## 1. Exchange Rate Depreciation, Inflation, and Monetary Policy Rate

## Source details

**Canonical URL:** [1. Exchange Rate Depreciation, Inflation, and Monetary Policy Rate](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16134.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16134.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16134.pdf.json)

---

### Introduction
- Monetary policy tightened after inflation breached the upper limit of the target range in 2015.
- The Colombian exchange rate depreciated around 76 percent since 2014Q2, contributing to inflation of 6.4 percent by end-2015.
- The authorities raised monetary policy rates by 200 bps to 6.5 percent since September 2015.
- Paper objectives:
  - Estimate the nominal neutral interest rate and assess the monetary policy gap.
  - Construct a Monetary Conditions Index (MCI) and Financial Conditions Indicator (FCI).
  - Estimate a Macro-Risk Premium (MRP) linked to intermediaries’ balance sheet data.
- High-level assessment:
  - Traditional price-based indices (MCI and FCI) indicate loosening since mid-2014.
  - The MRP suggests increasing risk premia and a moderation in credit and domestic demand growth.

### Neutral Interest Rate
- Staff estimates place the nominal neutral interest rate for Colombia between 5.9 and 8.4 percent.
- Four methodologies used: Uncovered Interest Parity; Expected-Inflation Augmented Taylor Rule; Consumption-based CAPM (standard consumption-smoothing model); Hodrick-Prescott (HP) filter on interest rates.
- Key figures (preserving source numbers):
  - Expected Inflation: 4.5
  - Actual Monetary Policy Rate: 6.5
  - Uncovered Interest Parity — Neutral Real Interest Rate (NRIR): 3.1; Neutral Nominal Interest Rate (NNIR): 7.6; Nominal Monetary Policy Gap (bps): 110
  - Expected-Inflation Augmented Taylor Rule — NRIR: 1.3; NNIR: 5.8; Nominal Monetary Policy Gap (bps): -66
  - Consumption Based CAPM — NRIR: 3.9; NNIR: 8.4; Nominal Monetary Policy Gap (bps): 190
  - HP-Filter — NRIR: 1.4; NNIR: 5.9; Nominal Monetary Policy Gap (bps): -60
  - Average — NRIR: 2.4; NNIR: 6.9; Nominal Monetary Policy Gap (bps): 43.5
- Implication: with inflation outside the target range and risks to expectations, the central bank has room to further increase the policy rate.

### Monetary Conditions Index (MCI)
- Definition and construction:
  - MCI_t = 100 + w1 (r_t − r_0) + w2 (q_t − q_0), where r is real short-term interest rate and q is the real effective exchange rate; reference period 2005Q1.
  - Weighting: ratio w1/w2 = 3 derived from a VAR analysis and consistent with literature.
- Key findings:
  - The MCI indicates accommodative monetary conditions in 2015.
  - Monetary conditions have been easing steadily since 2014Q1.
  - Despite policy rate hikes, the sharp currency depreciation caused net easing in the last two quarters of 2015.
- Caveat: MCI does not distinguish whether REER depreciation reflects easing monetary conditions or a fundamentals-driven adjustment (e.g., terms of trade shock).

### Financial Conditions Index (FCI)
- Two approaches:
  - VAR-FCI: weights from estimated impact of financial indicators on real GDP in a VAR/structural model.
  - PC-FCI: principal-component approach estimating a common factor from multiple financial variables.
- Correlations with GDP growth (2001Q3–2015Q2 and related horizons):
  - VAR-FCI contemporaneous correlation with annualized QoQ growth: 0.46
  - VAR-FCI correlation with two-quarter-ahead growth: 0.34
  - PC-FCI contemporaneous correlation with growth: uncorrelated
  - PC-FCI correlation with two-quarter-ahead growth: 0.33
- Trajectory and drivers:
  - Both FCIs broadly follow the same trajectory and point to accommodative financial conditions by end-2015.
  - Financial conditions eased continuously from 2014Q3 until 2015Q3, reaching the highest level since 2010.
  - VAR-FCI decomposition shows the currency depreciation as the largest contributor to easing.
  - By end-2015, financial conditions began reacting mildly to policy rate hikes.
- Correlations with Real GDP annualized QoQ percent change at t+1, t+2, t+3, t+4 (table values):
  - FCI: 0.45, 0.36, 0.34, 0.26, 0.30
  - Lending Rate: -0.17, -0.15, -0.14, -0.10, -0.05
  - Policy Rate: -0.06, -0.14, -0.18, -0.19, -0.18
  - Real exchange rate: -0.04, -0.06, -0.10, -0.09, -0.08
  - EMBI: -0.33, -0.13, 0.00, 0.05, 0.10
  - SP500: 0.10, -0.04, -0.12, -0.16, -0.23
  - Stock Exchange: 0.13, 0.05, -0.02, -0.05, -0.10
  - VIX: -0.40, -0.26, -0.15, -0.21, -0.13
- Principal-component FCI factor loadings:
  - S&P 500 and VIX indices drive the PC-FCI with factor loadings of more than 80 percent.
  - Example: factor loading of 0.8 for the EMBI spread implies the common factor explains 80 percent of the variance in Colombia’s sovereign spreads.
  - Domestic variables with significant loadings: EMBI spread, lending rate, policy rate.
- Interpretation:
  - Higher S&P 500 and lower VIX → easier financial conditions.
  - Lower policy and lending rates, lower EMBI spreads, and REER appreciation → better financial conditions.

### Macro-Risk Premium (MRP)
- Conceptual basis: Adrian, Moench, and Shin (2010) — intermediaries’ balance-sheet management affects the macro risk premium; when intermediaries are less capital constrained, risk appetite increases and MRP decreases, boosting credit supply.
- Methodology:
  - Estimate MRP by correlating risk premia with GDP and regressing MRP on measures of credit growth (total credit, commercial credit, consumer credit, mortgages).
- Empirical findings:
  - MRP is negatively correlated with GDP growth.
  - Higher credit growth is associated with lower MRP.
  - The MRP has been increasing since 2014Q3.
  - Drivers of the MRP increase: higher sovereign spreads, higher policy rates, and a steeper yield curve.
  - The increase in spreads is associated with decreases in slack of financial intermediary balance sheet capacity, reducing credit supply.
  - Unlike sovereign spreads, lending spreads have remained broadly stable, which has contained the increase in the MRP.
- Implication: expect a further slowdown in credit growth as monetary tightening reduces banks’ risk-taking capacity and lending spreads increase.

### Conclusion (monetary/financial conditions)
- Traditional monetary and financial conditions indices (MCI and FCI) imply an accommodative stance in 2015 that started reversing in early 2016.
- Price-based indices are useful for high-frequency forecasting but may overweight exchange rate movements and miss balance-sheet information.
- The MRP, which excludes the exchange rate and incorporates balance-sheet data, has been increasing since 2014Q3—indicating declining risk appetite of financial intermediaries and consistent with a slight moderation in credit growth.
- Ongoing monetary policy tightening is expected to further reduce banks’ risk-taking capacity and contribute to a slowdown in credit growth, consistent with the envisaged slowdown in aggregate demand in 2016.

*Source: _cr16134 - 1. Exchange Rate Depreciation, Inflation, and Monetary Policy Rate (IMF staff report, April 18, 2016).*

---

### Annex I. Financial Conditions Index: Methodologies
- VAR method:
  - FCI_t = Σ w_i (x_it − x̄_i) where x_it is the value of financial variable i in period t and x̄_i is its sample mean (2001Q3−2015Q2).
  - Weights reflect cumulative impulse responses (between 2−4 quarters) of GDP growth to a one-unit shock in each financial variable from a recursive VAR including financial variables, quarterly GDP growth, and the GDP deflator; identification via Cholesky decomposition.
  - Financial variables considered: lending rates, policy rates, EMBI spread, VIX, S&P 500, real effective exchange rate, domestic general stock exchange index.
- Factor Analysis (PC-FCI):
  - Step 1: Extract common factor from vector of financial variables: X_t = λ C_t + ε_t.
  - Step 2: Regress the common factor on economic activity variables and take residuals to address endogeneity: C_t = A(L) y_t + η_t; use residual η_t as the FCI (exogenous financial developments influencing future economic activity).

*Source: Annex I. Financial Conditions Index: Methodologies (IMF staff).*

---

### Macro-Risk Premium: Methodology (Annex II summary)
- Two-stage methodology (following Adrian, Moench, and Shin, 2010):
  - Stage 1: Estimate MRP by regressing real GDP growth y_t on financial spreads and other financial variables (EMBI spreads, bank lending-deposit spread, term premium, 5 year yield, monetary policy rate) using quarterly data from 2003 until 2015. The estimated MRP is the fitted value scaled to match the average level and volatility of EMBI spreads.
  - Stage 2: Regress estimated MRP on credit growth measures for total credit, commercial credit, consumption credit, and mortgages to explain MRP with changes in financial intermediaries’ balance-sheet data.

*Source: Annex II summary (IMF staff).*

---

### Corporate Vulnerabilities and Macrofinancial Implications (Colombia)
### High-level findings and mechanisms
- Nonfinancial corporate debt and leverage have increased in recent years, supported by easy access to capital markets, abundant global liquidity, and low interest rates.
- Some sectors more strained: oil, gas and airlines.
- Debt-servicing capacity improved with recent economic growth.
- Static shocks on earnings and FX depreciation for corporates do not significantly impact banking system soundness or overall macrofinancial stability but lead to lower credit growth and would impact GDP.
- Policy recommendations:
  - Enhance the quality and coverage of nonfinancial corporate data.
  - Improve quality of bank capital.
  - Maintain supervisory surveillance of corporate data and risk management practices, especially for FX shock impact and corporate hedges.

### Macrofinancial loops and channels
- Corporate sector measures:
  - Corporate debt represents around 45 percent of GDP.
  - Loans outstanding for private corporate sector stand at 52 percent of total credit issued by financial intermediaries.
- Vicious and virtuous loops described where corporate distress (or strength) affects bank balance sheets, credit growth, investment, and GDP.

### Stylized facts and metrics (as of June 2015 and 2015 totals)
- As of June 2015:
  - Debt of private sector corporates rose to around COP 183 billion pesos, representing about 52 percent of credit institutions total loan portfolio.
- Total private sector corporate debt (domestic and foreign, loans and securities) in 2015:
  - 44.5 percent of GDP (COP 344.5 billion), up 2.2 percentage points since 2014.
- Currency composition:
  - "70 percent of total debt equivalent to 31.2 of GDP in 2015."
  - Since 2013, U.S. dollar indebtedness as a percent of GDP has risen by 3.6 percentage points driven mainly by devaluation (3.4 percentage points) rather than increased U.S. dollar funding (0.2 percentage points of GDP).
- Funding sources:
  - Principal sources 2000–2015: loans with financial institutions both domestic and abroad; least used: bonds issued abroad and foreign credit providers.
  - Since 2003, substitution of foreign financial institution loans by domestic institution loans strengthened domestic macrofinancial linkages.

### Corporate balance-sheet and risk metrics
- Leverage and cash:
  - Gross corporate leverage increased; net leverage increased by about 10 percentage points between 2008 and 2014.
  - For large firms, total debt to total equity increased somewhat since 2008; for medium-sized firms, gross leverage increased more significantly.
  - Sectors with relatively high gross leverage compared to 2008: Airlines, Gas Utilities and Oil and Gas.
- Debt servicing capacity:
  - Earnings grew faster than interest expense from 2008 to 2014.
  - Interest Coverage Ratio (ICR = EBIT / Interest Expense) increased; median ICR across firms remains strong overall.
  - Some sectors with high leverage (Airlines and Utilities) still suffer from low ICRs.
- Default probabilities:
  - Nonfinancial corporate probability of default increased (per Moody’s KMV CreditEdge); banks’ own default frequencies increased but remain at much lower levels.

*Source: IMF staff analysis (Colombian corporate vulnerability analysis).*

---

### Debt-at-Risk (DAR) and Large-Firm Findings
- DAR trends (IMF large-corporate sample of 100 firms):
  - DAR fell over 2008−2014, reflecting robust domestic growth.
  - This contrasts with BanRep evidence showing DAR relatively flat around 25−30 percent.
  - IMF sample likely biased toward stronger-balance-sheet firms.
  - Under a 2015 scenario (earnings growth 10 percent and accounting for further depreciation 30 percent in 2015), DAR would increase from 5 to 12 percent for the IMF sample.
- Leverage and profitability dynamics:
  - For large firms, Debt/EBITDA almost doubled from 2008 to 2014.
  - Sectors with most year-on-year leverage growth: Airlines, Utilities, Oil and Gas.
  - Airlines and Utilities Debt/EBITDA surpass commonly suggested level of four.
- Debt distributions:
  - Debt-to-equity and debt-earnings have grown since 2008.
  - Debt/equity ratios show bi-modality in 2014 (a hump of highly indebted firms and a hump of firms with very little debt).
- Debt-servicing capacity (ICR):
  - Improvement between 2008 and 2014 with a growing mass of firms with ICR>3 in 2014.
  - Debt servicing likely suffered in 2015 with FX depreciation and slow growth and could continue to deteriorate in 2016.
- Corporate Financial Soundness Index (FSI):
  - BanRep FSI (2000−14) uses 34 metrics grouped into activity, leverage, profitability, liquidity and size via principal component analysis.
  - 2014 industry variation: strong — Accommodation and food service; weaker — Agriculture and fishing, mining, manufacturing, real estate, transport and storage, electricity/gas/water supplies, and (in past) financial intermediation.
- FSI and credit quality:
  - Defaulting firms median FSI: -0.33; Non-defaulting firms median FSI: -0.09 (2014 medians).
  - QI (risk loans / total gross loans) declined for both strong and weak firms since 2009; decline was less for stronger firms.
- Implication: as the economy slows, credit deterioration could be much greater for weaker firms due to limited loss-absorption capacity.

*Source: IMF staff analysis (Debt-at-Risk and FSI findings).*

---

### Staff Stress/Sensitivity Tests — Severe but Plausible Shocks
- Sample and assumptions:
  - Sample of firms with balance sheet information and data from 10 banks.
  - Severe but plausible shock assumptions:
    - 20 percent decline in earnings (median GFC change).
    - Currency depreciation against the U.S. dollar of 70 percent.
    - Corporate EBIT exposure to foreign currency of 60 percent.
    - Hedging assumption: financial hedges offset 10 percent of EBIT exposure.
  - DAR and NPL derivation:
    - ICR threshold for DAR: 2.00x.
    - Corporate Loans as 65 percent of Total Loans (weighted average of bank sample).
    - Probability of Default (PD): 25 percent.
    - Loss Given Default (LGD): 60 percent.
- Banking sector context:
  - Corporate lending accounts for around 65 percent (weighted average) of total loans for the 10-bank sample.
  - Total gross NPLs for the Colombian banking system: 2.83 percent in 2015.
  - For the 10 banks in the sample:
    - Median regulatory capital pre-shock: around 14.3 percent.
    - NPLs: 2.5 percent.
  - Countercyclical provisions: around 13−28 percent of total provisions depending on the bank.
  - Risk-weighted assets (RWA): 80 percent of total assets in Colombia; Latin America average ≈ 70 percent.
- Stress-test outcomes for the 10-bank sample:
  - Earnings shock alone:
    - No bank falls below regulatory minimum of 9 percent.
    - Total median capital falls to 13.6 percent (from 14.5 percent pre-shock).
  - Currency depreciation shock (60 percent EBIT FX exposure):
    - No bank falls below regulatory minimum of 9 percent.
    - Total median capital falls to 12.5 percent (from 14.5 percent).
  - Earnings and currency shocks jointly:
    - Two banks fall just below regulatory minimum of 9 percent.
    - Total median capital falls to 11.4 percent (from 14.5 percent).
    - Banco de La Republica static stress test also indicated one bank falls below the regulatory minimum.
  - Including forward-looking provisions and net income:
    - No bank falls below regulatory minimum of 9 percent.
    - Total median capital rises from 11.4 percent to 13.1 percent when buffers are included.

*Source: IMF staff stress-test summary (10-bank sample).*

---

### Stress Tests, Capital Measures, and Real-Sector Linkages
- Treatment of legacy goodwill and intangibles; Basel III sequencing:
  - Analysis assumes legacy capital recorded as goodwill/intangibles acts as fully-loss absorbing capital.
  - SFC regulations since 23 August 2012 require intangibles to be deducted from capital.
  - SFC exercises: deduction of intangibles registered before August 2012 would reduce total capital ratio by 0.7 percent; all credit institutions would continue fulfilling minimum capital ratios.
  - Since 2013 bank capital has improved transitioning to Basel III definition; sequencing to implement Pillar 2, systemic and countercyclical capital will add further loss absorbency.
  - Decree 2392 of 2015 empowered SFC to require additional capital levels; SFC issued stress test framework instructions.
- Vulnerabilities under standardized capital measures:
  - Under tangible common equity to RWA (excluding goodwill) Colombian bank capital tends to be at lower starting positions than many Latin American peers.
  - Under severe stress with standardized measures, Colombian banking system is second worst performer after Brazil in post-stress solvency.
  - Contributing factors: low standardized capital, relatively higher credit costs, limited efficiency but high earnings generation capacity.
  - SFC study: changing loans’ risk weights to LA average, deducting all intangibles from CET1 and switching CIC for Basel III countercyclical buffer might improve capital ratio by 3.6 percent.
- Banco de La Republica static stress test highlights:
  - Corporate test: fragile firms rise from 24.3% to 33.8% of sample under shock; debt of fragile firms increases from 30.1% to 41.1% of sample.
  - Static credit and market risk test: system solvency reduced by 0.5% (15.45% to 14.98%); 1 bank below regulatory minimum.
  - Static liquidity test: under joint shocks, banks’ IRL falls from 295 to 258; liquid assets remain >2.5 times net liquidity requirement.
- Linkages to real sector:
  - Depreciation shock increases Net Leverage by 7.85 percentage points.
  - Using Li, Magud and Valencia (2015) coefficients:
    - Reduction in investment rates: -0.50 percentage points (High Coefficient -0.064).
    - Reduction in investment rates: -0.33 percentage points (Low Coefficient -0.042).
  - With elasticity of real GDP growth to investment rate ≈ 0.1:
    - A reduction of investment rate between 0.33−0.5 would lower real GDP between 0.033–0.05.
- Bank capital, credit growth, and GDP quantitative impacts:
  - Literature on a 1-ppt shock to bank capital on lending/credit growth shows heterogeneous estimates (examples listed with exact values in source).
  - Earnings and depreciation shocks adversely impact NPLs and bank capital by around 1.48 percentage points; this may reduce bank credit growth by 0.09−4.44 percentage points.
  - Midpoint reduction in credit growth: 1.19 percentage points.
  - Using empirical elasticities between real GDP and bank credit growth (0−0.4), a 1.19 ppt reduction in bank credit growth could lower real GDP by 0−0.48 percentage points.
  - Adding the investment impact on GDP could add at maximum 0.05 percentage points, giving a combined maximum value of 0.53 percentage points.
- Caveats and potential for larger impacts:
  - The 0.53 percentage point GDP impact could be an underestimate due to larger shocks, omitted non-bank credit institutions, second-round effects, and dynamic adverse scenarios.
  - Funding linkages between collective investment funds and banks imply potential for solvency concerns to transform into liquidity problems and contagion.

### Policy recommendations and supervisory actions
- Increase quality and coverage of Colombian nonfinancial corporate data tracked by the Corporate Superintendency; IFRS move by firms will help.
- Increase supervisory capacity of the Corporate Superintendency and coordinate with BdR and Financial Superintendency.
- BdR should regularly update corporate FSI to link individual corporate data to a composite nonfinancial corporate vulnerability indicator at firm and industry level; provide more granular industrial breakdowns where sample-size issues are addressed.
- Bank capital should be increased early if risk-based supervision calls for it; Decree 2392 (December 2015) allowed Basel III compliant hybrid bonds.
- Raise equity capital via issuance or organic growth (reduce dividend payout) to add higher-quality loss absorbency.
- Financial Superintendency can limit dividend payout to increase organic capital growth (Colombian banks have higher payouts than Latin American peers).
- Maintain flexible and coordinated policy action to address real, fiscal, external, and financial imbalances to prevent a vicious corporate macrofinancial loop.

*Source: IMF staff analysis in _cr16134 (Colombia) — extracted chapter text.*

---

### Mark-up Model, Wages, Inflation Expectations, and Recent Inflation Dynamics
### Mark-up model and wages
- Model framework: stylized mark-up model where inflation is a function of domestic costs (wages) and imported input costs (NEER); estimated quarterly for the inflation-targeting period and longer sample 1990−2015.
- Key estimation results (Table 1: Stylized Mark-Up Model):
  - Full sample:
    - Wages: 1.11***
    - NEER: -0.10**
    - R2: 0.21
    - Obs: 96
  - IT period:
    - Wages: -0.20
    - NEER: -0.05***
    - R2: 0.36
    - Obs: 64
- Interpretation:
  - Wages are not significant in the inflation-targeting (IT) period but are significant in the longer sample.
  - ERPT implied by this model in the IT period is statistically significant and of similar magnitude to earlier estimates.
- Caution: model does not include real economic activity variable and does not allow for a lag structure.

### Role of inflation expectations
- Augmented mark-up model results (Table 2: The Role of Inflation Expectations):
  - Column (1):
    - NEER: -0.048**
    - Expectations t-4: 0.786***
    - R2: 0.23
    - Obs: 45
  - Column (2):
    - NEER: -0.048***
    - Expectations t-1: 1.400***
    - R2: 0.64
    - Obs: 48
- Conclusion: inflation expectations play a key role in explaining inflation dynamics and are important inputs for monetary policy.

### Recent inflation developments and policy response
- Observations:
  - "Inflation rose sharply in 2015 and remained at 8 percent in March."
  - "24-month-ahead inflation expectations are within the central bank’s target band (2−4 percent) but 12-month ahead expectations are not. Market-based expectations are also above the target band."
- Central bank action:
  - Policy rate raised to 6.5 percent in March, up from 4.50 percent a year earlier.
  - Objective: ensure convergence of inflation to the target band in 2017 and reduce the current account deficit.
  - Policy trade-off: raised rates weighing price pressures and indexation mechanisms against risk of a more pronounced slowdown in domestic demand.
- Drivers and projections:
  - Combined contribution from exchange rate depreciation and El Niño: up to 320 bps to inflation (temporary).
  - Exchange rate depreciation in 2015 contributed around 120bps to inflation in the first quarter of 2016.
  - El Niño estimated impact on headline inflation: 125−200bps.
  - As exchange rate bottoms out in 2016 with stabilizing oil prices, exchange-rate impact on inflation expected to recede quickly.
  - Food prices expected to fall sharply when wet season begins in second half of the year.
  - Staff projection: inflation expected to start declining in second half of 2016 and converge to target band next year (target band: 2−4 percent).
- Indexation and persistence:
  - No evidence of a strong link between minimum wage increases and inflation, but indexation mechanisms (e.g., health services) increase CPI early in the year with persistent effects.
  - Persistence of the depreciation shock may affect nontradable prices and inflation expectations.
  - Staff attributes expected convergence to tighter monetary policy and declining price pressures from depreciation and El Niño.

*Source: IMF staff analysis (mark-up model, inflation expectations, and recent inflation developments).*

### 1. Exchange Rate Depreciation, Inflation, and Monetary Policy Rate_______________________ 4

### 1. Exchange Rate Depreciation, Inflation, and Monetary Policy Rate

### Introduction
- Monetary policy tightened after inflation breached the upper limit of the target range in 2015.
- The Colombian exchange rate depreciated around 76 percent since 2014Q2, contributing to inflation of 6.4 percent by end-2015.
- The authorities raised monetary policy rates by 200 bps to 6.5 percent since September 2015.
- Paper objectives:
  - Estimate the nominal neutral interest rate and assess the monetary policy gap.
  - Construct a Monetary Conditions Index (MCI) and Financial Conditions Indicator (FCI).
  - Estimate a Macro-Risk Premium (MRP) linked to intermediaries’ balance sheet data.
- High-level assessment: traditional price-based indices (MCI and FCI) indicate loosening since mid-2014, whereas the MRP suggests increasing risk premia and a moderation in credit and domestic demand growth.

### Neutral Interest Rate
- Staff estimates place the nominal neutral interest rate for Colombia between 5.9 and 8.4 percent.
- Four methodologies used:
  - Uncovered Interest Parity
  - Expected-Inflation Augmented Taylor Rule
  - Consumption-based CAPM (standard consumption-smoothing model)
  - Hodrick-Prescott (HP) filter on interest rates
- The nominal monetary policy rate of 6.5 percent is within the neutral range but slightly below the average estimated neutral rate.
- Monetary policy gap definition: the difference between the nominal neutral interest rate and the actual monetary policy rate.
- Table of method estimates (preserving source numbers):
  - Expected Inflation: 4.5
  - Actual Monetary Policy Rate: 6.5
  - Uncovered Interest Parity — Neutral Real Interest Rate (NRIR): 3.1; Neutral Nominal Interest Rate (NNIR): 7.6; Nominal Monetary Policy Gap (bps): 110
  - Expected-Inflation Augmented Taylor Rule — NRIR: 1.3; NNIR: 5.8; Nominal Monetary Policy Gap (bps): -66
  - Consumption Based CAPM — NRIR: 3.9; NNIR: 8.4; Nominal Monetary Policy Gap (bps): 190
  - HP-Filter — NRIR: 1.4; NNIR: 5.9; Nominal Monetary Policy Gap (bps): -60
  - Average — NRIR: 2.4; NNIR: 6.9; Nominal Monetary Policy Gap (bps): 43.5
- Implication: with inflation outside the target range and risks to expectations, the central bank has room to further increase the policy rate.

### Monetary Conditions Index (MCI)
- MCI definition: weighted average of deviations of a short-term interest rate and the real exchange rate from baseline values (reference period 2005Q1).
- Formula structure: MCI_t = 100 + w1 (r_t − r_0) + w2 (q_t − q_0), where r is real short-term interest rate and q is the real effective exchange rate.
- Weighting: ratio w1/w2 = 3 derived from a VAR analysis and consistent with literature.
- Key findings:
  - The MCI indicates accommodative monetary conditions in 2015.
  - Monetary conditions have been easing steadily since 2014Q1.
  - Despite policy rate hikes, the sharp currency depreciation caused net easing in the last two quarters of 2015.
- Caveat: MCI does not distinguish whether REER depreciation reflects easing monetary conditions or a fundamentals-driven adjustment (e.g., terms of trade shock).

### Financial Conditions Index (FCI)
- Two approaches to construct FCIs:
  - VAR-FCI: weights assigned by estimated impact of financial indicators on real GDP in a VAR or structural model.
  - PC-FCI: principal-component approach estimating a common factor from multiple financial variables.
- Correlations with GDP growth (2001Q3–2015Q2 and related horizons):
  - VAR-FCI contemporaneous correlation with annualized QoQ growth: 0.46
  - VAR-FCI correlation with two-quarter-ahead growth: 0.34
  - PC-FCI contemporaneous correlation with growth: uncorrelated
  - PC-FCI correlation with two-quarter-ahead growth: 0.33
- Trajectory and drivers:
  - Both FCIs broadly follow the same trajectory and point to accommodative financial conditions by end-2015.
  - Financial conditions eased continuously from 2014Q3 until 2015Q3, reaching the highest level since 2010.
  - VAR-FCI decomposition shows the currency depreciation as the largest contributor to easing.
  - By end-2015, financial conditions began reacting mildly to policy rate hikes.
- Correlations (table of variable correlations with Real GDP annualized QoQ percent change at t+1, t+2, t+3, t+4):
  - FCI: 0.45, 0.36, 0.34, 0.26, 0.30
  - Lending Rate: -0.17, -0.15, -0.14, -0.10, -0.05
  - Policy Rate: -0.06, -0.14, -0.18, -0.19, -0.18
  - Real exchange rate: -0.04, -0.06, -0.10, -0.09, -0.08
  - EMBI: -0.33, -0.13, 0.00, 0.05, 0.10
  - SP500: 0.10, -0.04, -0.12, -0.16, -0.23
  - Stock Exchange: 0.13, 0.05, -0.02, -0.05, -0.10
  - VIX: -0.40, -0.26, -0.15, -0.21, -0.13
- Principal-component FCI factor loadings:
  - S&P 500 and VIX indices drive the PC-FCI with factor loadings of more than 80 percent.
  - Example: factor loading of 0.8 for the EMBI spread implies the common factor explains 80 percent of the variance in Colombia’s sovereign spreads.
  - Domestic variables with significant loadings: EMBI spread, lending rate, policy rate.
- Interpretation:
  - Higher S&P 500 and lower VIX → easier financial conditions.
  - Lower policy and lending rates, lower EMBI spreads, and REER appreciation → better financial conditions.

### Macro-Risk Premium (MRP)
- Conceptual basis: Adrian, Moench, and Shin (2010) — intermediaries’ balance-sheet management affects the macro risk premium; when intermediaries are less capital constrained, risk appetite increases and MRP decreases, boosting credit supply.
- Methodology:
  - Estimate MRP by correlating risk premia with GDP.
  - Regress MRP on measures of credit growth (total credit, commercial credit, consumer credit, mortgages).
- Empirical findings:
  - MRP is negatively correlated with GDP growth.
  - Higher credit growth is associated with lower MRP.
  - The MRP has been increasing since 2014Q3.
  - Drivers of the MRP increase: higher sovereign spreads, higher policy rates, and a steeper yield curve.
  - The increase in spreads is associated with decreases in slack of financial intermediary balance sheet capacity, reducing credit supply.
  - Unlike sovereign spreads, lending spreads have remained broadly stable, which has contained the increase in the MRP.
- Implication: expect a further slowdown in credit growth as monetary tightening reduces banks’ risk-taking capacity and lending spreads increase.

### Conclusion
- Traditional monetary and financial conditions indices (MCI and FCI) imply an accommodative stance in 2015 that started reversing in early 2016.
- Price-based indices are useful for high-frequency forecasting but may overweight exchange rate movements and miss balance-sheet information.
- The MRP, which excludes the exchange rate and incorporates balance-sheet data, has been increasing since 2014Q3—indicating declining risk appetite of financial intermediaries and consistent with a slight moderation in credit growth.
- Ongoing monetary policy tightening is expected to further reduce banks’ risk-taking capacity and contribute to a slowdown in credit growth, consistent with the envisaged slowdown in aggregate demand in 2016.

*Source: _cr16134 - 1. Exchange Rate Depreciation, Inflation, and Monetary Policy Rate (IMF staff report, April 18, 2016).*

### Annex I. Financial Conditions Index: Methodologies

### Annex I. Financial Conditions Index: Methodologies

### VAR – Method
- FCI in period t is calculated as: ݓݔൌ෍ ݔ൫ ݔെ  where ݔ௝௧ is the value that the financial variable x takes in period t, ݓ௝ denotes the weight attached to the financial variable, and ݔఫഥ is the mean of the variable over the sample period (2001Q3−2015Q2). Financial variables enter the FCI as deviations from the mean.
- Weights reflect the relative importance of each financial variable on output as derived from a VAR model, following Go and Lu (2013) and Onsorio et al. (2011).
- Financial variables considered: lending rates, policy rates, EMBI spread, VIX, S&P 500, real effective exchange rate, and the domestic general stock exchange index.
- Construction of weights:
  - Use cumulative impulse responses (between 2−4 quarters) of GDP growth to a one-unit shock to ݔ௝௧.
  - Estimated from a recursive VAR model consisting of all financial variables, plus quarterly GDP growth and the GDP deflator.
  - Identification of structural shocks via a Cholesky decomposition.

### Factor Analysis (PC-FCI)
- Two-step construction:
  - Step 1: Extract a common factor that captures the greatest common variation in a group of financial variables via ܺܨߚൌߤെܷ൅ where ܺ௧ is a vector of financial variables, ܨ௧ is the unobserved common factor, and ܷ௧ is a vector of errors orthogonal to the common factor.
  - Step 2: Regress the common factor on economic activity variables and take residuals to address endogeneity: ܨܣൌሺܮሻݕ߳൅ where A(L) is the lag operator, ݕ denotes year-on-year GDP growth rate, and ߳௧ is the error term uncorrelated with ݕ௧ and its lags.
- The residual ߳௧ is used as the FCI, reflecting only exogenous developments in financial conditions that influence future economic activity.

### Macro Risk Premium: Methodology (Annex II summary)
- Two-stage methodology (following Adrian, Moench, and Shin, 2010):
  - Stage 1: Estimate Macro Risk Premium (MRP) by regressing real GDP growth ݕ௧ on financial spreads:
    - Regression: ݕ௧ߚ൅ߙൌଵܾ݅݉ܧ௧ߚ൅ଶ݊݁ܮ݌ܵ௧ߚ൅ଷݎܲ݉ݎ݁ܶ௧ߚ൅ସݎݕ5௧ߚ൅ହ݁ݐܴ݈ܽ݋ܲ௧ߝ൅௧
    - Variables: ܾ݅݉ܧ௧ are EMBI spreads, ݊݁ܮ݌ܵ௧ is bank lending-deposit interest rate spread, ݎܲ݉ݎ݁ܶ௧ is term premium (slope of the yield curve), ݎݕ5௧ is a 5 year yield, ݁ݐܴ݈ܽ݋ܲ௧ monetary policy rate.
    - Data: quarterly from 2003 until 2015.
    - The estimated MRP is the fitted value; a linear transformation of regression coefficients is used to match the average level and volatility of the EMBI spreads.
  - Stage 2: Explain the MRP with changes in financial intermediaries’ balance sheet data by regressing estimated MRP ܴܲܯ௧ on credit growth measures ݐ݅݀݁ݎܿ௧ for total credit, commercial credit, consumption credit, and mortgages.

### Corporate Vulnerabilities and Macrofinancial Implications (Colombia)
### High-level findings and mechanisms
- Nonfinancial corporate debt and leverage have increased in recent years, supported by easy access to capital markets, abundant global liquidity, and low interest rates.
- Some sectors more strained: oil, gas and airlines.
- Debt-servicing capacity has improved with recent economic growth.
- Static shocks on earnings and FX depreciation for corporates do not significantly impact banking system soundness or overall macrofinancial stability but lead to lower credit growth and would impact GDP.
- Policy recommendations from analysis:
  - Enhance the quality and coverage of nonfinancial corporate data.
  - Improve quality of bank capital.
  - Maintain supervisory surveillance of corporate data and risk management practices, especially for FX shock impact and corporate hedges.

### Macrofinancial loops and channels
- Corporate sector represents around 45 percent of GDP in corporate debt; loans outstanding for private corporate sector stand at 52 percent of total credit issued by financial intermediaries.
- Vicious and virtuous loops described where corporate distress (or strength) affects bank balance sheets, credit growth, investment, and GDP.

### Transmission of shocks (Table 1 synthesis)
- Depreciation Shock (COP:USD): FX debt servicing increase leads to corporate distress; exporters and import-substitution may benefit.
- Global Monetary Unwind (Higher US interest rates): Higher FX debt servicing, lower capital inflows, lower investment.
- Commodity Price Shock (lower oil and other commodity prices): Oil, gas, metal, mining sectors most impacted with losses, liquidation and defaults; lower investment and layoffs.
- Interest Rate Rises (domestic): Higher loan servicing cost, lower investment, lower demand for corporate credit.
- Lower GDP Growth: Lower demand for corporate products, employment, investment, earnings; adverse for bank income, deposits, and credit demand.

### Colombian Corporate Debt: Stylized Facts and Metrics
### Key statistics and trends
- As of June 2015:
  - Debt of private sector corporates rose to around COP 183 billion pesos, representing about 52 percent of credit institutions total loan portfolio (banks, nonbanks, microcredit balances).
  - Total private sector corporate debt (domestic and foreign, loans and securities) in 2015 was 44.5 percent of GDP (COP 344.5 billion), up 2.2 percentage points since 2014.
- Currency composition:
  - Since 2004, corporates have tended to issue debt in pesos: "70 percent of total debt equivalent to 31.2 of GDP in 2015."
  - Since 2013, U.S. dollar indebtedness as a percent of GDP has risen by 3.6 percentage points driven mainly by devaluation (3.4 percentage points) rather than increased U.S. dollar funding (0.2 percentage points of GDP).
- Funding sources:
  - Principal sources from 2000 to 2015: loans with financial institutions both domestic and abroad.
  - Least used: bonds issued abroad and foreign credit providers.
  - Since 2003, substitution of loans with foreign financial institutions by loans with domestic financial institutions, strengthening domestic macrofinancial linkages.

### Corporate balance-sheet and risk metrics
- Leverage and cash:
  - Gross corporate leverage increased; net leverage (total debt minus cash, divided by total equity) increased by about 10 percentage points between 2008 and 2014.
  - For large firms, total debt to total equity increased somewhat since 2008; for medium-sized firms, gross leverage increased more significantly.
  - Sectors with relatively high gross leverage compared to 2008: Airlines, Gas Utilities and Oil and Gas.
- Debt servicing capacity:
  - Earnings grew faster than interest expense from 2008 to 2014.
  - Interest Coverage Ratio (ICR = EBIT / Interest Expense) increased; median ICR across firms remains strong and sufficient to cover debt interest payments.
  - Some sectors with high leverage (Airlines and Utilities) still suffer from low ICRs; Oil and Gas differs.
- Default probabilities:
  - Nonfinancial corporate probability of default increased (per Moody’s KMV CreditEdge); banks’ own default frequencies increased but remain at much lower levels.

*International Monetary Fund — Annex I and Annex II excerpts and Colombia corporate vulnerability analysis as presented in the source content.*

### 10.      Debt at Risk measures (ICR<2) for large fragile Colombian corporates have fallen over

### 10.      Debt at Risk measures (ICR<2) for large fragile Colombian corporates have fallen over

### Debt-at-Risk (DAR) trends and large-firm sample findings
- DAR for the IMF sample of 100 large corporates fell over 2008−2014, reflecting robust domestic growth (Figure 5b).
- This finding contrasts with Banco de La Republica (BanRep) evidence showing DAR at relatively flat levels of around 25−30 percent.
- The IMF sample likely reflects sample-specificities, including firms with stronger balance sheets.
- Under a 2015 scenario (earnings growth 10 percent and accounting for further depreciation 30 percent in 2015), DAR for the IMF large-corporate sample would increase from 5 to 12 percent.

### Leverage and profitability dynamics
- For large firms, the Debt/EBITDA ratio has almost doubled from 2008 to 2014, reflecting increased leverage rather than decreased profitability.
- Sectors with the most year-on-year growth in leverage: Airlines, Utilities, and Oil and Gas.
- Airlines and Utilities sectors’ Debt/EBITDA ratios surpass the commonly suggested level of four.
- Ratios higher than four or five:
  - Typically signal that a firm is less likely to be able to take on additional debt required to grow the business.
  - Are not necessarily indicative of immediate distress or default.

### Debt distributions and sectoral patterns
- Debt-to-equity and debt-earnings have grown since 2008, from lower starting levels for these large corporates.
- Debt/equity ratios show marked bi-modality in 2014, indicating:
  - A growing hump of highly indebted firms, and
  - A growing hump of firms with very little debt.
- In 2008 the proportion of firms with high debt/equity was much less evident.
- Colombian nonfinancial corporate debt dynamics are driven by sectoral differences influenced by domestic and global factors.

### Debt-servicing capacity (ICR) evolution
- Across all firms there has been an improvement in debt servicing capacity between 2008 and 2014.
- There has been a growing mass of firms with ICR>3 in 2014 compared to 2008.
- However, debt servicing capacity likely suffered in 2015 with FX depreciation and slow growth and could continue to deteriorate in 2016.

### Corporate Financial Soundness Index (FSI) — composition and industry results
- BanRep constructed a corporate FSI for 2000−14 using 34 metrics grouped into five categories: activity, leverage, profitability, liquidity and size, via principal component analysis.
- The corporate FSI can be constructed at firm and industry levels, but (per BanRep) is not constructed at the aggregate level for inter-temporal comparisons.
- Considerable variation by industry in 2014:
  - Strong performance (positive index): Accommodation and food service.
  - Weaker balance sheet and operating performance: Agriculture and fishing, mining, manufacturing, real estate activities, transport and storage, electricity, gas and water supplies, and financial intermediation (in past).
- Continued commodity price shocks and depreciation in 2015, and possible weaker growth into 2016, increase uncertainty and suggest diminished balance sheet scope to absorb macrofinancial spillovers.

### FSI distributions and credit quality linkages
- FSI distributions:
  - Non-defaulting firms’ FSI distributions are more symmetric.
  - Defaulting firms are skewed toward weak (negative FSI).
- 2014 medians:
  - Defaulting firms median FSI: -0.33.
  - Non-defaulting firms median FSI: -0.09.
- Credit quality (Quality Index, QI):
  - QI is the ratio of risk loans (loans firms find difficult to repay or have defaulted on) to total gross loans.
  - Firms with positive FSI consistently have lower QI (better credit worthiness) than those with negative FSI for 2000−13; the relationship is less marked in 2014.
  - QI has declined for both strong and weak firms since 2009; decline was much less for stronger firms.
- Implication: As the economy slows, credit deterioration could be much greater for weaker firms due to limited loss-absorption capacity.

### Staff stress/sensitivity tests — severe but plausible shocks (sample and assumptions)
- Sample and data:
  - Based on a sample of firms with balance sheet information and data from 10 banks.
- Severe but plausible shock assumptions:
  - A 20 percent decline in earnings (similar to median changes in firms’ EBIT across major EM countries during the GFC of 2008).
  - Currency depreciation against the U.S. dollar of 70 percent.
  - Corporate EBIT exposure to foreign currency of 60 percent (assumed mainly through increased import expenses).
- Hedging assumption:
  - Financial hedges offset 10 percent of EBIT exposure through derivatives.
- DAR and resulting NPLs derivation assumptions:
  - ICR threshold for DAR: 2.00x.
  - Corporate Loans as 65 percent of Total Loans (weighted average of bank sample).
  - Probability of Default (PD): 25 percent (based on World Bank recovery data for Latin America).
  - Loss Given Default (LGD): 60 percent (based on Banco de Bogota’s annual report estimates, applied to sample).

### Banking sector impacts from staff stress tests (key statistics and outcomes)
- Corporate lending share:
  - Corporate lending accounts for around 65 percent (weighted average) of total loans for the 10-bank sample.
- Systemwide and sample indicators:
  - Total gross nonperforming loans (NPLs) for the entire Colombian banking system reached 2.83 percent in 2015.
  - For the 10 banks in the sample:
    - Median regulatory capital pre-shock: around 14.3 percent.
    - NPLs: 2.5 percent.
  - Colombia imposes forward-looking expected loss provisioning and countercyclical provisioning.
    - Countercyclical provisions amounted to around 13−28 percent of total provisions depending on the bank in the sample.
  - Risk-weighted assets (RWA) in Colombia represent 80 percent of total assets; the average in the Latin America region is approximately 70 percent.

- Stress-test outcomes for the 10-bank sample:
  - Under the earnings shock alone:
    - No bank falls below the regulatory minimum of 9 percent.
    - Total median capital falls to 13.6 percent (from 14.5 percent pre-shock).
  - Under the currency depreciation shock (with 60 percent EBIT FX exposure):
    - No bank falls below the regulatory minimum of 9 percent.
    - Total median capital falls to 12.5 percent (from 14.5 percent pre-shock).
  - Under earnings and currency shocks jointly:
    - Two banks fall just below the regulatory minimum of 9 percent.
    - Total median capital falls to 11.4 percent (from 14.5 percent pre-shock).
    - Banco de La Republica (BdR) static stress test also indicated that one bank falls below the regulatory minimum.
  - Accounting for additional loss-absorbency buffers (forward-looking provisions and net income):
    - No bank falls below the regulatory minimum of 9 percent.
    - Total median capital for the 10 banks rises from 11.4 percent to 13.1 percent when these buffers are included.

*Source: IMF staff summary of IMF chapter "10. Debt at Risk measures (ICR<2) for large fragile Colombian corporates have fallen over" (figures and data drawn from S&P Capital IQ, Banco de La Republica, and IMF staff calculations).*

### 24.      While the impacts of static adverse earnings and currency depreciation shocks are

### _cr16134 - 24.      While the impacts of static adverse earnings and currency depreciation shocks are

### Stress test findings and capital resilience
- Static adverse earnings and currency depreciation shocks are manageable for the Colombian banking system under the assumptions used.
- Stress tests indicate system-level capital remains well above the 9 percent minimum under baseline and combined shocks.
- Robust provisioning and net income provide additional capital boost; capital buffers are larger at system level.

### Treatment of legacy goodwill and intangibles; Basel III sequencing
- The analysis assumes legacy capital (held before August 23, 2012) recorded as goodwill or intangibles acts as fully-loss absorbing capital similar to equity capital.
- Under Basel III rules goodwill and intangibles are deducted from bank capital; SFC regulations since 23 August 2012 require all intangibles whenever registered to be deducted from capital and do not act as fully-loss absorbing capital.
- SFC exercises show deduction of intangibles registered before August 2012 would reduce the total capital ratio by 0.7 percent and all credit institutions will continue fulfilling the minimum capital ratios.
- Since 2013 Colombian bank capital has been improved further transitioning to the Basel III capital definition by replacing legacy goodwill and intangibles; sequencing moves to fully implement Pillar 2, systemic and countercyclical capital would add further loss absorbency and resilience.
- Decree 2392 of 2015 empowered the SFC to require additional capital levels based on self-assessments of capital adequacy and supervision; SFC also issued instructions about the stress test framework institutions must follow.

### Vulnerabilities under standardized capital measures and rating-agency exercises
- Under standardized higher-quality capital measures (tangible common equity to risk-weighted assets) that exclude goodwill, Colombian bank capital tends to be at lower starting positions than many Latin American peers.
- Under a severe stress scenario using such standardized measures, the Colombian banking system is the second worst performer after Brazil in post-stress solvency.
- Contributing factors: low standardized measure of capital (low starting points), relatively higher credit costs (loan loss provisions over gross loans), limited efficiency (high cost-to-income), but still high earnings generation capacity.
- The SFC study: if loans’ risk weights are changed by the reported average in LA, deducting all intangibles assets from CET1 and switching the CIC (0.8 percent of RWA) for the Basel III countercyclical capital buffer then the capital ratio might improve by 3.6 percent.

### Banco de La Republica static stress test results (summary from Table 2)
- Corporate test: After a 25% shock to numerator/denominator, fragile firms increase from 24.3% to 33.8% of sample; debt of fragile firms increase from 30.1% to 41.1% of sample.
- Static credit and market risk test: All fragile corporate rating downgrades by one grade; 10% of consumer portfolio downgrade by one grade; 200 bp increase in TES and private debt curve; loss in value of shares of 20%. Bank system solvency reduced by 0.5% (15.45% to 14.98%); 1 bank is below regulatory minimum; nonbanks, financial cooperatives; finance companies improve capital by 0.04%-1.11%.
- Static liquidity test: Under joint materialization of shocks, banks’ IRL falls from 295 to 258. Liquid assets remain more than 2.5 times net liquidity requirement for banks.

### Linkages to real sector: corporate distress → investment and GDP
- Stress scenario: Depreciation Shock increases Net Leverage by 7.85 percentage points.
- Using Li, Magud and Valencia (2015) coefficients:
  - Reduction in investment rates: -0.50 percentage points (High Coefficient -0.064).
  - Reduction in investment rates: -0.33 percentage points (Low Coefficient -0.042).
- Reported elasticity of real GDP growth to investment rate ≈ 0.1 across emerging markets. Thus:
  - A reduction of investment rate between 0.33−0.5 would lower real GDP between 0.033–0.05.

### Bank capital, credit growth, and GDP quantitative impacts
- Table 4 summary of literature on effect of a 1-ppt shock to bank capital on lending/credit growth:
  - Bernanke & Lown (1991): Bank capital ratio → Bank lending growth +2-3 (ppt)
  - Berropside & Edge (2010): Bank capital-to-asset ratio → BHC lending growth +0.145 (ppt)
  - Carlson et al (2013): Total capital-to-asset ratio → Bank lending growth +0.13-2 (ppt)
  - Francis & Osborne (2012): Surplus bank capital ratio → Bank lending growth +0.060 (ppt)
  - Gambarcorta & Mistrulli (2004): Excess regulatory capital-to-assets ratio → Bank & credit cooperatives lending growth +0.0744 (ppt)
- The earnings and depreciation shocks studied previously adversely impact NPLs and bank capital by around 1.48 percentage points; this may reduce bank credit (loan) growth by 0.09−4.44 percentage points.
- A midpoint assessment between 0.09−4.44 ppt implies a reduction in credit growth of 1.19 percentage points.
- Using empirical elasticities between real GDP and bank credit growth ranging from 0-0.4, a 1.19 ppt reduction in bank credit growth could lower real GDP in the region of 0−0.48 percentage points.
- Adding the investment impact on GDP could add at maximum 0.05 percentage points, giving a combined maximum value of 0.53 percentage points.

### Caveats and potential for larger impacts
- The 0.53 percentage point GDP impact could be an underestimate because:
  - Earnings and depreciation shocks could be larger.
  - Credit growth reduction in stress tests assumed to arise from banks only; accounting for other credit institutions would imply larger credit growth and real GDP impacts.
  - Results reflect a single round of macrofinancial linkages; second-round effects could produce larger GDP impacts.
  - Static single- or combined-shock exercises ignore dynamic adverse macroeconomic scenarios that could lead to larger impacts over a longer period.
- Funding linkages between collective investment funds and banks imply potential for solvency concerns to transform into liquidity problems and raise contagion worries.

### Policy recommendations and supervisory actions
- Increase the quality and coverage of Colombian nonfinancial corporate data tracked by the Corporate Superintendency to improve timeliness and remove data gaps; IFRS move by firms will help.
- Increase supervisory capacity of the Corporate Superintendency to investigate nonfinancial corporate sector data and determine macrofinancial implications across government, financial, external, and household balance sheets; coordinate with BdR and Financial Superintendency.
- BdR should regularly update corporate financial soundness index (FSI) to link individual corporate data to a composite nonfinancial corporate vulnerability indicator at firm and industry level; provide more detailed and granular industrial breakdowns where sample-size issues are addressed.
- Bank capital should be increased early if risk-based supervision calls for it. Decree 2392 (December 2015) allowed Basel III compliant hybrid bonds; raising equity capital through issuance or organic growth (reducing dividend payout) would add higher-quality loss absorbency.
- Financial Superintendency has powers to further limit dividend payout for Colombian banks to increase organic capital growth (Colombian banks have higher payouts than Latin American peers).
- Continued flexible and coordinated policy action is needed to address multiple economic and financial imbalances (real, fiscal, external, financial) to prevent a vicious nonfinancial corporate macrofinancial loop and to reinforce a virtuous macrofinancial stability loop.

*Source: IMF staff analysis in _cr16134 (Colombia) — extracted chapter text.*

### Annex I. Bank Capital and Credit Growth Study Summary

### Annex I. Bank Capital and Credit Growth Study Summary

### Key findings on the relationship between bank capital and credit growth
- Empirical studies do not yield a uniform message: some studies suggest a strong positive impact of bank capital on credit growth, others a negative impact, and others find no relationship at all.
- Evidence of a positive relationship was most strongly evident during the recent financial crisis:
  - Albertazzi and Marchetti (2010) using Italian data in 2007−09 find evidence of a contraction of credit supply associated with low bank capitalization.
  - Kapan and Minoiu (2013) employ a sample of more than 800 banks from 55 countries during 2006−10 and show that better capitalized banks (with lower leverage ratio) exposed to financial market shocks decreased their supply of loans less than other banks.
  - Gambacorta and Mistrulli (2004) using data for Italian banks in 1992−2001 show that well-capitalized banks can better absorb temporary financial difficulties on the part of their borrowers and preserve long-term lending relationships.
- Mechanisms cited:
  - Higher capital makes the provision of credit more stable and robust even in economic downturns.
  - More capital allows banks to better withstand financial and real shocks.
  - Bank capital increases the capacity to raise non-insured debt and thus banks’ ability to limit the effect of a drop in deposits on lending (Ashcraft, 2001).

### Contrasting evidence and caveats
- Studies focused on banks in advanced economies during the 2008 crisis often find different or mixed results:
  - Huang and Ratnovski (2009) using OECD data find no relationship between pre-crisis bank capital and performance during the crisis.
  - Camara et al. (2010) for a sample of European banks report that well-capitalized banks took more risk before the 2008 crisis.
  - The IMF’s GFSR (2009), using a sample of 36 major global banks, finds that banks that received government support during the crisis had statistically higher capital metrics before the crisis.
- Empirical evidence therefore fails to provide a definitive answer on whether higher capital will always and everywhere enhance financial stability.
- The results and interpretation can depend on sample composition, time period (e.g., crisis vs. non-crisis), and economy type (advanced vs. emerging).

### Authors’ assessment
- Despite mixed empirical findings, the authors of this summary side with the view that:
  - Higher capital will increase banks’ resilience.
  - Higher capital will support credit to productive sectors of the economy.
  - Higher capital will reduce losses in a crisis period.

*Source: Annex I. Bank Capital and Credit Growth Study Summary*

### 15.      In the inflation-targeting period wages do not appear to be a significant driver of

### 15. In the inflation-targeting period wages do not appear to be a significant driver of

### Mark-up model and wages
- Model framework:
  - A stylized mark-up model (Brouwer and Ericsson 1998, Goodfriend 1997) is estimated; in the model inflation is a function of domestic costs (captured by wages) and the cost of imported inputs (captured by the NEER).
  - The source text states: "where ߨ∆ is y/y inflation, ݓ∆ is y/y industrial wage growth, and ܴܧܧܰ∆ is the y/y percent change in the NEER."
  - Estimation frequency: quarterly for the inflation-targeting period and for a longer sample covering 1990−2015.
  - Caveats: "The results should be interpreted with caution since the model does not include a variable to measure real economic activity and does not allow for a lag structure."

- Key estimation results (Table 1: Stylized Mark-Up Model):
  - Full sample:
    - Wages: 1.11***
    - NEER: -0.10**
    - R2: 0.21
    - Obs: 96
  - IT period:
    - Wages: -0.20
    - NEER: -0.05***
    - R2: 0.36
    - Obs: 64
  - Interpretation:
    - Wages are not significant in the inflation-targeting (IT) period but are significant in the longer sample.
    - The ERPT implied by this model in the IT period is statistically significant and of similar magnitude to earlier estimates in the document.

### Role of inflation expectations
- Augmented mark-up model with inflation expectations has significant explanatory power (Table 2).
- Specification details:
  - Column (1): inflation at time t explained by 12-month-ahead inflation expectations at time t-4.
  - Column (2): inflation at time t explained by expectations at t-1.
  - Results (Table 2: The Role of Inflation Expectations):
    - Column (1):
      - NEER: -0.048**
      - Expectations t-4: 0.786***
      - R2: 0.23
      - Obs: 45
    - Column (2):
      - NEER: -0.048***
      - Expectations t-1: 1.400***
      - R2: 0.64
      - Obs: 48
- Conclusion from estimates: inflation expectations play a key role in explaining inflation dynamics and are important inputs for monetary policy.

### Recent inflation developments and policy response
- Observations:
  - "Inflation rose sharply in 2015 and remained at 8 percent in March."
  - "24-month-ahead inflation expectations are within the central bank’s target band (2−4 percent) but 12-month ahead expectations are not. Market-based expectations are also above the target band."
- Central bank action:
  - Policy rate path: raised to 6.5 percent in March, up from 4.50 percent a year earlier.
  - Objective: ensure convergence of inflation to the target band in 2017 and reduce the current account deficit.
  - Policy trade-off: the central bank "raised rates weighing continued price pressures and the activation of indexation mechanisms against the risk of a more pronounced slowdown in domestic demand."

### Drivers of recent inflation and staff projections
- Quantified contributions to inflation:
  - Combined contribution from exchange rate depreciation and El Niño: up to 320 bps to inflation (temporary).
  - Exchange rate depreciation in 2015 contributed around 120bps to inflation in the first quarter of 2016.
  - El Niño estimated impact on headline inflation: 125−200bps.
- Dynamics and outlook:
  - As the exchange rate bottoms out in the course of 2016 in line with stabilizing oil prices, its impact on inflation is expected to recede quickly.
  - Food prices are expected to fall sharply when the wet season begins in the second half of the year.
  - Staff projection: inflation is expected to start declining in the second half of 2016 and converge to the target band next year (target band: 2−4 percent).
- Indexation and persistence:
  - "There is no evidence of a strong link between minimum wage increases and inflation but indexation mechanisms for a number of items such as health services are increasing the CPI early in the year with persistent effects on projected inflation rates in coming months."
  - The persistence of the depreciation shock may affect prices of nontradable goods and inflation expectations.
  - Nevertheless, staff attributes expected convergence to tighter monetary policy and declining price pressures from depreciation and El Niño.

### Methodological and interpretive cautions
- The mark-up model results:
  - Do not include a variable measuring real economic activity.
  - Do not allow for a lag structure.
  - Hence results should be interpreted with caution.

*Source: IMF staff analysis (section content supplied).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16134.pdf_