## Does Money Matter for Inflation in Ghana? — Section 1–3 (_wp11274)

## Source details

**Canonical URL:** [Does Money Matter for Inflation in Ghana? — Section 1–3 (_wp11274)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11274.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11274.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11274.pdf.json)

---

### Abstract and key conclusions
- Money has only limited information value for future inflation in Ghana over a typical monetary policy implementation horizon (four to eight quarters).
- Currency depreciation and demand pressures (as measured by the output gap) are important predictors of future price changes.
- Inflation inertia is high and inflation expectations are largely based on backward-looking information, suggesting that inflation expectations are not well anchored.
- Strengthening the credibility of Ghana’s inflation-targeting regime is needed.

### Financial reform and institutional background
- Reforms and outcomes:
  - Since early 1980s: dismantling of exchange and credit controls, discrete devaluations, development of local foreign exchange market, shift to indirect liquidity control with open market operations (early 1990s).
  - Treasury and central bank bills introduced prior to open market operations.
  - Later reforms: modernization of banking laws, further exchange system liberalization, partial external capital account liberalization, shift to a floating exchange rate regime.
  - Outcomes included: foreign investor access to longer-end of domestic bond market; residents permitted foreign-currency bank accounts; exchange rate largely market-determined; monetary policy focused on inflation targeting; domestic capital markets remain nascent and shallow; new payment instruments and electronic banking reducing cash demand.
- Institutional modernization:
  - 2002 Bank of Ghana Act strengthened central bank independence and mandated price stability as the primary objective.
  - Monetary policy committee established in 2002.
  - Formal inflation targeting announced May 2007; Bank of Ghana had already moved away from reserve money targeting.

### Rationale for inflation targeting in Ghana
- Financial system changes can destabilize money demand and monetary transmission.
- When money–inflation relationships shift, monetary targets lose transparency and signaling power.
- Inflation targeting does not rely on stable money demand but recognizes money may still convey useful information.

### Model framework and expectations formation
- Phillips-curve and forecasting variables:
  - π_{t+1}: four-quarter change in CPI at t+1.
  - E_{t}π_{t+1}: four-quarter expected inflation at t+1 based on information at t.
  - Output gap = y_{t} − y_{t}^{trend}, with y_{t}^{trend} estimated via Hodrick-Prescott filter.
  - X_{t}: other indicators (real money gap and exchange rate).
  - ε_{t+1}: normally distributed error term.
- Expectations specification and interpretation:
  - Proxy for survey expectations constructed following Gerlach and Svensson (2003) due to data limitations.
  - Assumed constant twelve-month implicit inflation objective: π* = 10 percent for estimation period.
  - Parameter α measures weight the public places on past inflation relative to the target:
    - α = 0 ⇒ expectations fully anchored to inflation objective.
    - Large α ⇒ expectations strongly backward-looking (weak credibility).

### Model specification (empirical approach)
- Estimable inflation-gap equation explains future inflation gaps using:
  - lagged inflation gap (inertia),
  - output gap,
  - real money gap (level),
  - four-quarter change in the real money gap (change),
  - currency depreciation (change in nominal exchange rate).
- Real broad money (M2+) includes foreign currency deposits; real money gap constructed relative to optimal long-run real broad money stock.

### Long-run money demand (1990Q1-2009Q4)
- Estimated properties:
  - Strong income elasticity.
  - Significant currency substitution: cedi depreciation raises demand for broad money via valuation and substitution effects.
  - No robust interest rate channel evidence over the sample, consistent with recent financial deepening.
- Cointegration and long-run estimates (selected reported parameters):
  - M2_PLUS_REAL_LOG(-1) -1.00
  - RGDP_LOG(-1) 1.58
  - STDEV 0.13
  - T-STAT 11.99
  - D_NEER(-1) -3.98; STDEV 1.29; T-STAT -3.09
  - DEP_LOG(-1) -2.91; STDEV 2.55; T-STAT -1.14
  - CONSTANT -4.00
  - ALPHA -0.11; STDEV -0.03; T-STAT -3.36
- Cointegration tests:
  - Trace test indicates 1 cointegrating eqn(s) at the 0.05 level.
  - Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level.
- Note: MacKinnon-Haug-Michelis (1999) p-values referenced.

### Empirical findings on money and inflation
- Cross-correlation evidence (Figure 2 descriptions):
  - Vertical axis range: – 0.8 to 0.8; T ranges from -8 to 8 quarters.
  - For real money and money growth gaps:
    - Same-period (T=0) cross-correlations with the inflation gap are highly positive.
    - Cross-correlations decline quickly for any T>0 → impacts are instantaneous and short-lived.
  - For the output gap:
    - Contemporaneous cross-correlation with the inflation gap ~ -0.7.
    - Positive cross-correlation emerges after a lag of four quarters and stabilizes around 0.2 thereafter.
  - For the exchange rate:
    - Changes in the nominal exchange rate have an immediate impact on the inflation gap, peaking at T=2; pass-through weakens quickly thereafter.
    - A decline (an increase) in the level of the exchange rate refers to a depreciation (appreciation); a negative cross-correlation therefore means that a currency depreciation is associated with rising inflation.
  - Example cross-correlation values referenced: contemporaneous money-gap/inflation cross-correlation small negative number (-0.2) at T=4; high positive cross-correlation (almost 0.6) at T=-2.
- Inflation-gap model estimates (four- and eight-quarters-ahead):
  - Lagged inflation gap (inflation inertia) statistically significant in all models; parameter estimates range: 0.52 to 0.97.
  - Output gap parameter estimates statistically significant:
    - average 0.94 across four-quarter lag models,
    - average 1.22 across eight-quarter lag models.
  - Currency depreciation:
    - Statistically significant at four-quarter interval; pass-through averages about 20 percent.
    - Not significant at the eight-quarter interval.
  - Money-gap indicators:
    - Both money-gap indicators fail to obtain statistically significant parameter estimates at four-quarter interval.
    - Real money growth gap indicator is statistically significant and positive in all model specifications at the eight-quarter interval (parameter estimate averages 0.11).
    - Real money gap indicator only significant when both money-gap indicators included at eight-quarter interval (Model G), likely reflecting strong cross-correlation between the two money-gap indicators.
- Diagnostics and robustness:
  - Residuals show significant autocorrelation and heteroscedasticity; Newey-West adjustments applied.
  - Ramsey’s RESET test highly significant, indicating possible measurement error endogeneity biasing OLS.
  - GMM estimates (column H of Table 2) used to address endogeneity; GMM results very similar to OLS.
  - Recursive estimation: parameter estimates stable, especially in second half of sample; some instability in 1990s.

### Comparative interpretation
- Demand pressures (output gap) transmit to inflation significantly within four quarters in Ghana—faster than evidence for the euro zone.
- Ghana exhibits strong inflation inertia (lagged inflation significant), unlike the euro zone findings.
- Empirical support for a persistent money channel to inflation in Ghana is weak; real money growth channel shows limited importance at policy horizons.
- Nominal effective exchange rate (imported inflation) is strongly important in Ghana.

### Monetary policy credibility and inflation expectations
- Using model estimates:
  - φ̄ averages 0.63.
  - Monetary policy credibility estimate equals 0.44 (characterized as rather low).
- Implications:
  - Low credibility ⇒ inflation expectations substantially influenced by backward-looking information.
  - Estimated high inflation inertia implies public expectations tied to past inflation outturns rather than being forward-looking and anchored.
  - Results depend on specification assumptions (constant implicit inflation objective π* = 10 percent) and are subject to estimation uncertainty.

### Model versus survey expectations (2005Q4-2009Q4)
- Empirical patterns:
  - When actual inflation rises (falls), inflation expectations adjust upwards (downwards) with a relatively short lag.
  - At end of 2009, expectations began to recede soon after actual inflation started to fall.
- Model vs Bank of Ghana survey:
  - Model-generated expectations are lower than survey-based expectations.
  - This suggests estimated inflation inertia may understate true (unobserved) inertia.
  - A higher assumed inertia value, = 0.9, produces model expectations that closely mimic survey patterns.
  - Model-based expectations track actual inflation in 2007 but lag when inflation rose rapidly thereafter.

### Policy recommendations and implications
- Strengthen the interest rate channel to enhance monetary policy implementation under inflation targeting.
- Continue reviewing money aggregates as part of the information set but avoid assigning significant policy weight to them given their limited information content.
- Maintain attention to exchange rate stability given strong and rapid pass-through to domestic prices.
- Enhance monetary policy credibility to reduce inflation inertia and limit the effect of temporary deviations of actual inflation on public expectations.

### Data and research priorities
- Progress in modeling Ghana’s inflation dynamics is needed; work is at an early stage.
- Strengthen data quality, particularly for real sector activities where data are missing (e.g., wages and employment) or at low frequency (annual).
- Publication milestone: Ghana Statistical Services began publishing quarterly national accounts in May 2011.

### Key numerical values and statistics (preserved verbatim)
- Money demand sample period: 1990Q1-2009Q4.
- Cross-correlation vertical axis range: – 0.8 to 0.8; T ranges from -8 to 8 quarters.
- Example cross-correlation values: (-0.2) at T=4; (almost 0.6) at T=-2.
- Lagged inflation gap parameter estimates range: 0.52 to 0.97.
- Output gap parameter averages: 0.94 (four-quarter lags models), 1.22 (eight-quarter lags models).
- Currency depreciation pass-through at four-quarter interval: averages about 20 percent.
- Real money growth gap parameter at eight-quarter interval: parameter estimate averages 0.11.
- φ̄ averages 0.63; monetary policy credibility estimate equals 0.44.
- Estimation methods referenced: OLS and GMM; robustness checks included Newey-West adjusted tests and Ramsey’s RESET test.

*Source: IMF Working Paper WP/11/274, “Does Money Matter for Inflation in Ghana?” — Sections 1–3.*

### Section 1

### Does Money Matter for Inflation in Ghana? — Section 1

### Abstract — key conclusions
- Money has only limited information value for future inflation in Ghana over a typical monetary policy implementation horizon (four to eight quarters).
- Currency depreciation and demand pressures (as measured by the output gap) are important predictors of future price changes.
- Inflation inertia is high and inflation expectations are largely based on backward-looking information, suggesting that inflation expectations are not well anchored.
- Strengthening the credibility of Ghana’s inflation-targeting regime is needed.

### I. Introduction and financial reform background
- Since early 1980s, Ghana progressively dismantled exchange and credit controls, realigned the exchange rate through a series of discrete devaluations, developed the local foreign exchange market, and shifted the Bank of Ghana to indirect control of liquidity with open market operations in early 1990s.
- Treasury and central bank bills were introduced prior to open market operations.
- Later reforms included modernization of banking laws, further liberalization of the exchange system, partial liberalization of the external capital account, and a shift to a floating exchange rate regime.
- Outcomes of reforms:
  - Foreign investors gained access to the longer-end of Ghana’s domestic bond market.
  - Ghanaian residents permitted to hold foreign currency denominated bank accounts.
  - Exchange rate largely determined by market forces; monetary policy focused on inflation targeting rather than targeting a particular exchange rate level.
  - Domestic capital markets began developing but remain nascent and shallow.
  - New payment instruments and electronic banking are reducing demand for cash and expanding rural banking services.

- Institutional modernization for monetary policy:
  - The 2002 Bank of Ghana Act strengthened central bank independence and mandated price stability as the primary objective of monetary policy (growth and exchange rate stability are seconded).
  - A monetary policy committee was established in 2002.
  - In May 2007 Ghana announced formal inflation targeting; the Bank of Ghana had already moved away from reserve money targeting toward analyzing a broader set of indicators.

- Rationale for inflation targeting in the context of money-demand instability:
  - Significant changes in financial systems can destabilize money demand and the monetary transmission mechanism.
  - When the relationship between money and inflation shifts, monetary targets lose transparency and their signaling power.
  - This argues for inflation targeting, which does not rely on stable money demand, while acknowledging that money may still convey useful information.

### II. Comparative evidence and Ghana’s inflation-targeting experience
- Literature overview (select findings preserved as stated):
  - Issing (2011): argues for a prominent role for money post-global financial crisis; “broad money growth is one of the ‘pillars’” in the euro area context.
  - Gerlach and Svensson (2003): real money gap contains statistically significant information about future euro zone inflation; broad money growth used by the ECB fails to do so.
  - Berger and Stavrev (2008a): money contains relevant information for future inflation in the euro area but marginal contribution is limited.
  - Scharnagl et al. (2010): money enhances monetary policy reaction function performance in the euro area; information value of money relates to uncertainty about current real output (available with a lag).
  - Berger and Österholm (2008b): contribution of money to U.S. inflation forecast is limited and has diminished.
  - Barnichon and Peiris (2008): real money gap has a statistically significant contemporaneous impact on inflation across 17 Sub-Saharan African countries (annual data 1970-2004).

- Ghana’s experience with inflation targeting:
  - Inflation rose from around 10 percent in mid-2007 (when formal inflation-targeting was adopted) to more than 20 percent by early 2009.
  - Inflation started to recede towards end-2009 and remained in single-digit levels since mid-2010, allowing the Bank of Ghana to ease monetary policy significantly.
  - Compared to other inflation-targeting countries, accommodative domestic policies contributed to higher inflation outcomes in Ghana.
  - Other emerging market inflation-targeting countries contained domestic inflationary pressures from higher global food and energy prices better partly by tighter monetary policies and could ease earlier than Ghana.

- Table 1 (selected preserved entries and notes):
  - Year-on-year inflation (in percent). (Entries are presented in original tabular form in the source; selected figures reproduced as quoted.)
  - Average (both groups): 6.2 4.9 -1.4 3.3 3.1 -0.2
  - Group I: 9.5 5.8 -3.7 4.1 3.9 -0.1
  - Group II: 3.0 4.0 1.0 2.5 2.3 -0.3
  - Ghana (2007q2 (May)): 10.2   20.7    10.5 10.2    20.7 10.5
  - Ghana (2002q3 (MPC formed)): 13.7   12.6 -1.1 .........

### III. Policy credibility and inflation expectations — model framework and implications
- Objective:
  - Develop a generic model describing how the public formulates inflation expectations and how these expectations relate to monetary policy credibility.
  - Key outcome: weak monetary policy credibility ⇒ the public places larger weight on past inflation outturns ⇒ expectations are backward-looking and exhibit high inertia.

- Phillips curve specification (as presented):
  - π_{t+1} = four-quarter change in the consumer price index at period t+1.
  - E_{t}π_{t+1} = four-quarter expected inflation at t+1 based on information available at t.
  - y_{t} = current-quarter real output; y_{t}^{trend} = current trend output (estimated with a Hodrick-Prescott filter).
  - Output gap = y_{t} − y_{t}^{trend}.
  - X_{t} = other exogenous indicators relevant for forecasting future inflation (such as a real money gap and the exchange rate).
  - ε_{t+1} = normally distributed error term.

- Expectations formation approach:
  - Data limitations preclude use of survey-based inflation expectations for Ghana; construct a proxy following Gerlach and Svensson (2003).
  - Public’s inflation expectations depend on the central bank’s implicit inflation objective and the credibility of that objective.
  - The implicit inflation objective may be time-varying or constant; due to data limitations the paper assumes a constant twelve-month inflation objective of 10 percent (i.e., π* = 10 percent) for the estimation period.
  - High inflation expectations are interpreted as reflecting weak monetary policy credibility (large deviations from the implicit inflation objective) rather than a high target inflation rate.
  - A constant implicit inflation objective performs better in estimation and captures the essence of inflation dynamics in Ghana.

- Parameter interpretation:
  - Parameter α in the expectations equation measures the weight that the public places on past inflation outturns relative to the targeted rate of inflation.
  - When monetary policy is fully credible (α = 0), inflation expectations are fully anchored to the central bank’s inflation objective and ex post deviations from the inflation target do not affect expectations.
  - When anchoring is weak (α large), inflation expectations are significantly influenced by past inflation trends; past deviations from the implicit inflation objective persistently shape public expectations.

*Source: IMF Working Paper WP/11/274, “Does Money Matter for Inflation in Ghana?” — Section 1.*

### Section 2

### _wp11274 - Section 2

### Model specification and empirical approach
- Combines equations (1) and (2) to produce an estimable inflation-gap equation (equation (3) reworked into equation (4)) that explains future inflation gaps in terms of:
  - lagged inflation gap (inertia),
  - output gap,
  - real money gap (level),
  - four-quarter change in the real money gap (change),
  - currency depreciation (change in the nominal exchange rate, proxy for imported inflation).
- Real broad money stock is denoted as the real broad money term (nominal broad money stock includes foreign currency deposits divided by the price level). The optimal real broad money stock is the long-run benchmark used to construct the real money gap.
- The model explicitly distinguishes between:
  - the real money gap in level (parameter expected to be positive), and
  - the change in the real money gap (parameter expected to be positive and also appearing in the level term).

### Long-run demand for money in Ghana (data and main properties)
- Long-run demand for real broad money (including foreign currency deposits) estimated over 1990Q1-2009Q4 (estimation results in Appendix I).
- Main empirical findings from the money demand estimation:
  - Evidence of strong income elasticity.
  - Significant currency substitution: when the Ghanaian cedi depreciates, demand for broad money increases via:
    - valuation effect (foreign-currency denominated deposit liabilities increase in domestic currency value), and
    - substitution effect (public shifts from cedi-denominated deposits to foreign-currency denominated deposits).
  - No robust evidence for a significant interest rate channel over the sample, consistent with relatively recent financial deepening.
- Results are consistent with Dagher and Kovanen (2011) estimating a similar money demand relation for Ghana.

### Empirical results: is money informative for future inflation in Ghana?
- Cross-correlation evidence (Figure 2):
  - Cross-correlations measured on vertical axis range from – 0.8 to 0.8; T ranges from -8 to 8 quarters on horizontal axis.
  - For real money and money growth gaps:
    - Same-period (T=0) cross-correlations with the inflation gap are highly positive.
    - Cross-correlations decline quickly for any T>0, suggesting impacts are instantaneous and short-lived.
    - Given typical monetary policy lags of four to eight quarters, these money-gap indicators are unlikely to be useful for policy.
  - For the output gap:
    - Contemporaneous cross-correlation with the inflation gap is highly negative (around -0.7).
    - A positive cross-correlation emerges after a lag of four quarters and stabilizes around 0.2 in subsequent quarters—indicating the output gap has more information value for inflation at longer horizons.
  - For the exchange rate:
    - Changes in the nominal exchange rate have an immediate impact on the inflation gap, peaking at T=2 in Figure 2; pass-through weakens quickly thereafter.
    - Note: a decline (an increase) in the level of the exchange rate refers to a depreciation (appreciation); a negative cross-correlation therefore means that a currency depreciation is associated with rising inflation.
- Table 2 — Estimates for four- and eight-quarters-ahead inflation-gap models:
  - Lagged inflation gap (inflation inertia) is statistically significant in all models; parameter estimates range from 0.52 to 0.97 across four-quarter and eight-quarter lag models.
  - Output gap parameter estimates are statistically significant in all specifications:
    - average 0.94 across four-quarter lag models,
    - average 1.22 across eight-quarter lag models.
  - Currency depreciation is statistically significant at the four-quarter interval; pass-through averages about 20 percent. Not significant at the eight-quarter interval.
  - Two money-gap indicators produce weak results:
    - Both money-gap indicators fail to obtain statistically significant parameter estimates at the four-quarter interval.
    - The real money growth gap indicator receives a statistically significant and positive parameter estimate in all model specifications at the eight-quarter interval (parameter estimate averages 0.11).
    - The real money gap indicator is only statistically significant when both money-gap indicators are included at the eight-quarter interval (Model G), but this likely reflects strong cross-correlation between the two money-gap indicators and should be treated cautiously.
- Diagnostics and estimation issues:
  - Estimated residuals show significant autocorrelation and heteroscedasticity; Newey-West test statistics were used to adjust inference.
  - Ramsey’s RESET test is highly significant (not reported), indicating correlation between explanatory variables and the error term—potentially due to measurement errors in explanatory variables (which would bias OLS).
  - Generalized Method of Moments (GMM) estimates (column H of Table 2) were used to address endogeneity; GMM results are very similar to OLS results.
- Recursive estimation:
  - Parameter estimates remained remarkably stable, particularly during the second half of the sample (period when the monetary operating framework changed).
  - Some parameter instability evident in the 1990s, likely related to macroeconomic instability and higher price/output volatility.

### Comparisons and interpretation
- Demand pressures (output gap) in Ghana are transmitted to inflation significantly within four quarters—faster than in the euro zone (Gerlach and Svensson (2003)) where only about one-quarter of demand pressures transmit within four quarters.
- Results differ from other studies (Gerlach and Svensson (2003); Barnichon and Peiris (2008)) in that:
  - Ghana shows strong inflation inertia (lagged inflation significant), unlike the euro zone where inertia is not a significant driver.
  - Ghana does not provide strong empirical support for a persistent money channel to inflation; real money growth channel shows weak importance for future inflation dynamics.
  - Nominal effective exchange rate (imported inflation) shows strong importance in Ghana and was not included in some other studies.

### Monetary policy credibility for Ghana
- Using estimates from the four-quarter and eight-quarter lag models:
  - φ̄ averages 0.63.
  - Using this, the estimate for monetary policy credibility equals 0.44, which is characterized in the text as rather low.
- Interpretation:
  - Low credibility implies that inflation expectations in Ghana are substantially influenced by backward-looking information (past inflation outturns).
  - This finding depends on assumptions about the specification of the inflation objective and the formation of inflation expectations, and the estimate for inflation inertia may be subject to estimation error.
  - Nevertheless, the results highlight the importance of past inflation outturns for public inflation expectations and underscore the need to strengthen monetary policy implementation in Ghana.

### Key numerical values and statistics (preserved verbatim)
- Sample period for money demand estimation: 1990Q1-2009Q4.
- Cross-correlation vertical axis range: – 0.8 to 0.8; T ranges from -8 to 8 quarters.
- Examples from cross-correlation discussion: contemporaneous money-gap/ inflation cross-correlation small negative number (-0.2) at T=4 in Figure 2; high positive cross-correlation (almost 0.6) at T=-2.
- Lagged inflation gap parameter estimates range: 0.52 to 0.97.
- Output gap parameter averages: 0.94 (four-quarter lags models), 1.22 (eight-quarter lags models).
- Currency depreciation pass-through at four-quarter interval: averages about 20 percent.
- Real money growth gap parameter at eight-quarter interval: parameter estimate averages 0.11.
- φ̄ averages 0.63; monetary policy credibility estimate equals 0.44.
- Estimation methods referenced: OLS and GMM; robustness checks included Newey-West adjusted tests and Ramsey’s RESET test.

*Source: _wp11274 - Section 2 (PDF: _wp11274 - Section 2)*

### Section 3

### _wp11274 - Section 3

### Inflation expectations: model versus survey
- Period analyzed: 2005Q4-2009Q4.
- Empirical patterns:
  - When actual inflation rises (falls), inflation expectations are adjusted upwards (downwards) with a relatively short lag.
  - At end of 2009, inflation expectations began to recede soon after actual inflation started to fall.
- Model versus Bank of Ghana survey:
  - The empirical model generates inflation expectations that are lower than the survey-based inflation expectations reported by the Bank of Ghana.
  - This suggests the estimated inflation inertia is likely to understate the “true” but unobserved inflation inertia.
  - A higher assumed inertia value, = 0.9, produces model expectations that mimic closely the pattern of survey-based inflation expectations.
  - Model-based inflation expectations are broadly in line with actual inflation outturns in 2007, but fell behind in the subsequent period when actual inflation increased rapidly.
- Credibility and interpretation:
  - Monetary policy credibility is difficult to measure; survey-based inflation expectations among businesses and consumers provide indications on alignment with the inflation target.
  - The paper cautions interpreting results on credibility given measurement challenges.

### Main conclusions on money and inflation dynamics
- Core question: whether money provides useful information for explaining Ghana’s inflation dynamics.
- Key findings:
  - A stable money demand relation for Ghana is found over the sample period, consistent with Dagher and Kovanen (2011).
  - Empirical support for money in the inflation process in Ghana is weak, at best.
  - The weak information value for money in Ghana is in line with empirical evidence for other countries and suggests its importance is limited and likely to decline.
  - Despite a structurally stable money demand function, money no longer remains a reliable indicator of monetary policy.
- Implication for Bank of Ghana:
  - Money aggregates may continue to be reviewed as part of the information set, but such information is likely of limited value and should not be given significant weight in policy deliberations.
  - This conclusion supports Bank of Ghana’s inflation-targeting regime (formally adopted in 2007), which underscores the role of the interest rate channel for monetary policy transmission.

### Transmission channels and dynamics
- Confirmed transmission channels:
  - Exchange rate and output gap provide key transmission channels for inflation.
- Exchange rate:
  - The pass-through to domestic prices from exchange rate changes is strong and rapid, underscoring linkages between a stable exchange rate and price stability in Ghana.
- Output gap:
  - The output gap is highly relevant for inflation, with an average impact lag to inflation about four to eight quarters.
- Inflation inertia and credibility:
  - Ghana’s weak policy credibility is embedded in the estimated high degree of inflation inertia, where inflation expectations are determined largely by past inflation outcomes rather than being forward-looking and anchored by the monetary authorities’ inflation objective.

### Policy recommendations and implications
- Strengthen the interest rate channel:
  - Efforts to strengthen the interest rate channel would enhance monetary policy implementation under inflation targeting.
- Role of money aggregates:
  - Continue reviewing money aggregates but avoid assigning significant policy weight to them given their limited empirically demonstrated information content.
- Exchange rate policy:
  - Maintain attention to exchange rate stability given its strong and rapid pass-through to domestic prices.
- Anchor inflation expectations:
  - Enhance monetary policy credibility to reduce inflation inertia and limit the effect of temporary deviations of actual inflation on public expectations.

### Further research and data priorities
- Modeling:
  - Progress in modeling inflation dynamics in Ghana is needed; this area is described as being at its infancy.
- Data quality:
  - Strengthen data quality, particularly for real sector activities where relevant data are missing (e.g., wages and employment) or available at low frequencies (e.g., annual).
  - Publication milestone: the Ghana Statistical Services began publishing quarterly national accounts in May 2011.

### Appendix I — Model of long-run money demand: selected estimates and tests
- Estimated cointegration equation parameters (selected):
  - M2_PLUS_REAL_LOG(-1) -1.00
  - RGDP_LOG(-1) 1.58
  - STDEV 0.13
  - T-STAT 11.99
  - D_NEER(-1) -3.98
  - STDEV 1.29
  - T-STAT -3.09
  - DEP_LOG(-1) -2.91
  - STDEV 2.55
  - T-STAT -1.14
  - CONSTANT -4.00
  - ALPHA -0.11
  - STDEV -0.03
  - T-STAT -3.36
- Unrestricted Cointegration Rank Test (Trace):
  - Hypothesized None: Eigenvalue 0.3655, Trace Statistic 5.0647, Critical Value 7.860, Prob. 0.01
  - Hypothesized At most 1: Eigenvalue 0.192, Trace Statistic 1.932, Critical Value 29.800, Prob. 0.30
  - Hypothesized At most 2: Eigenvalue 0.086, Trace Statistic 6.371, Critical Value 15.490, Prob. 0.65
  - Hypothesized At most 3: Eigenvalue 0.000, Trace Statistic 0.213, Critical Value 3.840, Prob. 0.65
  - Trace test indicates 1 cointegrating eqn(s) at the 0.05 level.
- Unrestricted Cointegration Rank Test (Maximum Eigenvalue):
  - Hypothesized None: Eigenvalue 0.363, Max-Eigen Statistic 3.1227, Critical Value 7.580, Prob. 0.01
  - Hypothesized At most 1: Eigenvalue 0.191, Max-Eigen Statistic 5.5721, Critical Value 21.130, Prob. 0.25
  - Hypothesized At most 2: Eigenvalue 0.086, Max-Eigen Statistic 6.1614, Critical Value 14.260, Prob. 0.59
  - Hypothesized At most 3: Eigenvalue 0.000, Max-Eigen Statistic 0.213, Critical Value 3.840, Prob. 0.65
  - Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level.
- Note on p-values:
  - **MacKinnon-Haug-Michelis (1999) p-values referenced for tests.

### Appendix II — Data definitions (as used in estimates)
- M2+ : Broad money, including foreign currency deposits (in millions of cedis)
- P : Consumer price index (100 = 2000)
- RGDP : Real GDP (in millions of cedis)
- NEER : Nominal effective exchange rate index (100 = 2000)
- DEP : Average deposit interest rate offered by banks (annual percent)
- Sources: Ghanaian authorities and author’s estimates.

*Source: _wp11274 - Section 3*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11274.pdf_