## Annex A. Monetary Policy Shocks

## Source details

**Canonical URL:** [Annex A. Monetary Policy Shocks](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026102-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026102-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026102-source-pdf.pdf.json)

---

### Purpose and identification approach
- Applies a Romer and Romer (2004)-style single-equation approach adapted to SSA economies to construct monetary policy shocks by purging the policy rate of systematic responses to observed and expected macroeconomic conditions, isolating the unanticipated component of policy actions.
- Uses identified shocks to trace transmission:
  - (i) through financial markets (speed and strength of pass-through across channels);
  - (ii) to economic activity and prices; and
  - (iii) to investigate how structural factors shape effectiveness across SSA EFMs.
- Identification: Romer and Romer (2004) Taylor-rule-type regression where residuals (ε_t) are interpreted as exogenous monetary policy shocks; regression includes ∆ii_t, lagged ii, one-year-ahead WEO projections E_t Δy_{t+4} and E_t π_{t+4}, two lags of GDP growth and inflation, and lagged changes in nominal effective exchange rate.

### Key empirical findings
- Interest rate channel
  - Strong evidence of an active interest rate channel across the region.
  - Short-term market interest rates, as well as bank deposit and lending rates, respond quickly and persistently to monetary tightening and are statistically significant in most economies.
  - Deposit and lending rate magnitudes reported:
    - In South Africa: a 100-basis-point contractionary monetary policy shock raises deposit rates by about 1.5 percentage points after three quarters.
    - In Ghana and Uganda: a 100-basis-point contractionary monetary policy shock raises deposit rates by about 0.5 percentage point.
    - In South Africa, Ghana, and Mozambique: a 100-basis-point monetary policy shock increases lending rates by about 1.5-2 percentage points after four quarters.
- Exchange rate channel
  - Weak exchange rate channel: exchange rate responses to contractionary monetary policy are generally muted and statistically insignificant and, in many cases, exhibit depreciation (the "exchange rate puzzle").
  - A 100-basis-point monetary policy shock leads to an appreciation of the NEER of less than 0.5 percent on impact on average.
  - Country examples: expected appreciation only in Kenya (positive and statistically significant); sharp depreciation in Nigeria; depreciation to a lesser extent in Zambia, South Africa, and Ghana.
- Real GDP and CPI responses
  - Contractionary monetary policy reduces both real GDP and consumer prices with a lag; output responds before inflation.
  - Magnitude is modest and notably weaker than in AEs and EMEs.
  - Pooled SSA EFM averages:
    - Real GDP: a 100-basis-point monetary policy shock reduces real GDP by about 0.2 percent at its peak, occurring between six and seven quarters after the shock (also reported as peak after roughly eight quarters in pooled sample).
    - Headline CPI: declines by roughly 0.4 percent about two years after the shock (effects remain statistically insignificant throughout in pooled sample).
  - Heterogeneity by framework:
    - SSA EFM IT: a 100-basis-point shock reduces real GDP by around 0.3 percent at its peak (after four to six quarters); CPI falls by approximately 1 percent two years after the shock (reported as stronger and statistically significant).
    - SSA EFM MT: smaller and statistically insignificant effects — real GDP decline of about 0.2 percent at the trough (around eight quarters); CPI decline of about 0.2 percent (insignificant).
  - Cross-region comparison:
    - In AEs: real GDP declines by approximately 0.4 percent after one year, reaching a trough of around 1 percent after eight quarters.
    - In EMEs: real GDP declines by about 0.3 percent after one year and remains at roughly that level after two years.
- Financial conditions
  - Tighter financial conditions, measured using a financial conditions index (FCI), exert a sizable contractionary effect on both output and inflation.
  - A one percent increase in the FCI (tighter financial conditions) exerts economically meaningful and statistically significant contractionary effects on output; peak output effects typically occur after about four to six quarters and remain persistent.
  - Disinflationary effects from FCI tightening are strong and statistically significant in most SSA EFMs, with prices continuing to decline for up to two years following the shock.

### Transmission across financial markets (selected findings)
- Money market (interbank) rates
  - Immediate and persistent rises occur in a limited number of economies; strongest, most persistent, and statistically significant responses in South Africa, Ghana, Mozambique, and to some extent, Uganda.
  - Positive but short-lived and statistically significant responses in Kenya, Tanzania, Zambia.
- Government treasury bill and bond yields
  - T-bill yields increase following contractionary shocks in most economies; T-bond responses generally positive but less pronounced than T-bills.
  - Positive, statistically significant, and persistent T-bond responses observed in Ghana, Tanzania, and Uganda.
- Stock prices
  - Contractionary shocks lower stock prices; statistically significant declines only in a small number of economies.
  - Most pronounced declines in Nigeria, South Africa, and Zambia.
- Deposit and lending rates
  - Broadly persistent increases across most economies, pointing to a dominant bank-based interest rate channel given bank-based financial systems in SSA EFMs.
  - Exceptions with muted lending-rate responses: Mauritius and Tanzania.
  - Cases where lending rates respond more strongly than deposit rates: Ghana, Mozambique, Zambia.

### Cross-country heterogeneity and structural determinants
- Stronger and faster transmission associated with:
  - Adoption of formal inflation-targeting frameworks (IT).
  - Greater monetary policy transparency.
  - Higher levels of financial development.
  - To a lesser extent, lower fiscal dominance.
- Non-significant or limited explanatory factors:
  - Sovereign-bank nexus does not appear to significantly explain cross-country differences in transmission.
  - Exchange rate flexibility does not appear to significantly explain cross-country differences in transmission for output; it is more relevant for inflation differences.
- Structural constraints linked to weaker transmission:
  - Underdeveloped financial systems, fiscal dominance, weak institutions, limited banking competition, persistently high excess reserves, shallow financial markets, restricted capital mobility, low global financial integration, and prevalence of de facto exchange rate management or fixed arrangements in several economies.
- Statistical tests (p-values) for group differences (selected results)
  - GDP response differences: Financial Development significant at multiple horizons (p-values include 0.023**, 0.034**, 0.028**, 0.048**, 0.052*, 0.059* across horizons 3–8); Monetary Policy Transparency significant at some horizons (e.g., 0.076* at horizon 3; 0.053* at horizon 5).
  - CPI response differences: Monetary Policy Transparency significant at horizon 1 (0.015**) and Exchange Rate Flexibility significant at horizons 3–8 (e.g., 0.046**, 0.008***, 0.008***, 0.006***, 0.000***, 0.000***).
  - Notes: ***, **, and * denote statistical significance at the 1, 5, and 10 percent levels, respectively.

### Data, sample, and empirical framework
- Sample and period
  - Quarterly data for 11 SSA EFMs: CEMAC, Ghana, Kenya, Mauritius, Mozambique, Nigeria, Tanzania, Uganda, WAEMU, South Africa, and Zambia.
  - Sample period: first quarter of 2002 to the fourth quarter of 2023.
  - Classification as of end-2023: Ghana, Kenya, Mauritius, South Africa and Uganda classified as inflation targeters; Nigeria and Tanzania target monetary aggregates; Mozambique and Zambia follow eclectic frameworks; the two monetary unions peg to the euro but retain some independent policy scope.
- Variables and sources
  - First-stage variables: policy rate, interbank rates, government treasury bill and bond yields, stock prices, nominal exchange rates, bank deposit and lending rates.
  - Second-stage variables: real GDP and headline consumer price indices; supplemented with international commodity price data.
  - Forecast proxies: central bank forecasts proxied by IMF WEO projections for real GDP growth and inflation; professional forecasters’ expectations (Consensus Forecasts) used for AEs and major EMEs where available.
  - Conditioning indices: Monetary policy frameworks from AREAER (2023); IMF Financial Development Index; monetary policy transparency index of Dincer et al. (2022); External Wealth of Nations database for international financial integration.
- Empirical methods
  - First-stage local projections (Jordà, 2005) for financial variables: y_{t+h} − y_{t−1} = α_h + β_h MPS_t + Σ_{k=1}^{K} θ_{k,h} X_{t−k} + ε_{t+h,h}, h = 0,...,8; Newey–West SEs.
  - Second-stage panel local projections for macro outcomes with country and time fixed effects and clustered SEs at country level.
  - Country-level local projections and panel interaction/specification splitting countries into High vs. Low groups based on sample means of conditioning variables.

### Robustness checks and validation
- Alternative exchange rate measures (USD bilateral) yield broadly consistent country-level results showing muted exchange rate responses or exchange rate puzzle in most economies.
- Using changes in the actual policy rate yields broadly similar output responses but produces a pronounced price puzzle across country groups; identified shocks reduce the price puzzle relative to actual policy rate.
- VAR-based robustness: parsimonious VARs produce qualitatively similar dynamics with generally smaller magnitudes and comparable cross-country heterogeneity.
- Interaction and statistical significance: differences in responses statistically significant for financial development (output) and monetary policy transparency (output and prices); exchange rate flexibility significant mainly for inflation; fiscal dominance and sovereign-bank nexus differences generally not statistically significant.

### Policy implications and recommended reform priorities
- Strengthen monetary policy frameworks:
  - Improvements in operational and institutional features of monetary policy, including greater transparency, can enhance transmission effectiveness.
- Develop domestic financial systems:
  - Higher levels of financial development bolster the effectiveness and persistence of monetary transmission—particularly the bank lending channel.
- Address fiscal dominance:
  - Lower fiscal dominance is associated with stronger transmission; fiscal discipline and limiting central bank credit to government may improve monetary policy effectiveness.
- Recognize limits of interest-rate interventions alone:
  - Interest rate hikes alone may be insufficient to systematically influence exchange rates due to limited capital mobility and low global financial integration.
- Consider broader financial conditions in policy design:
  - Financial conditions (as captured by FCIs) have sizable effects on output and inflation; policymakers should monitor and, where possible, address broader financial vulnerabilities and global financial spillovers.

*IMF Working Paper — Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets — Annex A. Monetary Policy Shocks*

### Annex A. Monetary Policy Shocks ........................................................................................

### Annex A. Monetary Policy Shocks

### Purpose and identification approach
- Applies a Romer and Romer (2004)-style single-equation approach adapted to SSA economies to construct monetary policy shocks by purging the policy rate of systematic responses to observed and expected macroeconomic conditions, isolating the unanticipated component of policy actions.
- Uses identified shocks to trace transmission: (i) through financial markets (speed and strength of pass-through across channels); (ii) to economic activity and prices; and (iii) to investigate how structural factors shape effectiveness across SSA EFMs.

### Key empirical findings
- Interest rate channel:
  - Strong evidence of an active interest rate channel across the region.
  - Short-term market interest rates, as well as bank deposit and lending rates, respond quickly and persistently to monetary tightening and are statistically significant in most economies.
- Exchange rate channel:
  - Weak exchange rate channel: exchange rate responses to contractionary monetary policy are generally muted and statistically insignificant and, in many cases, exhibit depreciation (the "exchange rate puzzle").
- Real GDP and CPI responses:
  - Contractionary monetary policy reduces both real GDP and consumer prices with a lag, with output responding before inflation.
  - The overall magnitude is modest and notably weaker than in AEs and EMEs.
  - A 100-basis-point monetary policy shock reduces real GDP by about 0.2 percent at its peak in SSA EFMs, compared with around 1 percent in AEs.
  - In inflation-targeting (IT) economies, a 100-basis-point shock reduces real GDP by around 0.3 percent at its peak (after four to six quarters); effects in monetary-aggregate-targeting (MT) regimes are smaller and statistically insignificant.
  - Disinflationary effects are more pronounced and statistically significant in IT economies.
  - Country-level results reveal weak and uncertain transmission in some economies and some evidence of a price puzzle in a few economies.
- Financial conditions:
  - Tighter financial conditions, measured using a financial conditions index (FCI), exert a sizable contractionary effect on both output and inflation, indicating broader financial channels matter beyond the direct policy rate effect.

### Cross-country heterogeneity and structural determinants
- Stronger and faster transmission is associated with:
  - Adoption of formal inflation-targeting frameworks (IT).
  - Greater monetary policy transparency.
  - Higher levels of financial development.
  - To a lesser extent, lower fiscal dominance.
- Non-significant or limited explanatory factors:
  - The sovereign-bank nexus does not appear to significantly explain cross-country differences in transmission.
  - Exchange rate flexibility does not appear to significantly explain cross-country differences in transmission.
- Structural constraints implicated in weaker transmission:
  - Underdeveloped financial systems.
  - Fiscal dominance.
  - Weak institutions.
  - Limited banking competition.
  - Persistently high excess reserves.
  - Shallow financial markets, restricted capital mobility, and low global financial integration; prevalence of de facto exchange rate management or fixed arrangements in several economies.

### Methodological and contribution notes
- Rationale for method:
  - Single-equation Romer and Romer (2004)-style approach is well suited to data constraints common in SSA and offers an alternative to VAR-based methods.
- Comparative contribution:
  - Provides a comprehensive assessment across multiple financial channels and macroeconomic outcomes for SSA EFMs using a consistent framework.
  - Systematically investigates how monetary policy frameworks and macro-financial conditions shape transmission strength and timing across countries.
- Relation to existing literature:
  - Builds on findings that monetary transmission is generally weaker in LICs and SSA, and on single-country loan-level studies documenting an active bank lending channel.
  - Cites micro-level evidence and prior cross-country and single-country studies indicating a dominant bank lending channel in LICs and SSA.

*IMF Working Paper — Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets — Annex A. Monetary Policy Shocks*

### Section V reports robustness checks and additional results. Section VI concludes.

### wpiea2026102-source-pdf - Section V reports robustness checks and additional results. Section VI concludes.

### II. Impediments to Effective Monetary Policy Transmission in SSA
- Monetary policy transmission is country-specific and depends on: financial market depth, international financial integration, exchange rate regime, banking system structure, and credibility of the monetary policy framework.
- Structural conditions in SSA—and in LICs more broadly—are often underdeveloped or unevenly distributed, implying weaker and more uncertain transmission than in AEs and major EMEs.
- Key channel-specific impediments:
  - Interest rate channel: impaired by shallow interbank markets, underdeveloped domestic securities markets, limited financial inclusion, and high excess reserves, reducing pass-through from the policy rate to market rates and borrowing costs.
  - Bank lending channel: constrained by concentrated banking systems, weak competition, high net interest margins, and limited non-bank financial institutions.
  - Asset price channel: weak because equity and real estate markets are generally small, illiquid, and narrow, with limited market capitalization outside a few countries (notably South Africa).
  - Exchange rate channel: limited by low integration with international capital markets, shallow and illiquid foreign exchange markets, frequent foreign exchange intervention, and, in some countries, fixed or heavily managed exchange rate regimes.
  - Expectations channel: weakened where monetary policy frameworks lack coherence and transparency, compounded by high exposure to supply shocks (food and fuel price fluctuations, weather shocks, terms-of-trade volatility) that raise inflation variability and complicate policy trade-offs.
- Fiscal dominance amplifies channel weaknesses:
  - When fiscal constraints limit the central bank’s ability to tighten policy, inflation expectations anchor to fiscal sustainability rather than to monetary policy.
  - Political pressure for low interest rates, high public debt, elevated default risk, and a strong sovereign-bank nexus (large bank holdings of government debt) make central banks more reluctant to raise rates.
- Substantial intra-regional heterogeneity: countries such as South Africa, Ghana, Kenya, and Mauritius have stronger institutional frameworks and more developed financial systems, supporting more effective transmission.

### III. Data and Methodology
- Sample and coverage:
  - Quarterly data for 11 SSA EFMs: CEMAC, Ghana, Kenya, Mauritius, Mozambique, Nigeria, Tanzania, Uganda, WAEMU, South Africa, and Zambia.
  - Sample period: first quarter of 2002 to the fourth quarter of 2023.
  - As of end-2023: five countries—Ghana, Kenya, Mauritius, South Africa and Uganda—classified as inflation targeters; Nigeria and Tanzania target monetary aggregates; Mozambique and Zambia follow eclectic frameworks; the two monetary unions peg to the euro but retain some independent policy scope.
- Financial and macro variables:
  - First-stage variables: policy rate, interbank rates, government treasury bill and bond yields, stock prices, nominal exchange rates, bank deposit and lending rates.
  - Second-stage variables: real GDP and headline consumer price indices; supplemented with international commodity price data.
  - Benchmarks: extended analysis to 10 AEs and 13 major EMEs.
- Forecast proxies:
  - Central bank forecasts proxied by IMF WEO projections for real GDP growth and inflation.
  - Professional forecasters’ expectations (Consensus Forecasts) used for AEs and major EMEs where available.
- Conditioning variables and indices:
  - Monetary policy frameworks from AREAER (2023).
  - Financial development: IMF Financial Development Index.
  - Monetary policy transparency: index of Dincer et al. (2022).
  - International financial integration: External Wealth of Nations database (external assets and liabilities to GDP).
- Identification of monetary policy shocks:
  - Adopt Romer and Romer (2004) Taylor-rule-type regression to purge systematic responses and interpret residuals as exogenous shocks.
  - Estimated regression (equation (1)):
    - ∆ii_t = α + γ1 ii_{t−1} + γ2 E_t Δy_{t+4} + γ3 E_t π_{t+4} + Σ_{k=1}^{2} γ4,k Δy_{t−k} + Σ_{k=1}^{2} γ5,k π_{t−k} + Σ_{k=1}^{2} γ6,k ΔNEER_{t−k} + ε_t
    - Δii_t denotes change in the policy rate; E_t Δy_{t+4} and E_t π_{t+4} denote one-year-ahead WEO projections of real GDP growth and inflation; two lags of GDP growth and inflation and lagged changes in nominal effective exchange rate included.
  - Residuals (ε_t) interpreted as exogenous monetary policy shocks.
- Empirical framework:
  - First-stage local projections (Jordà, 2005) for financial variables:
    - Equation (2): y_{t+h} − y_{t−1} = α_h + β_h MPS_t + Σ_{k=1}^{K} θ_{k,h} X_{t−k} + ε_{t+h,h}, estimated for h = 0,1,...,8.
    - MPS_t is the identified monetary policy shock; X includes four lags of dependent variable and shock and indicator-specific controls (e.g., VIX for stocks, U.S. federal funds rate).
    - Confidence intervals computed using Newey–West standard errors.
  - Second-stage panel local projections for macro outcomes:
    - Equation (3): y_{j,t+h} − y_{j,t−1} = α_{j,h} + α_{t,h} + β_h MPS_{j,t} + Σ_{k=1}^{K} θ_{k,h} X_{j,t−k} + ε_{j,t+h,h}, estimated for SSA EFMs and disaggregated by policy framework (SSA EFM IT vs SSA EFM MT).
    - Country fixed effects and time fixed effects included; standard errors clustered at country level.
  - Country-level local projections:
    - Equation (4): y_{t+h} − y_{t−1} = α_h + β_h MPS_t + Σ_{k=1}^{K} θ_{k,h} X_{t−k} + ε_{t+h,h}, with controls for global conditions (global commodity price growth in local currency, U.S. policy rate).
- Heterogeneity analysis:
  - Split sample into “High” or “Low” groups based on whether the historical mean of a conditioning variable lies above or below the SSA EFM sample mean.
  - Panel local projection for groups (equation (5)):
    - y_{j,t+h} − y_{j,t−1} = α_{j,h} + α_{t,h} + β_{g,h} MPS_{j,t} + Σ_{k=1}^{K} θ_{g,k,h} X_{j,t−k} + ε_{j,t+h,h}, with g ∈ {High, Low}.
  - Conditioning characteristics considered:
    - Monetary policy transparency index (high vs. low).
    - Financial development (IMF Financial Development Index; high vs. low).
    - International financial integration (total external assets and liabilities to GDP; high vs. low).
    - Fiscal dominance (central bank loans to government relative to GDP; high vs. low).
    - Sovereign-bank nexus (banks’ claims on public sector to total bank assets; high vs. low).
    - Exchange rate flexibility (volatility of NEER; high vs. low).

### IV. Baseline Results
A. Transmission to financial markets — empirical findings
- General:
  - Substantial heterogeneity in speed and strength of pass-through across financial variables and economies.
  - Impairment of channels consistent with structural features documented in Section II.
- Money market (interbank) rates:
  - Figures use a one-percentage-point contractionary monetary policy shock over eight quarters; shaded areas represent 90-percent confidence bands (Newey–West).
  - Immediate and persistent rise in money market rates occurs in a limited number of economies; magnitude varies.
  - Strongest, most persistent, and statistically significant responses in: South Africa, Ghana, Mozambique, and to some extent, Uganda.
  - Positive but short-lived and statistically significant responses in: Kenya, Tanzania, Zambia.
  - Weak pass-through in some countries may reflect lack of operational tools to align market rates with intended policy or liquidity management adjustments not captured by policy rate changes.
- Government treasury bill (T-bill) yields:
  - T-bill yields increase following contractionary shocks in most economies; responses largely mirror interbank rates in magnitude.
  - Positive, statistically significant, and persistent responses in economies where interbank responses are strongest (e.g., South Africa, Ghana, Mozambique).
  - Positive but short-lived responses in most remaining countries.
  - Muted responses in economies with modest interbank impacts (e.g., Mauritius, Nigeria).
- Government bond (T-bond) yields:
  - Contractionary shock generally raises bond yields, but responses are less pronounced than T-bills (reflecting term and risk premia).
  - Positive, statistically significant, and relatively persistent responses in: Ghana, Tanzania, and Uganda.
  - More modest, short-lived, or less precisely estimated effects in remaining economies.
- Stock prices:
  - Contractionary shocks lower stock prices; statistically significant declines only in a small number of economies with stock market data.
  - Most pronounced declines in: Nigeria, South Africa, and Zambia.
  - Stronger asset price channel in South Africa consistent with more developed capital markets.
- Exchange rates (NEER):
  - A one-percentage-point contractionary shock leads to expected appreciation only in Kenya (positive and statistically significant).
  - In most other countries, currencies depreciate following monetary tightening; examples:
    - Sharp depreciation in Nigeria.
    - Depreciation to a lesser extent in Zambia, South Africa, and Ghana.
  - Limited responsiveness likely reflects less liquid financial markets, restricted capital mobility, and low global financial integration; foreign exchange intervention and exchange rate management may further damp responses.
  - The domestic depreciation following tightening (exchange rate puzzle) observed in most SSA economies aligns with EME evidence and has been linked to fiscal dominance, high foreign-currency debt shares, and elevated risk premia.
  - Broader evidence: Bolhuis et al. (2024) find a 100-basis-point shock leads to sharp and persistent NEER appreciation in AEs but only modest and short-lived depreciation in EMEs; other studies (Checo et al., 2024) find no puzzle in some EME samples.

*Source: IMF Working Paper — Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets (sections II–IV).*

### 0.5 percent, with effects dissipating within days to weeks.

### wpiea2026102-source-pdf - 0.5 percent, with effects dissipating within days to weeks.

### Exchange rate responses to monetary policy shocks
- Impulse response functions of nominal effective exchange rates (NEER) are estimated for a one-percentage-point contractionary monetary policy shock over the eight quarters following the shock, using the local projections method.
- A decline in the response indicates a depreciation of the NEER.
- Monetary policy shocks are identified following Romer and Romer (2004).
- The shaded areas in the figures represent 90-percent confidence bands constructed using Newey–West standard errors.
- Country-specific NEER impulse responses are reported for: GHA = Ghana; KEN = Kenya; MUS = Mauritius; MOZ = Mozambique; NGA = Nigeria; TZA = Tanzania; UGA = Uganda; ZAF = South Africa; ZMB = Zambia.
- Empirical note: Solis (2023) shows, using intraday asset price data for Mexico, that significant exchange rate responses to policy rate surprises are only detectable at intraday frequency.

### Deposit rate responses
- A monetary tightening generally leads to an increase in deposit rates across most economies, with effects that tend to be persistent.
- Responses are strongest and statistically significant in South Africa and, to a lesser extent, in Ghana and Uganda; responses are more muted or short-lived in the remaining economies.
- Typical magnitudes reported:
  - In South Africa: a 100-basis-point contractionary monetary policy shock raises deposit rates by about 1.5 percentage points after three quarters.
  - In Ghana and Uganda: a 100-basis-point contractionary monetary policy shock raises deposit rates by about 0.5 percentage point.

### Lending rate responses
- A contractionary monetary policy shock leads to positive and persistent increases in lending rates across the region, except in Mauritius and Tanzania where effects are muted.
- Peak impact timing: in most economies the peak occurs after about three to four quarters.
- Strongest and statistically significant effects observed in South Africa, Ghana, and Mozambique.
- Typical magnitudes reported:
  - In South Africa, Ghana, and Mozambique: a 100-basis-point monetary policy shock increases lending rates by about 1.5-2 percentage points after four quarters, with effects persisting thereafter.
- Possible channel explanation: strong pass-through to lending rates in some economies may partly reflect the prevalence of variable-rate loans (for instance, most mortgage loans in South Africa are on variable terms (IMF, 2024)).
- In most economies, deposit and lending rates respond with broadly similar timing and magnitude, suggesting banks do not systematically maintain or widen interest margins following policy tightening.
- Exceptions where lending rates respond more strongly than deposit rates: Ghana, Mozambique and Zambia.

### Aggregate interpretation: bank-based interest rate channel
- Overall finding: bank deposit and lending rates exhibit significant and persistent increases following monetary tightening in the majority of countries, pointing to the bank-based interest rate channel as the dominant transmission mechanism.
- This pattern aligns with micro-level evidence for SSA countries cited in the text (Abuka et al., 2015; Conesa et al., 2025 for Uganda; Loate and Viegi, 2021; Pirozhkova and Viegi, 2024 for South Africa).
- Consistent with theoretical predictions about monetary transmission in financially shallow economies where households and firms rely heavily on bank-based financing (IMF, 2016).

### Macroeconomic responses to monetary policy shocks (real GDP and headline CPI)
- Method: impulse response estimates of real GDP and headline CPI for SSA EFMs as a whole and for two subgroups—SSA EFM IT (inflation-targeting) and SSA EFM MT (monetary-aggregate-targeting and eclectic regimes)—following contractionary monetary policy shocks; estimated using local projections with country and time fixed effects; 90-percent confidence bands constructed using standard errors clustered at the country level.
- General result: weak and uncertain monetary policy transmission in SSA EFMs, consistent with literature documenting weaker transmission mechanisms in LICs (Mishra et al., 2010; Mishra and Montiel, 2012).
- Average effects for SSA EFMs:
  - Real GDP: a 100-basis-point monetary policy shock reduces real GDP by about 0.2 percent at its peak, occurring between six and seven quarters after the shock; effects remain persistent thereafter but are statistically insignificant.
  - Headline CPI: declines by roughly 0.4 percent about two years after the shock; effects remain statistically insignificant throughout.
- Timing: output responds before inflation, consistent with conventional view that monetary policy affects the real economy with long and variable lags and transmission unfolds gradually over time.

### Heterogeneity by monetary policy framework
- Transmission is stronger in economies that have adopted, or are transitioning toward, formal inflation-targeting frameworks (SSA EFM IT) compared with monetary-aggregate-targeting economies (SSA EFM MT).
- Reported subgroup magnitudes:
  - SSA EFM IT: a 100-basis-point shock reduces real GDP by around 0.3 percent at its peak (four to six quarters after the shock).
  - SSA EFM MT: a smaller and statistically insignificant decline of about 0.2 percent at the trough (around eight quarters after the shock).
  - CPI responses:
    - SSA EFM IT: CPI falls by approximately 1 percent two years after the shock (reported as stronger and statistically significant).
    - SSA EFM MT: an insignificant decline of about 0.2 percent.

### Cross-region comparison (SSA EFMs vs AEs and EMEs)
- Using the same panel local projection framework, monetary policy transmission in SSA EFMs is compared with that in advanced economies (AEs) and major emerging market economies (EMEs).
- Key comparative findings:
  - Monetary policy transmission in SSA EFMs is notably weaker than in AEs and EMEs.
  - In AEs: real GDP declines by approximately 0.4 percent after one year, reaching a trough of around 1 percent after eight quarters.
  - In EMEs: real GDP declines by about 0.3 percent after one year and remains at roughly that level after two years.
  - By contrast, the peak contraction in SSA EFMs is about 0.2 percent (see subgroup and aggregate figures above).

*IMF Working Paper — Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets (figures and text excerpts provided in source PDF).*

### 0.2 percent and occurs after roughly eight quarters.

### wpiea2026102-source-pdf - 0.2 percent and occurs after roughly eight quarters.

### Major empirical findings on transmission magnitudes and timing
- A 100-basis-point contractionary monetary policy shock produces:
  - For the pooled SSA EFM sample: a peak output impact of 0.2 percent and occurs after roughly eight quarters.
  - Prices fall by approximately 1 percent in AEs and EMEs at their respective troughs, around two years after the shock, compared with a more modest decline of about 0.4 percent in SSA EFMs over the same horizon.
- Robustness of identification:
  - Using Consensus Forecasts instead of World Economic Outlook projections yields similar results for AEs and EMEs.
  - Correlations between sources: one-year-ahead inflation expectations correlation is 0.85 for AEs and 0.94 for EMEs; one-year-ahead GDP growth expectations correlation is 0.84 for AEs and 0.88 for EMEs.
- Country-level highlights:
  - South Africa: real GDP declines gradually to about 3 percent at its trough eight quarters after the shock.
  - Statistically significant output contractions also observed in Mozambique, Nigeria, Uganda, and Zambia.
  - In many SSA EFMs, output responses are muted and less precisely estimated, but the direction is broadly negative.

### Heterogeneity in price responses and the price puzzle
- Price responses across SSA EFMs:
  - Full disinflationary effects emerge only with long lags and are not statistically significant in most economies.
  - Statistically significant price effects are observed mainly in some inflation-targeting regimes (e.g., Ghana, Kenya, and Uganda).
  - Evidence of a price puzzle (an increase in price level following a monetary policy contraction) is present in a few economies (e.g., CEMAC, Mozambique, Nigeria, and South Africa).
- Contrast across country groups:
  - AEs and EMEs: larger disinflationary peak (~1 percent) around two years.
  - SSA EFMs: more modest price decline (~0.4 percent) over two years.

### Financial channel evidence (first-stage transmission)
- Interest rate channels:
  - Effects on money market rates, government securities yields, and bank interest rates are statistically significant in most SSA EFMs.
  - Strong and persistent response of bank lending rates implies the bank lending channel is broadly operative.
  - Given bank-based financial systems in SSA EFMs, the bank-based interest rate channel is likely the dominant transmission mechanism.
- Exchange rate and asset price channels:
  - Operative in only a limited number of economies, consistent with documented structural impediments.
  - Exchange rate channel appears notably weak; a 100-basis-point monetary policy shock leads to an appreciation of the NEER of less than 0.5 percent on impact on average.
- Table 1 summary (significance within two years):
  - Statistically significant variables (at least at the 10 percent level) with expected sign include money market, T-bill/T-bond, deposit, lending rates in many countries; NEER and stock prices significant in fewer economies.
  - Inflation-targeting countries (grey shaded in table) show more cases with statistically significant CPI responses.

### Structural drivers of cross-country heterogeneity
- Financial development:
  - Economies classified as highly financially developed: trough impact on real GDP reaches approximately 1 percent after seven to eight quarters and is statistically significant.
  - Less financially developed economies: trough impact on real GDP is less than 0.2 percent.
  - Price effects do not conform clearly to priors; some evidence of a price puzzle in more financially developed economies around six to eight quarters.
- Monetary policy transparency:
  - High-transparency economies exhibit larger and statistically significant effects of monetary policy shocks on output and prices.
  - Greater transparency strengthens transmission by improving private sector interpretation of policy and expectations formation.
- Fiscal dominance:
  - Fiscal dominance proxied by central bank credit to government relative to GDP.
  - Monetary policy transmission is stronger in economies with low fiscal dominance; output and consumer prices contract more sharply and responses are statistically significant.
- Sovereign-bank nexus:
  - Measured by banks’ public sector claims relative to bank assets.
  - Real GDP declines gradually in economies with a low sovereign-bank nexus; responses are smaller and less precisely estimated in high-nexus economies.
  - Price effects are not statistically significant in either group.
  - Elevated sovereign exposures across the sample may limit ability to detect strong nexus effects.
- Exchange rate flexibility:
  - Proxied by NEER volatility (standard deviation of quarterly log differences).
  - Monetary policy shocks have statistically insignificant effects on output across high- and low-flexibility groups.
  - Consumer prices decline more steeply and significantly in economies with lower exchange rate flexibility than in those with higher flexibility.
  - Only three sample economies classified as floating in 2023 according to IMF de facto classification: Mauritius, South Africa, and Uganda.

### Broader financial conditions and macroeconomic effects
- Financial conditions indices (FCIs) results:
  - A one percent increase in the FCI (tighter financial conditions) exerts economically meaningful and statistically significant contractionary effects on output in most economies.
  - Peak output effects typically occur after about four to six quarters and remain persistent.
  - Disinflationary effects are strong and statistically significant in most SSA EFMs, with prices continuing to decline for up to two years following the shock.
- Interpretation caveats:
  - FCIs capture monetary policy effects but also uncertainty and domestic or external shocks beyond central bank control.
  - SSA EFM FCIs incorporate global variables (e.g., VIX and EMBI), so estimated effects capture combined domestic monetary policy, broader domestic financial conditions, and global financial developments.

### Robustness checks and methodological validation
- Alternative exchange rate measures:
  - Bilateral U.S. dollar exchange rate analyses yield country-level results broadly consistent with baseline NEER specification, showing muted exchange rate responses or exchange rate puzzle in most economies.
- Alternative monetary policy measure:
  - Using changes in the actual policy rate yields broadly similar output responses for SSA EFM groups but produces a pronounced price puzzle across country groups.
  - The identified monetary policy shock reduces the price puzzle relative to the actual policy rate, supporting the identification strategy.
- Alternative estimation approaches:
  - VAR framework with identified monetary policy shock yields qualitatively similar dynamics to local projections, with generally smaller magnitudes and comparable cross-country heterogeneity.
  - Replacing identified shock with actual policy rate in the VAR produces smaller output effects and a pronounced price puzzle.
- Interaction and statistical significance:
  - Differences in responses are statistically significant for financial development (for output) and monetary policy transparency (for both output and prices).
  - Differences related to exchange rate flexibility are statistically significant mainly for inflation.
  - Fiscal dominance and sovereign-bank nexus differences generally not statistically significant.

### Policy implications and recommended reform priorities
- Strengthen monetary policy frameworks:
  - Improvements in operational and institutional features of monetary policy (including greater transparency) can enhance transmission effectiveness.
- Develop domestic financial systems:
  - Higher levels of financial development appear to bolster the effectiveness and persistence of monetary transmission—particularly the bank lending channel.
- Address fiscal dominance:
  - Lower fiscal dominance is associated with stronger transmission; fiscal discipline and limiting central bank credit to government may improve monetary policy effectiveness.
- Recognize limits of interest-rate interventions alone:
  - Interest rate hikes alone may be insufficient to systematically influence exchange rates in the region due to limited capital mobility and low global financial integration.
- Consider broader financial conditions in policy design:
  - Financial conditions (as captured by FCIs) have sizable effects on output and inflation, implying that policymakers should monitor and, where possible, address broader financial vulnerabilities and global financial spillovers.

*IMF Working Paper — Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets.*

### Annex A. Monetary Policy Shocks

### Annex A. Monetary Policy Shocks

### Sources and notes
- Sources: IMF World Economic Outlook; IMF International Financial Statistics; Haver and authors’ calculations.
- Notes: The figures show monetary policy shocks, identified following Romer and Romer (2004), in blue, and actual policy rate changes in red, expressed in percentage points.

### Panels and axis scales (figure: "Monetary Policy Shocks vs. Actual Policy Rate Changes")
- CEMAC
  - Vertical axis ticks: -1.0, -0.8, -0.6, -0.4, -0.2, 0.0, 0.2, 0.4, 0.6, 0.8, 1.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Ghana
  - Vertical axis ticks: -6.0, -4.0, -2.0, 0.0, 2.0, 4.0, 6.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Kenya
  - Vertical axis ticks: -8.0, -6.0, -4.0, -2.0, 0.0, 2.0, 4.0, 6.0, 8.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Mauritius
  - Vertical axis ticks: -2.0, -1.5, -1.0, -0.5, 0.0, 0.5, 1.0, 1.5, 2.0
  - Horizontal axis years: 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Mozambique
  - Vertical axis ticks: -8.0, -6.0, -4.0, -2.0, 0.0, 2.0, 4.0, 6.0, 8.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Nigeria
  - Vertical axis ticks: -5.0, -4.0, -3.0, -2.0, -1.0, 0.0, 1.0, 2.0, 3.0, 4.0, 5.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Tanzania
  - Vertical axis ticks: -8.0, -6.0, -4.0, -2.0, 0.0, 2.0, 4.0, 6.0, 8.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Uganda
  - Vertical axis ticks: -12.0, -10.0, -8.0, -6.0, -4.0, -2.0, 0.0, 2.0, 4.0, 6.0, 8.0, 10.0, 12.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- WAEMU
  - Vertical axis ticks: -1.0, -0.8, -0.6, -0.4, -0.2, 0.0, 0.2, 0.4, 0.6, 0.8, 1.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- South Africa
  - Vertical axis ticks: -3.0, -2.5, -2.0, -1.5, -1.0, -0.5, 0.0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0
  - Horizontal axis years: 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks
- Zambia
  - Vertical axis ticks: -12.0, -10.0, -8.0, -6.0, -4.0, -2.0, 0.0, 2.0, 4.0, 6.0, 8.0, 10.0, 12.0
  - Horizontal axis years: 2007 2009 2011 2013 2015 2017 2019 2021 2023
  - Legend entries: Policy Rate Changes; Monetary Policy Shocks

### Key visual elements (figure legend)
- Monetary policy shocks are shown in blue.
- Actual policy rate changes are shown in red.
- Values are expressed in percentage points.

*IMF WORKING PAPER Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets — INTERNATIONAL MONETARY FUND*

### Annex B. Robustness Checks

### Annex B. Robustness Checks

### B.1 Results for financial markets
- The figures present impulse response functions of USD exchange rates (USD per unit of local currency; an increase indicates an appreciation) to a one-percentage-point contractionary monetary policy shock over the eight quarters following the shock, estimated using the local projections method.
- Monetary policy shocks are identified following Romer and Romer (2004).
- Confidence bands:
  - Shaded areas represent 90-percent confidence bands constructed using Newey–West standard errors.
- Country abbreviations shown in figures:
  - GHA = Ghana; KEN = Kenya; MUS = Mauritius; MOZ = Mozambique; NGA = Nigeria; TZA = Tanzania; UGA = Uganda; ZAF = South Africa; ZMB = Zambia.

### B.2 Results for macroeconomic responses
- The figures show impulse response functions of real GDP and CPI to a one-percentage-point contractionary monetary policy over the eight quarters following the shock, estimated using the local projections method.
- Panel regressions include country and time fixed effects.
- Two response measures plotted:
  - Thick blue line: response to a one-percentage-point contractionary monetary policy shock.
  - Red dashed line: response to a one-percentage-point increase in the actual policy rate over the eight quarters following the shock.
- Identification and inference:
  - Monetary policy shocks identified following Romer and Romer (2004).
  - Shaded areas represent the 90 percent confidence bands, constructed using standard errors clustered at the country level.
- Country/group abbreviations:
  - SSA EFM = SSA emerging and frontier market economies; SSA EFM IT = SSA emerging and frontier market economies, inflation targeters; SSA EFM MT = SSA emerging and frontier market economies, monetary-aggregate–targeting and eclectic regime.
  - AE = Advanced economies; EME = Emerging market economies.
- Additional VAR robustness checks:
  - Parsimonious VARs comprising real GDP, consumer price inflation, and a cumulated measure of monetary policy shocks (or actual policy rate for alternative VARs) with recursive identification (monetary policy variable ordered last).
  - Shaded areas denote 90 percent confidence intervals.
  - Figures reported for: CEMAC, GHA, KEN, MOZ, MUS, NGA, TZA, UGA, WAEMU, ZAF, ZMB.

### B.3 Cross-country heterogeneity
- Alternative measures of monetary policy frameworks:
  - Figures contrast responses for economies with high vs. low monetary policy framework strength using the IAPOC index (Unsal et al. (2022)); CEMAC and WAEMU classified as having weak frameworks because index not available.
  - Confidence bands: 90 percent, clustered standard errors.
- Central bank independence:
  - Figures contrast high vs. low central bank independence using the CBI index (Romelli (2022, 2025)); Mozambique classified as low because index not available.
  - Confidence bands: 90 percent, clustered standard errors.
- Panel local projections with structural interaction terms:
  - Estimated specification (notation preserved from source):
    - y_{j,t+h} − y_{j,t−1} = α_{j,h} + α_{t,h} + β^{HIGH}_h D_j × MPS_{j,t−1} + β^{LOW}_h (1−D_j) × MPS_{j,t−1} + Σ_{k=1}^K θ_{k,h} X_{j,t−k} + ε_{j,t+h,h}
    - D_j = 1 if country’s mean value of given structural characteristic exceeds the SSA EFM sample mean; D_j = 0 otherwise.
    - β^{HIGH}_h and β^{LOW}_h capture cumulative impulse responses for economies with high and low levels of each characteristic to a monetary policy shock at horizon h, relative to horizon −1.
    - Null hypothesis tested: H0: β^{HIGH}_h = β^{LOW}_h.
  - Structural characteristics considered: financial development, monetary policy transparency, fiscal dominance, sovereign‑bank nexus, exchange rate flexibility.

- Statistical tests (p-values) for differences in GDP responses across groups (Table B.1):
  - Horizon 1: Financial Development 0.184; Monetary Policy Transparency 0.515; Exchange Rate Flexibility 0.063*; Fiscal Dominance 0.625; Sovereign‑Bank Nexus 0.258
  - Horizon 2: Financial Development 0.086*; Monetary Policy Transparency 0.117; Exchange Rate Flexibility 0.188; Fiscal Dominance 0.679; Sovereign‑Bank Nexus 0.848
  - Horizon 3: Financial Development 0.023**; Monetary Policy Transparency 0.076*; Exchange Rate Flexibility 0.190; Fiscal Dominance 0.462; Sovereign‑Bank Nexus 0.721
  - Horizon 4: Financial Development 0.034**; Monetary Policy Transparency 0.064*; Exchange Rate Flexibility 0.249; Fiscal Dominance 0.686; Sovereign‑Bank Nexus 0.946
  - Horizon 5: Financial Development 0.028**; Monetary Policy Transparency 0.053*; Exchange Rate Flexibility 0.516; Fiscal Dominance 0.881; Sovereign‑Bank Nexus 0.939
  - Horizon 6: Financial Development 0.048**; Monetary Policy Transparency 0.123; Exchange Rate Flexibility 0.519; Fiscal Dominance 0.701; Sovereign‑Bank Nexus 0.715
  - Horizon 7: Financial Development 0.052*; Monetary Policy Transparency 0.273; Exchange Rate Flexibility 0.616; Fiscal Dominance 0.500; Sovereign‑Bank Nexus 0.579
  - Horizon 8: Financial Development 0.059*; Monetary Policy Transparency 0.513; Exchange Rate Flexibility 0.478; Fiscal Dominance 0.393; Sovereign‑Bank Nexus 0.481
  - Notes: The table reports p-values for the null hypothesis that β^{HIGH}_h = β^{LOW}_h. ***, **, and * denote statistical significance at the 1, 5, and 10 percent levels, respectively.

- Statistical tests (p-values) for differences in CPI responses across groups (Table B.2):
  - Horizon 1: Financial Development 0.896; Monetary Policy Transparency 0.015**; Exchange Rate Flexibility 0.448; Fiscal Dominance 0.420; Sovereign‑Bank Nexus 0.141
  - Horizon 2: Financial Development 0.335; Monetary Policy Transparency 0.065*; Exchange Rate Flexibility 0.256; Fiscal Dominance 0.919; Sovereign‑Bank Nexus 0.440
  - Horizon 3: Financial Development 0.714; Monetary Policy Transparency 0.672; Exchange Rate Flexibility 0.046**; Fiscal Dominance 0.338; Sovereign‑Bank Nexus 0.886
  - Horizon 4: Financial Development 0.246; Monetary Policy Transparency 0.771; Exchange Rate Flexibility 0.008***; Fiscal Dominance 0.281; Sovereign‑Bank Nexus 0.934
  - Horizon 5: Financial Development 0.194; Monetary Policy Transparency 0.230; Exchange Rate Flexibility 0.008***; Fiscal Dominance 0.423; Sovereign‑Bank Nexus 0.543
  - Horizon 6: Financial Development 0.341; Monetary Policy Transparency 0.195; Exchange Rate Flexibility 0.006***; Fiscal Dominance 0.223; Sovereign‑Bank Nexus 0.629
  - Horizon 7: Financial Development 0.180; Monetary Policy Transparency 0.414; Exchange Rate Flexibility 0.000***; Fiscal Dominance 0.124; Sovereign‑Bank Nexus 0.859
  - Horizon 8: Financial Development 0.163; Monetary Policy Transparency 0.503; Exchange Rate Flexibility 0.000***; Fiscal Dominance 0.118; Sovereign‑Bank Nexus 0.881
  - Notes: The table reports p-values for the null hypothesis that β^{HIGH}_h = β^{LOW}_h. ***, **, and * denote statistical significance at the 1, 5, and 10 percent levels, respectively.

- Interaction-specification impulse response figures:
  - Financial Development: responses plotted for High vs. Low financial development (90 percent confidence bands, clustered SEs).
  - Monetary Policy Transparency: responses plotted for High vs. Low transparency.
  - Fiscal Dominance: responses plotted for High vs. Low degree of fiscal dominance proxied by central bank loans to government-to-GDP ratio; alternative measure proxied by government interest-to-revenue ratio also shown.
  - Sovereign‑Bank Nexus: responses plotted for High vs. Low public sector claims / bank assets.
  - Exchange Rate Flexibility: responses plotted for High vs. Low exchange rate flexibility.
  - All interaction figures estimated using local projections with country and time fixed effects; monetary policy shocks identified following Romer and Romer (2004); shaded areas represent 90 percent confidence bands constructed using standard errors clustered at the country level.

- VAR robustness:
  - Figures present VAR-based impulse responses for GDP and CPI to a one-percentage-point contractionary monetary policy shock over eight quarters using parsimonious VARs (real GDP, CPI, and cumulated monetary policy shocks or actual policy rate), recursive identification, and 90 percent confidence intervals.
  - Country abbreviations reported in VAR figures: GHA = Ghana; KEN = Kenya; MUS = Mauritius; MOZ = Mozambique; NGA = Nigeria; TZA = Tanzania; UGA = Uganda; ZAF = South Africa; ZMB = Zambia.

*IMF Working Papers — Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets*

### Annex   C. Data Sources

### Annex   C. Data Sources

### Table C.1 : List of Countries
- Emerging markets
  - South Africa
  - Australia
  - Canada
  - Euro area
  - Japan
  - Republic of Korea
  - Norway
  - Switzerland
  - Sweden
  - United Kingdom
  - United States
- Frontier markets
  - Ghana
  - Kenya
  - Mauritius
  - Mozambique
  - Nigeria
  - Tanzania
  - Uganda
  - Zambia
- Monetary Unions
  - CEMAC
    - Cameroon
    - Central African Rep.
    - Chad
    - Rep. of Congo
    - Equatorial Guinea
    - Gabon
  - WAEMU
    - Benin
    - Burkina Faso
    - Côte d’Ivoire
    - Guinea-Bissau
    - Mali
    - Niger
    - Senegal
    - Togo
- Advanced Economies
  - Australia
  - Canada
  - Euro area
  - Japan
  - Republic of Korea
  - Norway
  - Switzerland
  - Sweden
  - United Kingdom
  - United States
- Emerging Markets (additional listed)
  - Brazil
  - Chile
  - Colombia
  - Hungary
  - India
  - Indonesia
  - Mexico
  - Peru
  - Philippines
  - Poland
  - Russian Federation
  - Thailand
  - Turkey
- Frontier Markets (additional listed)
  - Ghana
  - Kenya
  - Mauritius
  - Mozambique
  - Nigeria
  - Tanzania
  - Uganda
  - Zambia

### Table C.2 : Classification of Exchange Rate Arrangements and Monetary Policy Frameworks in SSA Emerging and Frontier Markets
- Sources: AREAER database 2023; and websites of central banks.
- Notes:
  - 1/ Includes countries that have no explicitly stated nominal anchor, but rather monitor various indicators in conducting monetary policy.
  - 2/ *: The policy rate is the main instrument to signal the stance of monetary policy. +: The central bank uses a variety of instruments to align the operating target with the policy rate.
  - 3/ The central bank is in transition toward inflation-targeting.
- Exchange rate anchor listings (as presented):
  - US dollar
  - Euro
  - Composite
  - Other
  - Other 1/
- Country-level markers (symbols preserved as in source):
  - Cameroon **
  - Central African Rep. **
  - Chad **
  - Congo, Rep. of *
  - Equatorial Guinea **
  - Gabon **
  - Benin **
  - Burkina Faso *
  - Côte d’Ivoire **
  - Guinea-Bissau **
  - Mali *
  - Niger **
  - Senegal **
  - Togo **
  - Mozambique 3/ **
  - Tanzania *+
  - Nigeria *+*
  - Kenya **
  - Zambia **
  - Ghana * (Other managed arrangement (1))
  - Mauritius **
  - South Africa * *
  - Uganda * *
- Exchange rate arrangement categories (as presented):
  - Other managed arrangement (1)
  - Stabilized arrangement (2)
  - Conventional peg (14)
  - Central African Economic and Monetary Union (CEMAC) (6)
  - West African Economic and Monetary Union (WAEMU) (8)
  - Crawl-like arrangement (3)
  - Floating (3)
- Monetary policy framework headings (as presented):
  - Exchange rate anchor
  - Monetary aggregate target
  - Inflation-targeting framework
  - Main instruments
  - Policy interest rate 2/
  - Open market operations

### Table C.3 : Data Sources and Description
- Real GDP
  - Description: Real GDP in local currency units, interpolated to quarterly frequency using the quadratic match method if quarterly series are not available.
  - Source: Haver Analytics
- Nominal GDP
  - Description: Nominal GDP in local currency units, interpolated to quarterly frequency using the quadratic match method if quarterly series are not available.
  - Source: Haver Analytics
- Central bank loans to government
  - Description: Claims on the central government.
  - Source: IMF MFS
- General interest payments
  - Description: Interest payments on government debt (including long-term bonds, long-term loans, and other debt instruments) to domestic and foreign residents.
  - Source: IMF MFS
- General government revenue
  - Description: Total inflow of resources (taxes, fees, charges, grants, and other income) received by the government.
  - Source: IMF WEO
- Output gap
  - Description: Measured as the Hodrick–Prescott (HP)–filtered quarterly real GDP with a smoothing parameter λ = 1600.
  - Source: IMF WEO
- Policy rate
  - Description: Main nominal official (policy) interest rate.
  - Source: Haver Analytics
- NEER
  - Description: Nominal effective exchange rate.
  - Source: IMF IFS
- USD exchange rate
  - Description: Exchange rate; U.S. dollar per domestic currency.
  - Source: IMF IFS
- CPI inflation
  - Description: Consumer price index.
  - Source: Haver Analytics
- Inflation expectations
  - Description: One-, three-, and five-year-ahead forecasts for CPI inflation. Data are interpolated to quarterly frequency using the quadratic match method.
  - Source: IMF WEO
- GDP growth expectations
  - Description: One-, three-, and five-year-ahead forecasts for GDP growth. Data are interpolated to quarterly frequency using the quadratic match method.
  - Source: IMF WEO
- Commodity prices, all index
  - Description: World commodity price index (2016 = 100), USD, including both fuel and non-fuel price indices.
  - Source: Refinitiv Eikon; IMF
- Commodity prices, excl. fuel
  - Description: World commodity price index excluding fuel (2016 = 100), USD, including precious metals, food and beverages, and industrial inputs price indices.
  - Source: Refinitiv Eikon; IMF
- Commodity food prices
  - Description: World food price index (2016 = 100), USD, including cereals, vegetable oils, meat, seafood, sugar, and other food items (apples, bananas, chana [legumes], fishmeal, groundnuts, milk [dairy], tomatoes).
  - Source: Refinitiv Eikon; IMF
- Oil prices
  - Description: Brent crude oil price in USD.
  - Source: Refinitiv Eikon; IMF
- Bank lending rates
  - Description: Rate charged by other depository corporations that typically meets the short- and medium-term financing needs of the private sector.
  - Source: IMF IFS
- Bank deposit rates
  - Description: Rates offered to resident customers for demand, time, or savings deposits.
  - Source: IMF IFS
- Money market interest rates
  - Description: Short-term lending rate between financial institutions or interbank interest rate, depending on country availability.
  - Source: IMF IFS
- Government treasury bills
  - Description: Yield on short-term government debt securities (maturity of less than one year), depending on country availability.
  - Source: IMF IFS; Refinitiv Eikon; Haver Analytics
- Government treasury bonds
  - Description: Yields to maturity on government bonds or other securities indicating longer-term rates (maturity greater than two years), depending on country availability.
  - Source: IMF IFS; Refinitiv Eikon; Haver Analytics
- Stock market indices
  - Description: Price indices of national stock exchanges.
  - Source: Refinitiv Eikon; Haver Analytics
- U.S. monetary policy rate
  - Description: Federal Funds target rate.
  - Source: Haver Analytics
- Euro area policy rate
  - Description: Main refinancing rate.
  - Source: Haver Analytics
- VIX
  - Description: Chicago Board Options Exchange Volatility Index.
  - Source: Haver Analytics
- FCI (EFMs)
  - Description: Defined following Tiedemann et al. (2025).
  - Source: Tiedemann et al. (2025)
- Financial development index
  - Description: Defined following Sahay et al. (2015).
  - Source: IMF Financial Development Index Database
- Monetary policy transparency index
  - Description: Defined following Dincer et al. (2022).
  - Source: Dincer et al. (2022)
- Bank exposures to the public sector
  - Description: Ratio of banks’ total claims on the public sector (loans and debt securities) to total bank assets.
  - Source: IMF MFS
- International financial integration
  - Description: Ratio of a country’s stock of foreign assets and liabilities to GDP.
  - Source: External Wealth of Nations Database

*IMF Working Paper: Monetary Policy Transmission in Sub-Saharan Africa: Evidence from Emerging and Frontier Markets, Annex C. Data Sources*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026102-source-pdf.pdf_