## _wp1260 - References .............................................................................................................

## Source details

**Canonical URL:** [_wp1260 - References .............................................................................................................](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1260.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1260.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1260.pdf.json)

---

### I. INTRODUCTION
- Over the past decade, large capital inflows and significant foreign exchange (FX) intervention in many emerging market (EM) central banks have led to major shifts in the composition and size of central bank balance sheets (CBBS), including:
  - Large expansions of net foreign assets (NFA) and domestic (interest-bearing) liabilities due to sterilization operations.
  - Wider currency mismatches and a secular decline in central bank capital (interrupted temporarily by 2008 depreciations).
  - Particularly pronounced dynamics in emerging Asia and under inflation targeting (IT) regimes, where lower tolerance for inflation led to reduced seigniorage and revaluation gains.
- Accumulation of foreign-currency assets and short-term local-currency liabilities increased central bank capital sensitivity to short-term domestic interest rate movements via:
  - Higher interest payments on liabilities (with no symmetric effect on asset revenues).
  - Exchange rate movements producing capital losses.
- Two opposing views on central bank financial strength (CBFS) relevance to policy:
  - CBFS is irrelevant because central banks can print money and be recapitalized via seigniorage or through institutional recapitalization agreements with the Treasury (consolidated fiscal position matters).
  - CBFS matters because financial weakness can trigger political-economy pressures, reduce independence, and induce sub-optimal monetary policy to avoid loss of independence (Jeanne and Svensson, 2007).
- Research gap: limited theoretical and empirical exploration of whether CBFS affects monetary policy decisions directly.

### Key empirical focus and headline result
- Objective: Assess empirically whether CBFS constrains monetary policy.
- Approach overview:
  - (i) Define an empirical measure of CBFS.
  - (ii) Construct a proxy for monetary policy constraint (MPC) as deviations from an estimated “optimal policy.”
  - (iii) Test statistical linkage between CBFS and MPC.
- Sample: 41 emerging and advanced market countries over the period 2002:M12011:M3 (alternative reporting 2002:m1 to 2011:m4).
- Main finding: CBFS matters for monetary policy. The effect is nonlinear—statistically significant and economically meaningful for very sub-optimal monetary policies (lower deciles). CBFS explains deviations of up to 72 basis points in policy interest rates when rates are below the “optimal” level.

### II. METHODOLOGY AND DATA

#### A. Central Bank Financial Strength (CBFS)
- Definition: CBFS is defined as the ratio of a broad measure of capital to assets (Capital/Total Assets).
- Rationale for measure:
  - Widely available and standardized monthly from IMF International Financial Statistics (IFS).
  - Reflects the degree of currency mismatch in EM central bank balance sheets (Assets largely in foreign currency; Liabilities largely in domestic currency).
  - Total Assets chosen as denominator to factor in currency mismatch; Capital/Assets can be rewritten as 1-(Domestic Liabilities/Foreign Assets).
- Limitations and caveats:
  - Accounting nature may not capture market values of assets/liabilities or contingent liabilities.
  - "Other Items Net" inclusion desirable as it captures valuation changes and reserves serving as buffers, though idiosyncratic across countries.
  - Adoption of IFRS by some central banks may reduce opacity in valuations.

#### B. Indicator of Constraints on Monetary Policy (MPC)
- MPC proxy: Deviations of observed policy interest rates from an estimated “optimal” level (forecast errors from fitted interest rate rules); emphasis on MPC_t < 0 cases.
- Estimation procedure: three steps
  - Step 1: Estimation of Interest Rate Rules
    - Baseline specification: policy rate depends on expected 12-months-ahead CPI inflation gap (relative to inflation target), 3-month ahead expected output gap, and last-quarter observed exchange rate depreciation (vis-à-vis the US dollar).
    - Alternative proxies for activity: industrial production or unemployment rate; potential output via HP filter or linear-quadratic trend.
    - Number of rule specifications per country: 8.
    - Estimation method: instrumental variable-generalized method of moments (IV-GMM). Instruments: lags of all independent variables and interest rate, plus log of a broad commodity price index.
  - Step 2: Selection Algorithm
    - Out-of-sample forecast performance tested in the spirit of Clark and West (2006, 2007); choose rule that outperforms a random walk at the largest number of horizons (lowest mean p-value of the equal-predictive-ability test).
  - Step 3: Measure of Monetary Policy Constraint (MPC)
    - MPC_t = observed interest rate − predicted interest rate from selected rule (equation (5)).

#### C. Empirical linkage: CBFS → MPC
- Baseline empirical model: pooled ordinary least squares (OLS) with MPC as a function of CBFS (equation (6)).
- Nonlinearity assessed via quantile regressions (Koenker and Hallock (2001)) to capture heterogeneous effects across MPC distribution.
- Intuition: quantile regressions reveal whether CBFS influences MPC differently for countries with large sub-optimal deviations versus average cases.

#### D. Sample and Data
- Sample size: 41 countries.
- Sample periods as reported: "2002:M12011:M3" and "2002:m1 to 2011:m4".
- Data sources: IMF International Financial Statistics (IFS) and Haver Analytics.
- Variables: central bank balance sheet items, policy rates or money market rates, CPI index, industrial production, unemployment rates, exchange rates, broad commodity price index.

### Method to derive the MPC measure (Annex Table A.2)
- Benchmark estimation:
  - Interest rate rule estimated over the pre-crisis period (2002:1-2008:8).
  - MPC variable measured using the dynamic forecast for the out-of-sample period (2008:9-2010:4).
- Robustness checks include panel fixed effects and quantile regressions; sensitivity to sub-period choices explored.

### III. RESULTS

#### III.A. Assessing whether CBFS interferes with the conduct of monetary policy — main findings
- Pooled OLS (Equation (6)), benchmark:
  - Coefficient (pooled OLS, benchmark): 0.0414*** (standard error 0.00905)
  - Constant: -1.520*** (standard error 0.166)
  - Observations: 813
  - R-squared: 0.032
- Interpretation:
  - Lower central bank capital is correlated with a wider (negative) gap between observed and optimal interest rate (the MPC measure).
  - A 1 percent increase in the CBFS measure is associated with an average effect of 4 basis points in the MPC measure.
- Non-linearity documented via quantile regressions (benchmark quantile coefficients and standard errors):
  - q10: 0.0718*** (0.0137)
  - q20: 0.0385*** (0.0136)
  - q30: 0.0300*** (0.0102)
  - q40: 0.0188 (0.0128)
  - q50: 0.0145*** (0.00485)
  - q60: 0.0224*** (0.00612)
  - q70: 0.00949 (0.0104)
  - q80: 0.0265* (0.0139)
  - q90: 0.0378** (0.0164)
- Magnitude example: For the first decile (interest rates below “optimal”), a 10 percent increase in CBFS decreases the size of the interest rate gap by 72 basis points.

#### Heterogeneity by monetary regime and development stage
- Inflation Targeting (IT) countries (EMEs):
  - Pooled OLS: CBFS coefficient: 0.0235 (standard error 0.0184); Constant: 2.861*** (0.249); Observations: 339; R-squared: 0.005.
  - Quantile regression for IT countries: q10: 0.199*** (0.0356) (other quantiles reported in annex).
- Developed economies:
  - Pooled OLS: CBFS coefficient: 0.00453 (0.00515); Constant: -2.025*** (0.120); Observations: 200; R-squared: 0.004.
  - Quantile regression coefficients generally small and insignificant across deciles (selected: q10: -0.000898 (0.00648), q20: 0.00284 (0.00546), q30: 0.00327 (0.00891), q40: 0.00986 (0.0102), q50: 0.00941 (0.0106), q60: 0.00407 (0.0101), q70: 0.00810 (0.0103), q80: 0.00305 (0.00714), q90: 0.000533 (0.00674)).
- Possible interpretation: stronger institutional frameworks or absence of currency mismatches in developed economies may explain null result.

#### III.B. Robustness analysis — checks and outcomes
- Sample period sensitivity:
  - Re-estimated rules for 2002-2006:12 and forecast 2007:1-2008:12; OLS and quantile results qualitatively similar to benchmark but with level effect and wider 95 percent confidence bands.
- In-sample versus out-of-sample forecasts:
  - Using in-sample forecasts and re-estimating Equation (6) yields qualitatively similar results; nonlinearities somewhat more pronounced.
- Sensitivity to CBFS definition:
  - Alternative CBFS: capital-to-total-assets ratio (simpler); non-linearity survives qualitatively, but exclusion of "Other Items Net" complicates interpretation.
- Exchange rate misalignments as controls:
  - Controlled using deviation from a five-year moving average and deviation from trend via Hodrick-Prescott filter; results robust and CBFS remains relevant.
- Reverse causation and country fixed-effects:
  - Re-estimating Equation (6) with three-month lagged CBFS produces qualitatively robust results (estimates slightly smaller).
  - Introducing country fixed-effects does not affect results.

### Overall interpretation of econometric evidence
- CBFS plays a statistically significant role in explaining “suboptimal” monetary policies in a nonlinear manner.
- Quantile regression evidence: large negative deviations from optimal policy (“undershooting”) can be partially explained by weak CBFS.
- Magnitude summary:
  - 1 percent increase in CBFS → average effect of 4 basis points improvement in MPC.
  - 10 percent increase in CBFS at first decile → 72 basis points reduction in the interest rate gap.

### IV. CONCLUSIONS AND DIRECTIONS FOR FURTHER RESEARCH
- Shifts in central bank balance sheet composition and size have made central bank capital sensitive to domestic interest rate movements, raising concerns over implications for monetary policy conduct.
- This paper is the first empirical attempt to analyze the importance of central bank capital for monetary policy conduct.
- Main conclusions:
  - Weak central bank balance sheets can influence monetary policy.
  - The relationship between central bank balance sheets and monetary policy is highly non-linear: large deviations from optimal policy associated with very weak balance sheets.
- Further research recommended:
  - Refine the measure of central bank financial strength to fully capture currency mismatches and interest rate risk while preserving cross-country comparability.
  - Improve understanding of mechanisms through which CBFS influences monetary policy, including macroeconomic and institutional factors (e.g., fiscal arrangements).
  - Apply methodology to examine potential side effects of rapidly expanding central bank balance sheets in advanced economies from non-traditional policies (e.g., quantitative easing).

### Key statistics and methodological particulars (preserve exact figures/terms)
- Sample size: 41 countries.
- Sample periods as reported: "2002:M12011:M3" and "2002:m1 to 2011:m4".
- Number of rule specifications per country: 8.
- Estimation methods: IV-GMM; selection algorithm based on out-of-sample forecast performance (Clark and West framework).
- Main empirical result: CBFS explains deviations of up to 72 basis points in policy interest rates when rates are set below the estimated optimal level.

### Select references cited in the text
- Clarida, Richard, Jordi Galí, and Mark Gertler, 2000, “Monetary Policy Rules and Macroeconomic Stability: Evidence and Some Theory,” Quarterly Journal of Economics, February, pp. 147–80.
- Clark, Todd, and Kenneth West, 2006 and 2007, related methodology on out-of-sample forecast tests.
- Koenker, Roger and Kevin Hallock, 2001, “Quantile Regression,” Journal of Economic Perspectives, Vol. 15, No.4, pp. 143–56.
- Moura, Marcelo, and Alexandre Carvalho, 2010, application of Clark-West selection for interest rate rules.
- Jeanne, Olivier, and Lars E. O. Svensson, 2007, theoretical motivation on central bank balance sheets and policy credibility.
- Ize, Alain, 2005; Ize and Oulidi, 2009; Stella, Peter, 1997 and 2005 — selected IMF work on central bank capital and financial strength.

*Source: Excerpt from the IMF working paper content unit titled "_wp1260 - References ............................................................................................................."*

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

### _wp1260 - References .............................................................................................................

### I. INTRODUCTION
- Over the past decade, large capital inflows and significant foreign exchange (FX) intervention in many emerging market (EM) central banks have led to major shifts in the composition and size of central bank balance sheets (CBBS), including:
  - Large expansions of net foreign assets (NFA) and domestic (interest-bearing) liabilities due to sterilization operations.
  - Wider currency mismatches and a secular decline in central bank capital (interrupted temporarily by 2008 depreciations).
  - Particularly pronounced dynamics in emerging Asia and under inflation targeting (IT) regimes, where lower tolerance for inflation led to reduced seigniorage and revaluation gains.
- The accumulation of foreign-currency assets and short-term local-currency liabilities increased central bank capital sensitivity to short-term domestic interest rate movements via:
  - Higher interest payments on liabilities (with no symmetric effect on asset revenues).
  - Exchange rate movements producing capital losses.
- Two opposing views on central bank financial strength (CBFS) relevance to policy:
  - CBFS is irrelevant because central banks can print money and be recapitalized via seigniorage or through institutional recapitalization agreements with the Treasury (consolidated fiscal position matters).
  - CBFS matters because financial weakness can trigger political-economy pressures, reduce independence, and induce sub-optimal monetary policy to avoid loss of independence (Jeanne and Svensson, 2007).
- Research gap: limited theoretical and empirical exploration of whether CBFS affects monetary policy decisions directly.

### Key empirical focus and headline result
- Objective: Assess empirically whether CBFS constrains monetary policy.
- Approach overview:
  - (i) Define an empirical measure of CBFS.
  - (ii) Construct a proxy for monetary policy constraint (MPC) as deviations from an estimated “optimal policy.”
  - (iii) Test statistical linkage between CBFS and MPC.
- Sample: 41 emerging and advanced market countries over the period 2002:M12011:M3.
- Main finding: CBFS matters for monetary policy. The effect is nonlinear—statistically significant and economically meaningful for very sub-optimal monetary policies (lower deciles). CBFS explains deviations of up to 72 basis points in policy interest rates when rates are below the “optimal” level.

### Paper organization
- Section II: Methodology and data (CBFS measure, MPC estimation, empirical linkage).
- Section III: Results and robustness checks.
- Section IV: Conclusions, limitations, and future research avenues.

### Figures and empirical patterns (as described)
- Figure 1: Dynamics of main CBBS components in EMEs—shows changes in Net Foreign Assets to GDP, Net Domestic Liabilities to GDP, and Capital to GDP and Capital to Total Assets (percent changes and levels for 2002–10 and 2010 NFA/GDP).
- Figure 2: Dynamics of key CBBS items by monetary regime (Inflation Targeting vs Non-Inflation Targeting) — time series for Capital to GDP (Percent), Capital to Total Assets (Percent), Net Foreign Assets to GDP (Percent), and Inflation (Percent).

---

### II. METHODOLOGY AND DATA

#### A. Central Bank Financial Strength (CBFS)
- Definition: CBFS is defined as the ratio of a broad measure of capital to assets (Capital/Total Assets). Formally calculated as in equation (1) in the source.
- Rationale for measure:
  - Widely available and standardized monthly from IMF International Financial Statistics (IFS).
  - Reflects the degree of currency mismatch in EM central bank balance sheets (Assets largely in foreign currency; Liabilities largely in domestic currency).
  - Total Assets chosen as denominator to factor in currency mismatch; Capital/Assets can be rewritten as 1-(Domestic Liabilities/Foreign Assets).
- Limitations and caveats:
  - Accounting nature may not capture market values of assets/liabilities or contingent liabilities.
  - "Other Items Net" inclusion desirable as it captures valuation changes and reserves serving as buffers, though idiosyncratic across countries.
  - Adoption of IFRS by some central banks may reduce opacity in valuations.

#### B. Indicator of Constraints on Monetary Policy (MPC)
- MPC proxy: Deviations of observed policy interest rates from an estimated “optimal” level (forecast errors from fitted interest rate rules); focus on MPC_t < 0 (interest rates set below optimal).
- Estimation procedure: three steps

  - Step 1: Estimation of Interest Rate Rules
    - Baseline specification given as equation (2): policy rate depends on expected 12-months-ahead CPI inflation gap (relative to inflation target), 3-month ahead expected output gap (y* from HP-trend or linear-quadratic trend), and last-quarter observed exchange rate depreciation (vis-à-vis the US dollar).
    - Alternative proxies for activity: industrial production or unemployment rate; different potential output definitions (HP filter or linear-quadratic trend).
    - Models estimated with and without an exchange rate component; eight different specifications per country.
    - Estimation method: instrumental variable-generalized method of moments (IV-GMM) following Clarida, Gali and Gertler (1998, 2000).
    - Instruments: lags of all independent variables and interest rate, plus log of a broad commodity price index. Forward-looking variables (inflation and output gap) are first-stage fitted using instruments; fitted values used in second-stage rule.

  - Step 2: Selection Algorithm
    - Select the best interest rate rule per country based on out-of-sample forecast performance at different horizons.
    - Algorithm in the spirit of Clark and West (2006, 2007); implementation follows Moura and Carvalho (2010) application for interest rate rules.
    - Procedure: estimate rule on subsample D of full sample T (D<T), compute fitted interest rate and mean-correction equation (3) for forecasts at horizon k, derive statistic f for each forecast and rolling window to obtain T-D observations of f for each k, and test whether specification outperforms a random walk.
    - Selection criterion: choose the rule that beats the random walk at the largest number of horizons, operationalized as the rule that rejects with the lowest mean p-value the null of equal predictive ability to a random walk.

  - Step 3: Measure of Monetary Policy Constraint (MPC)
    - MPC constructed as forecast error: MPC_t = observed interest rate − predicted interest rate from selected rule (equation (5)).
    - Emphasis on MPC_t < 0 cases as indicative of potential balance-sheet-driven policy constraint.

#### C. Empirical linkage: CBFS → MPC
- Baseline empirical model: simple bivariate specification where MPC is a function of CBFS (equation (6)); estimated by pooled ordinary least squares.
- Recognizing potential nonlinearity/discontinuities, quantile regressions are also employed to assess heterogeneous effects across the MPC distribution (robust to outliers and useful when dependent variable distribution is highly non-linear).
- Intuition: quantile regressions reveal whether CBFS influences MPC differently for countries with large sub-optimal deviations versus average cases.

#### D. Sample and Data
- Sample: 41 emerging and advanced market economies with some exchange rate flexibility.
- Sample period reported in two places:
  - Main description: 2002:M12011:M3.
  - Methodology implementation note: 2002:m1 to 2011:m4.
- Data sources: IMF International Financial Statistics (IFS) and Haver Analytics.
  - Variables: central bank balance sheet items (for CBFS), policy rates or money market rates, CPI index, industrial production, unemployment rates, exchange rates, broad commodity price index.

---

### Key statistics and methodological particulars (preserve exact figures/terms)
- Sample size: 41 countries.
- Sample periods as reported: "2002:M12011:M3" and "2002:m1 to 2011:m4".
- Number of rule specifications per country: 8.
- Estimation methods: IV-GMM; selection algorithm based on out-of-sample forecast performance (Clark and West framework).
- Main empirical result: CBFS explains deviations of up to 72 basis points in policy interest rates when rates are set below the estimated optimal level.
- Focus on lower deciles of MPC distribution for significant CBFS effects (nonlinear/quantile regression evidence).

*Source: Excerpt from the IMF working paper content unit titled "_wp1260 - References ............................................................................................................."*

### Annex Table A.2).

### _wp1260 - Annex Table A.2)

### Method to derive the MPC measure
- The measure of MPC is derived by estimating the selected interest rate rule over a subsample period and then constructing the dynamic forecast over the remaining sample period.
- Benchmark estimation:
  - Interest rate rule estimated over the pre-crisis period (2002:1-2008:8).
  - MPC variable measured using the dynamic forecast for the out-of-sample period (2008:9-2010:4).
  - Focus: deviations from optimal monetary policy during the post crisis period to study the link to CBFS.
- Robustness of sub-period choices and other dimensions are explored (see Section III.C).
- Additional methodological notes:
  - Panel fixed effects are introduced as one robustness check.
  - Quantile regressions are used as a semi-parametric approach fitting regressions at different points of the dependent variable distribution (Koenker and Hallock (2001) referenced).

### III.A. Assessing whether CBFS interferes with the conduct of monetary policy — main findings
- Bivariate pooled OLS (Equation (6)) indicates CBFS coefficient positive and statistically significant:
  - Coefficient (pooled OLS, benchmark): 0.0414*** (standard error 0.00905)
  - Constant: -1.520*** (standard error 0.166)
  - Observations: 813
  - R-squared: 0.032
- Interpretation:
  - Lower central bank capital is correlated with a wider (negative) gap between observed and optimal interest rate (the MPC measure).
  - A 1 percent increase in the CBFS measure is associated with an average effect of 4 basis points in the MPC measure.
- Non-linearity documented via quantile regressions:
  - Quantile estimates vary across deciles of MPC; the relationship is stronger in tails (larger deviations).
  - Example: For the first decile (indicator negative, interest rates below “optimal”), a 10 percent increase in CBFS decreases the size of the interest rate gap by 72 basis points.
  - Quantile regression benchmark coefficients (selected):
    - q10: 0.0718*** (0.0137)
    - q20: 0.0385*** (0.0136)
    - q30: 0.0300*** (0.0102)
    - q40: 0.0188 (0.0128)
    - q50: 0.0145*** (0.00485)
    - q60: 0.0224*** (0.00612)
    - q70: 0.00949 (0.0104)
    - q80: 0.0265* (0.0139)
    - q90: 0.0378** (0.0164)
  - Constant terms across quantiles reported (selected) range from -6.020*** to 2.285*** with associated standard errors as in the annex table.
- Monetary policy regime and development stage:
  - Inflation Targeting (IT) countries (within EMEs): same qualitative pattern as benchmark — CBFS positively related to MPC and non-linearities matter.
    - Pooled OLS: CBFS coefficient for IT countries: 0.0235 (standard error 0.0184); Constant: 2.861*** (0.249); Observations: 339; R-squared: 0.005.
    - Quantile regression for IT countries: q10: 0.199*** (0.0356) with other quantiles reported in the annex.
  - Developed economies: baseline results do not hold; CBFS appears to have no role even at lower MPC levels.
    - Pooled OLS: CBFS coefficient for Developed Countries: 0.00453 (0.00515); Constant: -2.025*** (0.120); Observations: 200; R-squared: 0.004.
    - Quantile regression for Developed Countries shows small and generally insignificant coefficients across quantiles (selected: q10: -0.000898 (0.00648), q20: 0.00284 (0.00546), q30: 0.00327 (0.00891), q40: 0.00986 (0.0102), q50: 0.00941 (0.0106), q60: 0.00407 (0.0101), q70: 0.00810 (0.0103), q80: 0.00305 (0.00714), q90: 0.000533 (0.00674)).
  - Possible interpretations for developed economies’ null result include stronger institutional frameworks or absence of currency mismatches in central bank balance sheets.

### III.B. Robustness analysis — checks and outcomes
- Sample period for interest rate rule estimation:
  - Re-estimated interest rate rules for 2002-2006:12 and forecast policy rate for 2007:1-2008:12 to assess structural break sensitivity.
  - OLS and quantile regression results qualitatively similar to benchmark but with a level effect and wider 95 percent confidence bands due to smaller sample size.
- In-sample versus out-of-sample forecasts:
  - Constructed MPC using in-sample forecasts and re-estimated Equation (6) over the whole sample and pre-August 2008 sample.
  - Results qualitatively similar and confirm robustness; nonlinearities somewhat more pronounced than benchmark.
- Sensitivity to CBFS definition:
  - Alternative CBFS: capital-to-total-assets ratio (simpler ratio).
  - Non-linearity survives qualitatively, but interpretation less clear because baseline CBFS includes "Other Items Net" to capture valuation changes (reserves, valuation effects) that affect central bank capital.
  - Warning: excluding valuation changes could invalidate the relationship between capital and financial strength.
- Exchange rate misalignments as control variables:
  - Controlled for exchange rate misalignment using:
    - Deviation from a five-year moving average.
    - Deviation from trend using Hodrick-Prescott filter.
  - Results qualitatively robust; CBFS appears more relevant in explaining MPC when accounting for exchange rate misalignment.
- Reverse causation and country fixed-effects:
  - Tested reverse causation by re-estimating Equation (6) using a three-month lagged value for CBFS (motivated by AR-2 term in interest rate rule).
  - Results qualitatively robust; OLS and quantile estimates slightly smaller.
  - Introducing country fixed-effects (dummies) does not affect results.

### Overall interpretation of econometric evidence
- CBFS plays a statistically significant role in explaining “suboptimal” monetary policies in a nonlinear manner.
- Quantile regression evidence: large negative deviations from optimal policy (“undershooting”) can be partially explained by weak CBFS.
- Magnitude example reiterated: 1 percent increase in CBFS → average effect of 4 basis points improvement in MPC; 10 percent increase in CBFS at first decile → 72 basis points reduction in the interest rate gap.

### IV. Conclusions and directions for further research
- Over the past decade, shifts in central bank balance sheet composition and size have made central bank capital sensitive to domestic interest rate movements, raising concerns over implications for monetary policy conduct.
- This paper is the first empirical attempt to analyze the importance of central bank capital for monetary policy conduct.
- Main conclusions:
  - Weak central bank balance sheets can influence monetary policy.
  - The relationship between central bank balance sheets and monetary policy is highly non-linear: large deviations from optimal policy associated with very weak balance sheets.
- Further research recommended:
  - Refine the measure of central bank financial strength to fully capture currency mismatches and interest rate risk while preserving cross-country comparability.
  - Improve understanding of mechanisms through which CBFS influences monetary policy, including macroeconomic and institutional factors (e.g., fiscal arrangements).
  - Apply methodology to examine potential side effects of rapidly expanding central bank balance sheets in advanced economies from non-traditional policies (e.g., quantitative easing).

*Source: _wp1260 - Annex Table A.2).*

### REFERENCES

### REFERENCES

### Reference list
- Adler, Gustavo, and Camilo E. Tovar, 2011, “Foreign Exchange Intervention: A Shield against Appreciation Winds?” IMF Working Paper 11/165 (Washington: International Monetary Fund).
- Benecká, Soňa, Tomáš Holub, Narcisa Liliana Kadlčáková, Ivana Kubicová, 2012, “Central Bank Financial Strength and Inflation: Is There an Empirical Link?” Mimeo.
- Clarida, Richard, Jordi Galí, and Mark Gertler, 2000, “Monetary Policy Rules and Macroeconomic Stability: Evidence and Some Theory,” Quarterly Journal of Economics, February, pp. 147–80.
- __________, 1998, “Monetary Policy Rules in Practice: Some International Evidence,” European Economic Review, Vol. 42, pp 1033–67.
- Clark, Todd, and Kenneth West, 2006, “Using Out-of-Sample Mean Squared Prediction Errors to Test the Martingale Difference Hypothesis,” Journal of Econometrics, Vol. 135, pp. 155–86.
- __________, 2007, “Approximately Normal Tests for Equal Predictive Accuracy in Nested Models,” Journal of Econometrics, Vol. 138, pp. 291–311.
- Ize, Alain, 2005, “Capitalizing Central banks: A Net Worth Approach,” IMF Working Paper 05/15 (Washington: International Monetary Fund).
- __________, and Nada Oulidi, 2009, “Why do Central Banks Go Weak,” IMF Working Paper 09/13 (Washington: International Monetary Fund)
- Jeanne, Olivier, and Lars E. O. Svensson, 2007, “Credible Commitment to Optimal Escape from a Liquidity Trap: The Role of the Balance Sheet of an Independent Central Bank,” The American Economic Review, Vol. 97, No. 1, pp. 474–90.
- Kluh, Ulrich, and Peter Stella, 2008 “Central Bank Financial Strength and Policy Performance: An Econometric Evaluation,” IMF Working Paper 08/17 (Washington: International Monetary Fund).
- Koenker, Roger and Kevin Hallok, 2001, “Quantile Regression,” Journal of Economic Perspectives, Vol. 15, No.4, pp. 143–56.
- Moura, Marcelo, and Alexandre Carvalho, 2010, “What Can Taylor Rules Say about Monetary Policy in Latin America?” Journal of Macroeconomics, Vol. 32. pp. 392–404.
- Nelson, Mark, 1995, “Exchange Rates and Fundamentals: Evidence on Long-Horizon Predictability,” The American Economic Review, Vol. 85, No. 1, pp. 201–18.
- Restrepo, Jorge, Luis Salomó, and Rodrigo Valdés, 2009, “Macroeconomía, Política Monetaria y Patrimonio del Banco Central de Chile,” Economía Chilena, Vol. 12, No. 1.
- Sims, Christopher, 2001, “Fiscal Aspects of Central Bank Independence,” CESIfo Working Paper No. 547. August.
- __________, 2004, “Limits to Inflation Targeting,” in The Inflation-Targeting Debate,” ed. by Ben Bernanke and Michael Woodford. (Chicago: University of Chicago Press).
- Stella, Peter, 1997, “Do Central Banks Need Capital?” IMF Working Paper 97/83 (Washington: International Monetary Fund).
- __________, 2005, “Central Bank Financial Strength, Transparency and Policy Credibility,” IMF Staff Papers. Vol. 52, No. 2. Pages 335–65.
- __________, and Åke Lonnberg, 2008, “Issues in Central Bank Finance Independence,” Federal Reserve Bank of Atlanta Working Paper 2008–13 (Atlanta: Federal Reserve Bank of Atlanta).

*Source: _wp1260 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1260.pdf_