - Strengthening the resolution regime: improved resolution regimes to manage SIFI failure in an orderly manner; ensure counterparties are not sheltered from loss to strengthen market discipline ex ante.
- Note on SIFI surcharges internationally: systemic capital surcharge adopted varying from 0.5-2.5 percent of capital (work ongoing on modalities).

---

### Systemic risk from dollarization (yuanization) and FX regulation
- Dollarization metrics and vulnerabilities:
  - Share of deposits and loans in foreign exchange is around 25-30 percent.
  - Instruments to hedge FX risk (forwards, derivatives) are lacking.
- Existing prudential FX measures:
  - BOM limits on net open foreign exchange positions: limits set at 15 percent of capital for a single currency with an overall open position limit set at 40 percent of capital.
- Identified gaps:
  - Limited market awareness of active analysis and management of FX risk.
  - Some regulations remain currency-blind despite risks from dollarization and flexible exchange rate regime.
- Proposed FX-related tools:
  - Reserve Requirement: unremunerated reserve requirement should vary by currency, with togrog deposits carrying a lower reserve requirement than foreign currencies.
  - Currency-Matching Regulation: calculate open positions by maturities via a maturity ladder for local and foreign currency; impose limits on open positions in each period.
  - Requirements for foreign exchange loans: FX loans to non-foreign exchange earners should be subject to more stringent capital requirements (example from other countries: a 150 percent weight for such loans when calculating capital adequacy ratio).
- Suggested limit example: a limit of 100 percent on the ratio of USD denominated loans to USD denominated deposits would help limit liquidity risk inherent to dollar lending.

---

### Discretionary versus automatic macroprudential measures
- Political economy challenge: decisions by FSC/BOM to activate macroprudential policies will likely face resistance from financial sector and political actors during good times.
- Advantages of automatic/stable rules:
  - Automatic stabilizers reduce need for continual justification, leave less room for error, and minimize political pressure (citing Borio and Drehmann (2009) argument preserved).
  - Credibly pre-committed rules can better enforce buildup of provisions, capital reserves, and liquidity buffers in good times, reducing policy inertia.
- Role for discretion:
  - Supervisors should retain discretionary powers to address idiosyncratic risks and to require bank-specific buffers beyond regulatory minima when warranted.
  - Supervisory interventions: periodic stress tests with updated macro and sector scenarios and early warning indicators; require banks to adjust risk-absorbing buffers accordingly.

---

### Interaction of macroprudential and monetary policies
- Evolving view: monetary policy should continue to focus on inflation; macroprudential tools should target systemic risk. Monetary policy is too blunt to address financial imbalances alone.
- IMF recommendation noted: when credit booms coincide with general overheating, monetary policy should act first and foremost.
- Complementarities and tensions:
  - Price stability and financial stability are mutually reinforcing; sound monetary policy is prerequisite for financial stability.
  - Macroprudential policy is more targeted but works through similar transmission channels as monetary policy (balance sheet and bank lending channels).
  - Potential conflict: macroprudential tightening can contract aggregate activity and exert downward pressure on prices; low interest rates can encourage risk-taking and excessive credit growth.
- Research implication: need for research and statistical analysis to understand complementarities and potential conflicts between macroprudential and monetary policies.

---

### Concluding remarks
- Mongolia’s high growth trajectory associated with very strong demand for credit; combination of factors (loose fiscal policy, rising credit activity, heightened risk appetite from commodity boom) fuel price volatility in asset markets and system-wide financial risks.
- Compounding vulnerabilities: rising interconnectedness, potential increase in dollarization, and concentrated exposures.
- Policy stance:
  - Need to limit emerging system-wide financial risks through enhanced macroprudential framework.
  - Macroprudential tools should prevent buildup of systemic risk and constrain banking sector leverage but are not substitutes for necessary fiscal and monetary adjustments.
  - Sound and prudent monetary policy is essential for financial stability and for maintaining the effectiveness of the monetary transmission mechanism.
  - Maintaining banking system health will help preserve price stability.

_Italic: Content derived from the supplied IMF PDF chapter/section._

### 2009. To enable countries to proactively mitigate the risks from procyclicality (or rising common

### _wp1318 - 2009. To enable countries to proactively mitigate the risks from procyclicality (or rising common exposure)

### Measuring systemic risk: limitations and core dimensions
- Deficiencies in measuring systemic risk are numerous; advances in risk measures and models had led many to believe risks were under control, but the global financial crisis revealed important unmeasured risks were often ignored.
- Main dimensions of the measurement problem:
  - (i) short-sighted risk measures, driven mainly by concurrent volatility;
  - (ii) insufficient recognition of the buildup of latent, endogenous systemic instability;
  - (iii) non-measurable risks (See Huang, Zhou and Zhu, 2009).25
- Market prices may not always be timely indicators of actual systemic risk, particularly in periods of calm or high stress.
- Financial stability analysis often focuses on identifying leading indicators rather than attempting to measure systemic risk directly; a parsimonious set is commonly derived from the behavior of credit and asset prices (see Borio and Drehmann, 2009).
- Evidence: sustained rapid credit growth combined with large increases in property prices increases the probability of an episode of financial instability.
- Key surveillance challenges for Mongolia:
  - Differentiate between credit expansion due to successful financial deepening and credit expansion suggestive of imprudent borrowing.
  - Obtain a good understanding of credit—who is borrowing, how much and why—as a basic building block of macro-prudential surveillance.
  - Analyze terms of credit (e.g., length) in aggregate and by industry for insights into evolving risk.
- Sectoral balance sheet analysis (household, corporate, public) is more data-intensive but more useful:
  - Tracking debt and debt-servicing requirements over time aims to anticipate potential for higher default if growth falters.
  - Analysis of state and local government balance sheets may be rewarding if doubts exist over their debt servicing capabilities and central government backing.
  - Such estimates may be more difficult in emerging market economies due to limitations in national accounts data.

### Early Warning Systems (EWS): purpose and design
- Surveillance is only useful if it provides early warning of potential problems; crises can be triggered by macroeconomic, financial, or geo-political shocks—some unforeseen, others realizations of known risks.
- Systemic risk increases can present as:
  - higher probability of a shock (e.g., housing price "bust" when housing prices exceed fundamentals), or
  - expansion in the number and/or size of underlying vulnerabilities (e.g., higher financial leverage that converts a price shock into adverse real effects).26
- An EWS needs to:
  - organize surveillance to identify vulnerabilities and rank them as threats to financial stability;
  - map problems from different angles and drill down to underlying issues to make constructive contributions to policy debate.
- No EWS is a fail-safe crisis prediction device; multiple angles and drilling down help policymakers assess threats and prioritize actions.

### Measuring latent financial instability: comparative assessment of approaches
- Borio and Drehmann (2009) assess tools along three dimensions:
  - (i) extent to which leading indicators are used, as opposed to contemporaneous indicators;
  - (ii) extent to which behavioral interactions that amplify systemic distress are considered;
  - (iii) extent to which approaches "tell a story" about transmission mechanisms of financial distress.
- Overall conclusion: the usefulness of most tools is generally limited; tools typically provide too little lead time for adequate remedial action.
- Notable exceptions with better lead characteristics:
  - deviations from long-term trend in the credit/GDP ratio;
  - deviations from long-term trend in real estate prices and equity prices.
- Summary findings for specific approaches (as compiled from Borio and Drehmann, 2009):
  - Financial Soundness Indicators (FSIs) or related indices
    - (i) Leading Indicator? No. Backward looking or contemporaneous.
    - (ii) Behavioral Interactions? No
    - (iii) Does it Help to "Tell a Story"? No
  - Credit ratings
    - (i) Forward looking in theory, downgrades "sticky" in practice.
    - (ii) Behavioral Interactions? No
    - (iii) Does it Help to "Tell a Story"? No
  - Market price-based indicators
    - (i) Narrow coverage.
    - (ii) Market biases embedded. Lead too short for policy action.
    - (iii) Does it Help to "Tell a Story"? No
  - Early warning indicators (EWIs)
    - (i) Most promising: those drawing on endogenous cycle perspective. Good EWIs: deviations from long-term trend in credit/GDP and in deflated real estate and equity prices.
    - (ii) Can help identify interactions among risk factors and suggest system-endogenous effects.
    - (iii) Can help frame broad stories.
  - Single-module measures: VaRs
    - (i) Can provide forecasts of financial distress. But unable to incorporate boom-bust cycles.
    - (ii) Take into account feedback effects and can trace shock propagation. But macro-financial linkages are poorly dld.
    - (iii) Lack of structure, and too simplistic. Little to say about the dynamics of distress.
  - Multiple-module measures: Macro stress tests
    - (i) Explicitly forward-looking, can cover a broad range of scenarios. But failed to anticipate the recent turmoil.
    - (ii) Can trace shock propagation. Do not capture systemic-endogenous instability.
    - (iii) Much more structure and granularity. Helpful for storytelling and communicating concerns.

### Policy implications and practical recommendations
- Use EWIs—especially deviations from long-term trends in credit/GDP and in deflated real estate and equity prices—to gain lead time for policy action.
- Strengthen data collection and analysis of sectoral balance sheets (household, corporate, public; state and local government where relevant) to assess debt levels and debt-servicing capacities over time.
- Emphasize understanding credit composition and terms (borrower identity, amounts, purpose, maturity/length) to distinguish beneficial financial deepening from imprudent credit growth.
- Design EWS frameworks that map vulnerabilities from multiple angles and rank threats to financial stability to support prioritized policy responses.
- Recognize limitations of individual tools (FSIs, credit ratings, market-price indicators, VaRs, stress tests) and employ a mix of approaches to "tell a story" about transmission mechanisms and to capture system-endogenous effects where possible.

*Source: Excerpt from _wp1318 (2009) content provided.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1318.pdf_