## Section III presents

## Source details

**Canonical URL:** [Section III presents](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019268-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019268-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019268-print-pdf.pdf.json)

---

### Overview of stylized facts
- Expectation: monetary policy rates in GCC countries would tend to move in line with the federal funds rate given the pegged exchange rate regimes.
- For monetary policy to be effective in achieving desired objectives, liability and asset rates would tend to be correlated with the GCC policy rates.
- The section presents stylized facts based on the interest rate series constructed by the authors.

### Empirical observations
- GCC policy rates largely follow U.S. policy rates.
- GCC bank asset rates have broadly followed the policy rates in their home countries.

### Figure summary (Figure 1)
- Title: GCC-Wide Bank and Policy Rates and U.S. Federal Funds Rate (Percent)
- Vertical axis labels shown: 0, 1, 2, 3, 4, 5, 6, 7
- Horizontal axis years shown: 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017
- Series indicated in figure legend: Liability Rate; Asset Rate; Fed Funds Rate; GCC Policy Rate
- Figure source: Fitch Connect; and IMF staff calculations.

*Source: wpiea2019268-print-pdf - Section III presents*

### Appendix II, Figure 1 shows the evolution of

### Overview and context
- GCC banks’ liability rates only partially responded to the increase in the U.S. Federal Funds rate during 2004-06 due to:
  - Relatively higher oil prices during this period with positive implications for banks’ liquidity.
  - Possible greater bank market power when policy rates are high, allowing banks to set deposit rates well below policy rates without depositors switching to cash.
- At the country level:
  - Policy rates are strongly positively correlated with the U.S. Federal Funds Rate.
  - Liability and asset rates also move with U.S. policy rates, with stronger co-movement on the liability side (other than in Saudi Arabia).

### Definitions and data coverage
- Policy rate definitions:
  - Bahrain: one-week central bank deposit rate.
  - Kuwait: central bank discount rate.
  - Oman: average repo rate.
  - Qatar: central bank deposit rate.
  - Saudi Arabia: reverse repo rate.
  - UAE: central bank CD rate.
- Data start dates:
  - Kuwait, Qatar, and Saudi Arabia: starting in 2003.
  - Bahrain and Oman: starting in 2004.
  - UAE: starting in 2007.
- Rate definitions:
  - Liability rate = total interest expense divided by interest-bearing liabilities.
  - Deposit rate = interest expense on customer deposits divided by total customer deposits.
  - Asset rate = gross interest and dividend income divided by interest-earning assets.
  - Loan rate = interest income on loans divided by net loans.
  - Net interest margin (NIM) = net interest income scaled by total interest-earning assets.
  - Islamic compensation included in calculations.
- Liquidity: subset of central bank domestic currency liabilities vis-à-vis commercial banks readily available for payments (commercial bank excess reserves).

### Time-series (bank-level) empirical evidence
- Panel regressions: dependent variables = changes in liability and asset rates; independent variable = change in the U.S. Federal Funds Rate.
- Δ Federal Funds Rate coefficient estimates (from Table 2):
  - Δ Liability Rate: 0.35 (t-statistics: (7.51) and (7.33) across specifications)
  - Δ Asset Rate: 0.19 and 0.18 (t-statistics: (2.41) and (2.41))
- Interpretation taken at face value:
  - A 100 basis point rise in U.S. rates → GCC banks’ liability rates rise by 35 basis points and asset rates by close to 20 basis points.
- Specifications include country or bank fixed effects; standard errors clustered by country and year.
- Limitation: time-series regressions cannot fully isolate policy-rate effects from shocks to oil prices, liquidity conditions, or global financial developments.

### Liability pass-through (LPT) — cross-sectional measure of interest-sensitivity
- LPT definition:
  - From bank-level regressions ΔLiability Rate_i,t = α0 + α1 ΔFederal Funds Rate_t + ε_i,t.
  - Estimated coefficient α̂1 is LPT: α̂1 ≈ 1 → “sensitive”; α̂1 ≈ 0 → “insensitive.”
- Deposit mix and liability composition:
  - Wide variation in fraction of deposits that are interest bearing across banks; some banks have almost no compensated deposits while others rely almost exclusively on compensated deposits.
  - In Saudi Arabia, many of the most profitable banks fund themselves largely with uncompensated deposits.
  - Non-deposit liabilities (e.g., wholesale funding) matter in Kuwait and Bahrain.
- Relationship between LPT and interest-bearing deposit fraction:
  - In Saudi Arabia, LPT is closely related to the interest-bearing fraction of deposits.
  - Across GCC, LPT broadly aligns with interest-bearing deposit fraction for many banks, but notable divergences exist due to differing deposit shares of liabilities, varying deposit franchise strength/market power, and reliance on wholesale funding.

### Cross-sectional panel analysis (interaction of ΔFFR and LPT)
- Estimation specification:
  - Δ Interest Rate_i,t = α_t + α_i + α1 (Δ Federal Funds Rate_t × Liability Passthrough_i) + ε_i,t with year fixed effects and bank fixed effects; standard errors double clustered by country and year.
- Key estimates (Table 3):
  - Δ Federal Funds Rate × Liability Passthrough coefficients:
    - Δ Liability Rate: 0.74 (t-statistic (3.34))
    - Δ Asset Rate: 0.47 (t-statistic (1.3))
    - Δ Net Interest Margin: -0.11 (t-statistic (-1.02))
- Interpretation:
  - Relative to a bank with no LPT, a bank with one-to-one LPT experiences an additional 74 basis points increase in liability rates for a 100 basis point rise in U.S. rates.
  - Banks with higher LPT also tend to raise asset rates more when U.S. rates rise, but by less than liability-rate increases.
  - The differential impact on profitability (NIM) is negative in sign but not statistically distinguishable from zero at the GCC-wide level.

### Quantity responses: liabilities, asset growth, and deposit composition
- Response of quantities to Δ U.S. rates (Table 4). Dependent variables are log-changes × 100 (interpretable as percentage-point changes).
- Δ Federal Funds Rate × Liability Passthrough coefficients:
  - Δ ln (Liabilities) × 100: 7.52 (t-statistic (1.91))
  - Δ ln (Assets) × 100: 6.26 (t-statistic (1.64)) — coefficient not statistically significant from zero
- Interpretation:
  - Compared with a bank with no LPT, a bank with one-to-one LPT sees additional overall liability growth of 7.5 percentage points for a 100 basis point increase in U.S. rates; assets grow by close to an additional 7 percentage points (not statistically significant).
- Composition of liability growth (Table 5):
  - Δ ln (Interest Bearing Deposits): 6.78 (t-statistic (3.55))
  - Δ ln (Non Interest Bearing Deposits): -8.47 (t-statistic (-2.31))
  - Δ ln (Other Liabilities): 2.1 (t-statistic (0.19))
- Interpretation:
  - Sensitive banks’ new liabilities are largely in the form of compensated (interest-bearing) deposits; uncompensated deposits shrink.
  - Little effect on other liabilities.
- Implication:
  - Additional inflows of compensated deposits when U.S. rates rise may force banks to expand assets, constraining their ability to raise asset rates and thereby limiting the pass-through to loan pricing.

### Bank competition, concentration, and heterogeneous profitability effects
- Market concentration in GCC banking (Figure 6) — top four banks account for:
  - more than 40 percent of assets in the UAE and Bahrain,
  - about 60 percent in Saudi Arabia,
  - about 70 percent in Kuwait, Qatar, and Oman.
- Relationship between concentration and impact on profitability (Figure 7):
  - The impact of Δ FFR × LPT on NIM plotted against Herfindahl-Hirschman Index (HHI, range 0-10000).
  - Rising U.S. rates matter more for sensitive banks’ profitability in more concentrated banking systems.
  - Statistically significant negative impacts of rising U.S. rates on sensitive banks’ profitability found for Bahrain, Oman, and Qatar.
- Regression evidence with competition interactions (Table 6 highlights):
  - Examples of coefficients:
    - Liability Passthrough: -0.54 (t-statistic (-1.65))
    - Δ FFR × Liability Passthrough: 0.29 and 0.34 (t-statistics (0.85), (0.95))
    - Δ FFR × HHI: -0.75 and -0.55 (t-statistics (-1.14), (-0.81))
    - Liability Passthrough × HHI: 3.13 (t-statistic (1.70))
    - Δ FFR × Liability Passthrough × HHI: 2.84 and 2.39 (t-statistics (1.61), (1.3))
  - Results are suggestive but not statistically robust given limited sample (six countries).

### Conclusions and policy implications
- Main findings:
  - Stronger pass-through from U.S. to GCC policy rates and bank liability rates than to asset rates and bank profitability.
  - Joint evolution of prices and quantities helps explain limited sensitivity of asset rates: banks that raise liability rates in response to U.S. tightening tend to attract compensated deposits and expand assets, leaving little scope to raise asset rates.
  - Presence of large banks with relatively insensitive liabilities constrains more sensitive banks’ ability to raise asset rates.
- Policy perspective:
  - There may be room to improve monetary policy transmission to loan rates.
  - Measures to enhance competition with emphasis on the loan market could help enhance the transmission of monetary policy.

*Source: IMF staff calculations and analysis from the referenced PDF content unit.*

### APPENDIX I. DATA

### Sample and data sources
- Annual bank-level data from Fitch Connect.
- Sample period: 2004-2017.
- For consistency, banks for which coverage begins after 2004 are excluded.
- Policy rates obtained from Haver:
  - Bahrain: one-week deposit rate
  - Kuwait: central bank discount rate
  - Oman: repo rate
  - Qatar: deposit rate
  - Saudi Arabia: reverse repo rate
  - United Arab Emirates: central bank CD rate
  - United States: Federal Funds rate

### Bank-level rate and margin calculations
- Liability side:
  - Liability rate = interest expense scaled by interest-bearing liabilities.
  - Deposit rate = interest expense on customer deposits scaled by customer deposits.
- Asset side:
  - Asset rate = gross interest and dividend income scaled by interest-earning assets.
  - Loan rate = interest income on loans divided by net loans.
- Profitability:
  - Net interest margin (NIM) = net interest income scaled by total interest-earning assets.
- Aggregation:
  - Country and GCC-level rates are calculated as bank size-weighted averages by treating the relevant group of banks as a single bank.

### Coverage of Islamic banking
- Islamic banking included in the data.
- Example (Saudi Arabia): financial statements report ‘special commission’ compensation on assets and liabilities associated with Islamic banking operations; Fitch classifies these receipts and payments as interest income and interest expense.

### Sample size by country (Table 1)
- Bahrain: 20
- Kuwait: 16
- Oman: 8
- Qatar: 8
- Saudi Arabia: 11
- United Arab Emirates (UAE): 20

*Source: APPENDIX I. DATA, wpiea2019268-print-pdf*

### Section III presents

### Section III presents

### Overview of stylized facts
- The section presents stylized facts based on the interest rate series constructed by the authors.
- Expectation: monetary policy rates in GCC countries would tend to move in line with the federal funds rate given the pegged exchange rate regimes.
- For monetary policy to be effective in achieving desired objectives, liability and asset rates would tend to be correlated with the GCC policy rates.

### Empirical observations
- GCC policy rates largely follow U.S. policy rates.
- GCC bank asset rates have broadly followed the policy rates in their home countries.

### Figure summary (Figure 1)
- Title: GCC-Wide Bank and Policy Rates and U.S. Federal Funds Rate (Percent)
- Vertical axis labels shown: 0, 1, 2, 3, 4, 5, 6, 7
- Horizontal axis years shown: 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017
- Series indicated in figure legend: Liability Rate; Asset Rate; Fed Funds Rate; GCC Policy Rate
- Figure source: Fitch Connect; and IMF staff calculations.
- Footnotes referenced in text: 3; 4

*Source: wpiea2019268-print-pdf - Section III presents*

### Appendix II, Figure 1 shows the evolution of

### wpiea2019268-print-pdf - Appendix II, Figure 1 shows the evolution of

### Overview and context
- GCC banks’ liability rates only partially responded to the increase in the U.S. Federal Funds rate during 2004-06 due to:
  - Relatively higher oil prices during this period with positive implications for banks’ liquidity (see IMF (2017) and Adedeji and others (2019)).5
  - Possible greater bank market power when policy rates are high (Drechsler and others 2017), allowing banks to set deposit rates well below policy rates without depositors switching to cash.
- At the country level:
  - Policy rates are strongly positively correlated with the U.S. Federal Funds Rate (Table 1).
  - Liability and asset rates also move with U.S. policy rates, with stronger co-movement on the liability side (other than in Saudi Arabia).
- Definitions and data coverage (as stated in the source):
  - Policy rate definitions vary by country (one-week central bank deposit rate in Bahrain; central bank discount rate in Kuwait; average repo rate in Oman; central bank deposit rate in Qatar; reverse repo rate in Saudi Arabia; central bank CD rate in the UAE).3
  - Data obtained starting in 2003 for Kuwait, Qatar, and Saudi Arabia; 2004 for Bahrain and Oman; 2007 for the UAE.3
  - Liability rate = total interest expense divided by interest-bearing liabilities.4
  - Deposit rate = interest expense on customer deposits divided by total customer deposits.4
  - Asset rate = gross interest and dividend income divided by interest-earning assets.4
  - Loan rate = interest income on loans divided by net loans.4
  - Islamic compensation included in calculations.4
  - “Liquidity” defined as subset of central bank domestic currency liabilities vis-à-vis commercial banks readily available for payments (commercial bank excess reserves).5

### Time-series (bank-level) empirical evidence
- Panel regressions where dependent variables are changes in liability and asset rates and independent variable is change in the U.S. Federal Funds Rate (Table 2):
  - Δ Federal Funds Rate coefficient estimates:
    - Δ Liability Rate: 0.35 (t-statistics: (7.51) and (7.33) across specifications)
    - Δ Asset Rate: 0.19 and 0.18 (t-statistics: (2.41) and (2.41))
  - Interpretation taken at face value:
    - A 100 basis point rise in U.S. rates → GCC banks’ liability rates rise by 35 basis points and asset rates by close to 20 basis points.
  - Specifications include country or bank fixed effects; standard errors clustered by country and year.6
- Limitations noted:
  - Time-series regressions cannot fully isolate policy-rate effects from shocks to oil prices, liquidity conditions, or global financial developments.

### Liability pass-through (LPT) — cross-sectional measure of interest-sensitivity
- LPT constructed from bank-level regressions: ΔLiability Rate_i,t = α0 + α1 ΔFederal Funds Rate_t + ε_i,t
  - Estimated coefficient α̂1 is LPT: banks with α̂1 ≈ 1 are “sensitive” (liability rates respond one-for-one); α̂1 ≈ 0 are “insensitive.”
- Deposit mix and liability composition:
  - Wide variation in fraction of deposits that are interest bearing across banks; some banks have almost no compensated deposits while others rely almost exclusively on compensated deposits (Figure 2).
  - In Saudi Arabia, many of the most profitable banks fund themselves largely with uncompensated deposits.
  - Non-deposit liabilities (e.g., wholesale funding) matter in Kuwait and Bahrain (Figure 3).
- Relationship between LPT and interest-bearing deposit fraction:
  - In Saudi Arabia, LPT is closely related to the interest-bearing fraction of deposits (Figure 4).
  - Across GCC, LPT broadly aligns with interest-bearing deposit fraction for many banks, but notable divergences exist (Figure 5). Reasons include differing deposit shares of liabilities, varying deposit franchise strength/market power, and reliance on wholesale funding.

### Cross-sectional panel analysis (interaction of ΔFFR and LPT)
- Specification used (Table 4 / Table 3 description):
  - Δ Interest Rate_i,t = α_t + α_i + α1 (Δ Federal Funds Rate_t × Liability Passthrough_i) + ε_i,t
  - Includes year fixed effects (α_t) and bank fixed effects (α_i); standard errors double clustered by country and year.
- Key estimates (Table 3):
  - Δ Federal Funds Rate × Liability Passthrough coefficients:
    - Δ Liability Rate: 0.74 (t-statistic (3.34))
    - Δ Asset Rate: 0.47 (t-statistic (1.3))
    - Δ Net Interest Margin: -0.11 (t-statistic (-1.02))
  - Interpretation:
    - Relative to a bank with no LPT, a bank with one-to-one LPT experiences an additional 74 basis points increase in liability rates for a 100 basis point rise in U.S. rates.
    - Banks with higher LPT also tend to raise asset rates more when U.S. rates rise, but by less than liability-rate increases.
    - The differential impact on profitability (NIM) is negative in sign but not statistically distinguishable from zero at the GCC-wide level.

### Quantity responses: liabilities, asset growth, and deposit composition
- Response of quantities to Δ U.S. rates (Table 4):
  - Dependent variables are log-changes × 100 (interpretable as percentage-point changes).
  - Δ Federal Funds Rate × Liability Passthrough coefficients:
    - Δ ln (Liabilities) × 100: 7.52 (t-statistic (1.91))
    - Δ ln (Assets) × 100: 6.26 (t-statistic (1.64)) — coefficient not statistically significant from zero
  - Interpretation:
    - Compared with a bank with no LPT, a bank with one-to-one LPT sees additional overall liability growth of 7.5 percentage points for a 100 basis point increase in U.S. rates; assets grow by close to an additional 7 percentage points (not statistically significant).
- Composition of liability growth (Table 5):
  - Δ ln (Interest Bearing Deposits): 6.78 (t-statistic (3.55))
  - Δ ln (Non Interest Bearing Deposits): -8.47 (t-statistic (-2.31))
  - Δ ln (Other Liabilities): 2.1 (t-statistic (0.19))
  - Interpretation:
    - Sensitive banks’ new liabilities are largely in the form of compensated (interest-bearing) deposits; uncompensated deposits shrink.
    - Little effect on other liabilities.
- Implication:
  - Additional inflows of compensated deposits when U.S. rates rise may force banks to expand assets, constraining their ability to raise asset rates and thereby limiting the pass-through to loan pricing.

### Bank competition, concentration, and heterogeneous profitability effects
- Market concentration in GCC banking (Figure 6):
  - Top four banks account for:
    - more than 40 percent of assets in the UAE and Bahrain,
    - about 60 percent in Saudi Arabia,
    - about 70 percent in Kuwait, Qatar, and Oman.
- Relationship between concentration and impact on profitability (Figure 7):
  - The impact of Δ FFR × LPT on NIM plotted against Herfindahl-Hirschman Index (HHI, range 0-10000).
  - Rising U.S. rates matter more for sensitive banks’ profitability in more concentrated banking systems.
  - Statistically significant negative impacts of rising U.S. rates on sensitive banks’ profitability found for Bahrain, Oman, and Qatar (those countries labeled in bold).
- Regression evidence with competition interactions (Table 6 highlights):
  - Coefficients (examples):
    - Liability Passthrough: -0.54 (t-statistic (-1.65))
    - Δ FFR × Liability Passthrough: 0.29 and 0.34 (t-statistics (0.85), (0.95))
    - Δ FFR × HHI: -0.75 and -0.55 (t-statistics (-1.14), (-0.81))
    - Liability Passthrough × HHI: 3.13 (t-statistic (1.70))
    - Δ FFR × Liability Passthrough × HHI: 2.84 and 2.39 (t-statistics (1.61), (1.3))
  - Results are suggestive but not statistically robust given limited sample (six countries).

### Conclusions and policy implications
- Main findings:
  - Stronger pass-through from U.S. to GCC policy rates and bank liability rates than to asset rates and bank profitability.
  - Joint evolution of prices and quantities helps explain limited sensitivity of asset rates: banks that raise liability rates in response to U.S. tightening tend to attract compensated deposits and expand assets, leaving little scope to raise asset rates.
  - Presence of large banks with relatively insensitive liabilities constrains more sensitive banks’ ability to raise asset rates.
- Policy perspective:
  - There may be room to improve monetary policy transmission to loan rates.
  - Measures to enhance competition with emphasis on the loan market could help enhance the transmission of monetary policy.

*Source: IMF staff calculations and analysis from the referenced PDF content unit.*

### APPENDIX I. DATA

### APPENDIX I. DATA

### Sample and data sources
- Annual bank-level data from Fitch Connect.
- Sample period: 2004-2017.
- For consistency, banks for which coverage begins after 2004 are excluded.
- Policy rates obtained from Haver:
  - Bahrain: one-week deposit rate
  - Kuwait: central bank discount rate
  - Oman: repo rate
  - Qatar: deposit rate
  - Saudi Arabia: reverse repo rate
  - United Arab Emirates: central bank CD rate
  - United States: Federal Funds rate

### Bank-level rate and margin calculations
- Liability side:
  - Liability rate = interest expense scaled by interest-bearing liabilities.
  - Deposit rate = interest expense on customer deposits scaled by customer deposits.
- Asset side:
  - Asset rate = gross interest and dividend income scaled by interest-earning assets.
  - Loan rate = interest income on loans divided by net loans.
- Profitability:
  - Net interest margin (NIM) = net interest income scaled by total interest-earning assets.
- Aggregation:
  - Country and GCC-level rates are calculated as bank size-weighted averages by treating the relevant group of banks as a single bank.

### Coverage of Islamic banking
- Islamic banking included in the data.
- Example (Saudi Arabia): financial statements report ‘special commission’ compensation on assets and liabilities associated with Islamic banking operations; Fitch classifies these receipts and payments as interest income and interest expense.

### Sample size by country (Table 1)
- Bahrain: 20
- Kuwait: 16
- Oman: 8
- Qatar: 8
- Saudi Arabia: 11
- United Arab Emirates (UAE): 20

*Source: APPENDIX I. DATA, wpiea2019268-print-pdf*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019268-print-pdf.pdf_