## Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea

## Source details

**Canonical URL:** [Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea](https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023084.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2023/english/wpiea2023084.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2023/english/wpiea2023084.pdf.json)

---

### Introduction — context, scope, and key findings
- Context:
  - Inflation in low-income countries (LICs) is often high and volatile and driven by external developments; fragile states face additional domestic instability, weak fiscal positions, and a shallow financial system.
  - Weak institutional frameworks, imperfect banking competition, and resistance to exchange rate flexibility hamper traditional monetary transmission (Mishra, Montiel, and Spilimbergo 2012).
- Guinea background:
  - Low-income fragile state in Sub-Saharan Africa (SSA) with resource endowments including more than a quarter of the world’s reserves of bauxite, large untapped high-grade iron ore deposits, and significant gold reserves.
  - Recent shocks: Ebola and commodity-price shock during 2014-15; recurrent social unrest; military coup in September 2021.
  - Real GDP growth averaged 6 percent over 2012-21.
- Monetary framework:
  - Guinean franc used since 1959; not currently pegged nor in a currency union.
  - Central Bank of Guinea (BCRG) autonomous, operational target: growth rate of reserve money; generally has not actively managed liquidity and has lent extensively to the government, highly correlated with reserve money evolution.
- Empirical approach:
  - Focus period since 2000; use VARX with sign restrictions and local projections to assess exchange rate pass-through.
- Summary empirical findings:
  - External factors (petrol prices, exchange rate fluctuations, global shipping costs, commodity prices) play a substantial role in inflation dynamics.
  - Exchange rate pass-through is relatively high: complete to import prices over one year and increasing up to two years, suggesting second-round effects on non-imported goods.
  - Monetary policy is effective and buffers external shocks mainly via the exchange rate channel; contractionary policy helped contain inflation in 2021–22.
- Relation to prior literature:
  - Findings align with studies for LICs and country-specific studies for Mali, Chad, and earlier work for Guinea.

*IMF Working Paper — Introduction (excerpt)*

### Inflation and monetary policy in Guinea — stylized facts and operational context
- CPI and composition:
  - CPI for Conakry compiled monthly; CPI basket includes 534 goods.
  - Food products: "over one third" of the CPI basket; fresh produce accounted for 20 percent of the basket.
  - Share of food in CPI: about 38 percent (published table vintages 1987–2002 show Food 41.9 / 37.6 and Fresh produce 20.0).
  - Imported products share: 41.2 / 40.6 percent (local products 58.9 / 59.5 percent in the table).
- Inflation developments:
  - Inflation averaged more than 12 percent over the last two decades; among the highest in SSA.
  - Record episode in 2006 when reserve money growth reached 84 percent at end-2006.
  - Inflation peaked during the 2009–10 military regime, increasing to 23 percent in August 2011.
  - Declined to 7.3 percent at end-2015 after 2010 reforms despite Ebola shocks.
  - Over the last decade inflation remained around 10 percent and was in the top decile of SSA.
  - Rapid food price growth a key driver despite food representing about 38 percent of the CPI basket.
- Persistence:
  - Country-by-country AR(1) Δpt = c + ρ Δpt−1 + β Δet + εt for 48 SSA economies over 2005–21: Guinea’s ρ̂ = 0.8 (half-life slightly above three months), well above SSA average of 0.2.
- Monetary policy framework and issues:
  - BCRG mandate: primary price stability; secondary: "supporting the general economic policies of the Government..." BCRG interpreted price stability as inflation below 10 percent; longer-term aim 5 percent.
  - Policy rate used for signaling; adjusted only five times in past decade; cut by 100 basis points in April 2020 after being unchanged since 2016.
  - Toolkit: repos, sterilization bills, collateralized intraday refinancing facility.
  - Transmission weaknesses: weak pass-through of policy rate to market rates; liquidity largely determined by autonomous factors; shallow government debt market; BCRG refrains from active liquidity management.
  - 2017 recapitalization: US$300 million via non-marketable treasury securities not usable for open market operations; interest paid via additional non-marketable instruments.
  - Monetary financing: central bank advances were GNF 5,306 billion (3.6 percent of GDP) at end-August 2020 (more than tripling legal limit), declining to GNF 3,180 billion by end-2020; limit defined as 5 percent of average fiscal revenues over past three years.
  - Under IMF programs (2012–2020) quantitative performance criteria guided base money growth; average broad money growth 17 percent and base money growth 13 percent over 2012+ period.
- Banking sector:
  - Reserve requirement reduced from 22 percent in 2012 to 15 percent in 2022; requirement unremunerated.
  - Lending-deposit rate spread oscillates between 15 and 20 percent.
  - Private sector credit to GDP below 10 percent; central bank funding of government accounts for 40 percent of total credit; commercial bank funding of government >35 percent of banks’ total lending.
  - Dollarization: deposits in foreign currency ≈ 30 percent of total liquidity.
- Exchange rate policy and reforms:
  - Historical crawl-like arrangement with base money anchor.
  - Since October 2018 de facto gradual depreciation via crawl-like arrangement and FX auctions; FX operations generally not sterilized and affect domestic liquidity.
  - November 2020 rule-based FX intervention: no auction to purchase (sell) FX unless exchange rate moves >0.25 percent vs its 5-day moving average; intervention amount capped at US$8 million; rule public and replicable.
  - Prior to rule, trigger would have been activated in about 2 percent of trading days vs actual FX sales in 20 percent.
  - Since November 2020: monthly FX sales substantially reduced; greater exchange rate flexibility; daily exchange rate distribution more disperse; BCRG stopped participating on both sides in single session.

*Source: wpiea2023084 - 2. Inflation and monetary policy in Guinea*

### Theoretical determinants of inflation — model and implications
- Model setup:
  - Stylized small open economy with tradable and non-tradable sectors following Durevall, Loening, and Birru (2013) and Kinda (2013).
  - Aggregate log price level: p_t = α p_t^N + (1−α) p_t^T with 0<α<1.
- Tradable sector (external market equilibrium):
  - p_t^T = (1−β + βγ) (e_t + tc_t + p_t^*).
  - Variables: e_t = exchange rate; tc_t = transport costs; p_t^* = foreign price level; β = fraction of tradeables invoiced with limited pass-through; γ = pass-through measure.
  - Purchasing power parity need not hold; tradable prices depend on e_t, tc_t, p_t^*, and limited pass-through factors.
- Non-tradable sector:
  - p_t^N = φ [ m_t^s − m_t^d ].
  - Real money demand: m_t^d = η y_t − μ E(π_t) with adaptive expectations E(π_t) = Δp_{t−1}, hence m_t^d = η y_t − μ Δp_{t−1}.
- Combined log-linear price level:
  - p_t = α φ [ m_t^s − η y_t + μ Δp_{t−1} ] + (1−α) (1−β + βγ) ( e_t + tc_t + p_t^* ).
  - Compact form: p_t = f( m_t^s, y_t, Δp_{t−1}, e_t, tc_t, p_t^* ).
- Key channels and policy-relevant implications:
  - Monetary channel: excess money supply raises p_t via non-tradable prices.
  - External channel: e_t depreciation, higher p_t^*, or tc_t raise tradable prices and p_t.
  - Expectation channel: past inflation (Δp_{t−1}) raises current inflation via money demand.
  - Real income channel: higher y_t raises money demand and can lower p_t if money supply unchanged.
  - Policy levers: targeting m_t^s can stabilize p_t but effectiveness depends on link between money supply and government financing; exchange-rate management and reducing transport costs mitigate external pressures; managing inflation expectations crucial because of μ Δp_{t−1} channel.

*IMF Working Paper — Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea (section 3).*

### Empirical estimates — VARX, decompositions, and pass-through
- VARX specification:
  - A·Y(L) = C·X(L) + ε with endogenous Y: base money (BM), broad money (M2, 푚푡푠), NEER (푒푡), petrol price at the pump (P@P), real output (푦푡), and CPI inflation (푝푡).
  - Exogenous X: Baltic Dry index (푡푐푡) and global non-fuel commodity price index ((1−훽)푝푡∗).
  - Quarterly data 2004Q1–2021Q4; variables measured by annual log difference.
- Identification via sign restrictions:
  - Monetary policy (MP) shock: contemporaneous increase in BM, M2, NEER depreciation, and CPI increase.
  - Demand shock: contemporaneous increase in output, M2, inflation, and NEER appreciation.
  - Petrol price (regulated) shock: increase in domestic petrol price and inflation, fall in M2; conditioned on petrol price rising in foreign currency terms and not driven by NEER depreciation.
  - FX shock: GNF fall, M2 drop, inflation increase; constrained so domestic petrol price change does not exceed exchange rate depreciation.
- Impulse-response insights:
  - Output growth largely unresponsive except to petrol price which causes a lagged contraction.
  - Monetary shock produces larger M2 change and stronger NEER response than petrol shock → sharper, shorter-lived inflation increase.
  - Quantitative example: one standard deviation shock to base money shifts average inflation by about 1 percentage point one year after the shock, largely driven by average NEER change of 3 percent given roughly 40 percent import share in CPI.
- Variance decomposition (short-term):
  - Identified shocks explain ~80 percent of variation in broad money and inflation, and ~60 percent of NEER.
- Approximate drivers (shares reported):
  - Broad money (M2): monetary shock ~30 percent; demand shocks ~15 percent; non-fuel commodity prices ~15 percent; petrol price shocks ~10 percent; transport costs negligible; exchange rate shocks ~5 percent.
  - NEER: monetary policy shocks ~30 percent; domestic demand ~20 percent; exchange rate shocks ~10 percent; residual large; petrol price and transport costs limited.
  - Inflation (short horizon): non-fuel commodity prices ~20 percent; transport costs almost 10 percent; petrol price shocks 15–20 percent depending on horizon; exchange rate shocks ~5 percent; domestic demand shocks 15–20 percent; monetary policy shocks 15–20 percent. Residual grows over longer horizons.
- Historical decompositions:
  - 2005: loose monetary policy contributed to strong depreciation and high inflation.
  - 2012–16 disinflation largely due to decline in non-fuel commodity prices; domestic petrol adjustments pushed inflation up by 1 to 2 percentage points.
  - 2021–22: contractionary monetary policy reduced broad money growth and helped contain international price pressures via NEER appreciation despite transport costs and non-fuel commodity prices rising to historical highs.
  - In 2005 and 2021–22 episodes, monetary shock contributed about a five-percentage point deviation of inflation from deterministic path.
  - A 20 percent petrol price increase in June 2022 should put upward pressure on inflation in 2023.
- Exchange-rate pass-through (local projections, monthly Sep 2010–Dec 2021):
  - Dependent variable: cumulative change in CPI over horizons h = {0,...,24} months; regressors include monthly NEER changes (∆e_t), output gap, and lags of price changes.
  - Pass-through estimates:
    - Overall CPI: a one-percent NEER depreciation → ~0.3 percent CPI increase after six months, 0.5 percent after one year, 0.8 percent after two years.
    - Cumulative implication: one-percent depreciation associated with inflation roughly 0.4 percent higher over subsequent two years.
    - Food CPI: one-percent depreciation → 0.9 percent rise in food prices after one year.
  - Comparative note: 12-month pass-through coefficient 0.5 > 0.3 estimate for 30 emerging economies in Carrière-Swallow and others (2021); 0.5 slightly exceeds Guinea’s imported goods share 0.4, indicating complete pass-through to imported goods and second-round effects to domestic goods over two years.
  - Precision limits prevent rejecting equality of pass-through between food and non-food.
  - Policy implication: strong NEER incidence on domestic prices limits real exchange rate adjustment and weakens expenditure switching; significant foreign currency debt may further limit exchange-rate flexibility effectiveness.

*Italic: IMF Working Papers — Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea (section 4. Empirical estimates).*

### Conclusion — interpretation and policy implications
- Challenges:
  - Central bankers in LICs and fragile states face severe challenges to deliver price stability; traditional transmission hampered and fiscal pressure strong.
  - Many LICs choose pegs to import inflation stability at the cost of losing exchange-rate shock absorption.
- Main findings for Guinea:
  - High and persistent inflation over past two decades; largest autoregressive coefficient among SSA countries.
  - External factors (non-fuel commodity prices, transport costs, oil prices) explain most inflation dynamics.
  - Robust relationship between monetary aggregates, the exchange rate, and inflation — monetary policy, when resolute, effective in Guinea.
- Transmission and pass-through:
  - Monetary policy transmits largely via exchange rate channel; money growth drives exchange rate movements and inflation.
  - Strong exchange-rate pass-through, stronger for food, with evidence of second-round effects over two years.
  - In 2021–22, contractionary monetary policy and exchange rate appreciation nearly offset international price pressures that would have otherwise added 6 percentage points to an already high inflation rate.
- Central bank performance and reforms:
  - Recent reforms modernizing policy framework likely contributed to exchange rate outcomes; BCRG implemented simple transparent FX intervention rule and allowed greater flexibility.
  - Despite improvements, BCRG has struggled to meet its inflation objective over past five years.
  - Strictly implemented tight monetary policy can lower near-term inflation; strengthening operational and financial independence would reduce fiscal dominance risk.
- Medium-term strategy:
  - BCRG may transition from base money growth framework to an inflation targeting framework over the medium-term.
  - Given dependence on commodity exports and potential diversification, moving towards inflation targeting could contribute to price and macroeconomic stability.
  - Empirical evidence indicates regimes with more flexible exchange rates and strong nominal anchors have delivered lower and less persistent inflation without increasing inflation volatility in SSA.
  - Recommended next steps: gradually increase exchange rate flexibility to act as shock absorber while maintaining close control over monetary conditions through active operations.

*IMF Working Paper — Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea*

### Introduction ...........................................................................................................

### Introduction

### Context and motivation
- Inflation in low-income countries (LICs) is often high and volatile, driven by external developments.
- Inflation in fragile states is affected by domestic instability, weak fiscal positions, and a shallow financial system.
- These factors, together with often weak institutional frameworks (e.g., unclear nominal anchors or lack of central bank independence), imperfect competition in the banking sector and resistance to exchange rate flexibility hamper the traditional monetary transmission through interest rate, bank lending, asset price, and exchange rate channels (Mishra, Montiel, and Spilimbergo 2012).
- This makes it challenging for monetary policy to deliver price stability.

### Guinea: background and recent macroeconomic performance
- Guinea is a low-income and fragile state in Sub-Saharan Africa (SSA).
- Resource endowments: more than a quarter of the world’s reserves of bauxite, one of the largest untapped high-grade iron ore deposits, and significant gold reserves.
- Socio-political events:
  - Ebola health crisis and a commodity-price shock during 2014-15.
  - Recurrent social unrest, protests, and strikes in the subsequent period.
  - Military coup in September 2021.
- Real GDP growth averaged 6 percent over 2012-21, supported by booming mining activities and substantial infrastructure investment.

### Currency and monetary policy framework
- Guinea has issued its own domestic currency since 1959; the Guinean franc is not currently pegged to a foreign currency and does not belong to a currency union.
  - Footnote: Between 1974 and 1985, the Guinean franc was replaced by the syli. In the past, the Guinean franc has at times been pegged to the CFA franc and the US dollar.
- The Central Bank of Guinea (BCRG) has an autonomous monetary policy pursuing price stability with a moderately flexible exchange rate.
- Operational target: growth rate of reserve money.
- Empirical observation: the central bank has generally not actively managed liquidity conditions over the period surveyed, and its extensive lending to the government has been highly correlated with the evolution of reserve money.

### Scope and empirical approach of the paper
- Focus period: since 2000.
- Stylized facts documented: inflation in Guinea was among the highest and most persistent in SSA; food prices dominated overall CPI inflation.
- Empirical strategy:
  - Use a vector autoregressive model with exogenous variables (VARX) to analyze the impact of changes to monetary aggregates, exchange rates, domestic demand, and international prices on inflation.
  - Impose sign restrictions to identify the main structural shocks driving inflation and estimate their contributions to observed inflation dynamics.
  - Provide a quantitative assessment of exchange rate pass-through to domestic prices using a local projections approach.

### Key empirical findings (summary)
- External factors play a substantial role in determining inflation dynamics in Guinea, including:
  - Petrol prices
  - Exchange rate fluctuations
  - Global shipping costs
  - Commodity prices
- Exchange rate pass-through in Guinea is relatively high by international standards:
  - Pass-through to import prices is estimated to be complete over a one-year horizon and to increase further up to a two-year horizon.
  - This suggests the presence of second-round effects on non-imported goods.
- Monetary policy effectiveness:
  - Evidence that monetary policy in Guinea is effective and has played an important role in buffering the impact of external shocks on domestic inflation.
  - The exchange rate is central in monetary policy transmission: monetary shocks affect inflation significantly through their impact on the exchange rate.
  - Historical episodes:
    - Monetary accommodation of large fiscal expansions in the mid-2000s and before the military coup in 2008 were associated with visible exchange-rate/monetary interactions.
    - Contractionary monetary policy helped contain inflation during 2021–22 following another military coup, offsetting international price pressures related to the Covid-19 pandemic and the war in Ukraine.

### Relation to prior literature
- Findings align with previous studies of LIC economies (Ha and others 2019; Durevall, Loening, and Birru 2013).
- Country-specific studies:
  - Diouf (2007) for Mali: monetary and external factors matter for inflation in the short and long term.
  - Kinda (2013) for Chad: supply-side shocks, public spending, and external factors are main drivers.
  - Blavy (2004) for Guinea (1992-2003): strong long-run relationship between money supply growth and inflation; episodes of high money supply growth linked to civil unrest and structural change had significant impact on headline inflation.
- Recent broader findings:
  - Carare and others (2022) confirm the importance of the monetary regime in allowing LICs to adjust to external shocks.
  - Abuka and others (2019) and Brandão-Marques and others (2021) provide evidence on transmission channels in specific LIC and emerging market contexts.

### Structure of the paper
- Section II: evolution of inflation in Guinea, the monetary policy framework, and the exchange rate regime, with comparisons to SSA peers.
- Section III: motivation of the empirical strategy for modeling inflation.
- Section IV: data, estimation methods, and results.
- Section V: conclusion.

*IMF Working Paper — Introduction (excerpt)*

### 2. Inflation and monetary policy in Guinea:

### 2. Inflation and monetary policy in Guinea

### Stylized facts — CPI and composition
- CPI for Conakry compiled monthly by the National Institute of Statistics; CPI basket includes 534 goods.
- Food products: "over one third" of the CPI basket; fresh produce accounted for 20 percent of the basket.
- Share of food in CPI: about 38 percent (noting the basket vintages 1987–2002 show Food 41.9 / 37.6 and Fresh produce 20.0 in the published table).
- Imported products share: 41.2 / 40.6 percent (local products 58.9 / 59.5 percent in the table).

### A. Inflation developments — key findings
- Inflation has averaged more than 12 percent over the last two decades and has been among the highest in Sub-Saharan Africa.
- Record inflation episode in 2006 associated with reserve money growth reaching 84 percent at end-2006.
- Inflation peaked during the 2009–10 military regime, increasing to 23 percent in August 2011.
- Following 2010 democratic elections and reforms, inflation declined to 7.3 percent at end-2015 despite Ebola-related supply shocks.
- Over the last decade inflation remained around 10 percent and was in the top decile of SSA.
- Rapid food price growth has been an important driver of overall inflation, despite food representing about 38 percent of the CPI basket.
- Inflation persistence:
  - Estimated country-by-country AR(1) regression Δpt = c + ρ Δpt−1 + β Δet + εt for 48 SSA economies over 2005–21.
  - Guinea’s estimated AR(1) coefficient ρ̂ = 0.8, implying shocks to inflation dissipate very gradually (half-life of slightly above three months).
  - Guinea’s ρ̂ is broadly stable over rolling five-year windows and remains well above the SSA average of 0.2.

### B. Monetary policy framework — structure and operational issues
- BCRG primary mandate: ensure price stability; secondary objective: “supporting the general economic policies of the Government with a view to ensuring sound and sustainable economic growth”, without undermining price stability.
- BCRG interpreted price stability as maintaining inflation below 10 percent; longer-term aim is to bring inflation down to 5 percent.
- Monetary policy rate:
  - Used to signal policy stance and anchor inflation expectations.
  - Adjusted only five times in the past decade.
  - After remaining unchanged since 2016, policy rate was reduced by 100 basis points in April 2020.
- Operational toolkit includes repos, sterilization bills auctioned at variable rates, and a collateralized intraday refinancing facility.
- Transmission weaknesses and market structure:
  - Transmission of the policy rate to market rates is weak.
  - Liquidity conditions largely determined by autonomous factors (Treasury financing needs, external financing disbursement).
  - Market for government debt is shallow; no secondary market; first longer-dated government bond issued only in early 2022.
  - BCRG refrains from active liquidity management; autonomous liquidity changes are not regularly sterilized.
- 2017 recapitalization:
  - US$300 million recapitalization involved issuance of non-marketable treasury securities that cannot be used for open market operations.
  - Interest on recapitalization bonds has been paid via issuance of additional non-marketable instruments.
- Monetary financing and monetary aggregates:
  - Over last two decades, BCRG operated a base money targeting regime with a managed exchange rate.
  - Central bank lending to the government accounted for about three quarters of base money growth over this period.
  - Limit on central bank advances: defined as 5 percent of average fiscal revenues over the past three years.
  - In 2020, outstanding central bank advances surged to GNF 5,306 billion (3.6 percent of GDP) at end-August 2020, more than tripling the legal limit, before declining to GNF 3,180 billion by end-2020.
  - Under IMF programs (2012–2020), quantitative performance criteria for central bank net domestic assets and net foreign assets defined a path for base money growth to contain monetary financing.
  - Average broad money growth of 17 percent and base money growth of 13 percent over the 2012+ period—relatively low compared to Guinea’s prior record but high relative to SSA peers.
- Banking sector and credit conditions:
  - Reserve requirement reduced from 22 percent in 2012 to 15 percent in 2022; requirement is unremunerated.
  - High reserve requirement contributes to a high lending-deposit rate spread oscillating between 15 and 20 percent.
  - Private sector credit to GDP below 10 percent; outstanding private sector credit is shallow.
  - Central bank funding of the government accounts for 40 percent of total credit to the economy.
  - Commercial bank funding of the government accounts for more than 35 percent of banks’ total lending.
  - Credit to the economy dominated by lending to the official sector, crowding out private sector borrowing.
  - Dollarization measured by deposits in foreign currency as a share of total liquidity is estimated at 30 percent.

### C. Exchange rate policy — regime and recent reforms
- Historical regime: crawl-like arrangement with occasional large adjustments and a base money anchor.
- Empirical cross-country evidence (SSA):
  - Pegs have the lowest inflation rates and persistence.
  - Intermediate regimes—including Guinea’s—are associated with the highest average inflation (almost 10 percentage points higher than pegs and about 7 percentage points higher than floats) and higher inflation persistence.
- Operational FX policy:
  - Since October 2018, BCRG allowed a gradual depreciation via a de facto crawl-like arrangement; FX auctions used to buy/sell foreign exchange from domestic banks.
  - FX operations generally not sterilized and thus affect domestic liquidity.
- November 2020 rule-based FX intervention:
  - Central bank does not hold an auction to purchase (sell) FX unless the exchange rate has appreciated (depreciated) by more than 0.25 percent with respect to its 5-day moving average.
  - Amount of intervention not to exceed US$8 million.
  - The rule is public and replicable by market participants.
  - Trigger statistics: in the two years prior to rule adoption the trigger would have been activated in about 2 percent of trading days, versus actual FX sales in 20 percent of trading days.
- Effects of the rule:
  - Since November 2020 the rule-based policy has been consistently applied.
  - Monthly FX sales by the BCRG have been substantially reduced.
  - Exchange rate has been allowed greater flexibility; distribution of daily exchange rate movements became substantially more disperse.
  - BCRG stopped participating on both sides of the market in a single session (no longer both buyer and seller in same session).

*Source: wpiea2023084 - 2. Inflation and monetary policy in Guinea*

### 3. Theoretical determinants of inflation

### 3. Theoretical determinants of inflation

### Model setup and scope
- Framework: stylized small open economy model with two sectors — tradable and non-tradable — following Durevall, Loening, and Birru (2013) and Kinda (2013).
- Aggregate log price level: p_t = α p_t^N + (1−α) p_t^T, where variables are in natural logarithms and 0<α<1 is the share of nontradable goods in total expenditures.
- Objective: motivate the empirical specification by expressing the log price level as a function of monetary variables, external sector variables, and past inflation.

### Tradable sector and pass-through (external market equilibrium)
- Tradable price equation (external market equilibrium):
  - p_t^T = (1−β + βγ) (e_t + tc_t + p_t^*)
  - Components and interpretation:
    - e_t: exchange rate (in domestic currency).
    - tc_t: transport costs.
    - p_t^*: foreign price level (price of tradable goods in foreign currency determined exogenously).
    - β: fraction of tradeables invoiced with limited pass-through due to local currency pricing or regulated domestic prices.
    - γ: pass-through measure reflecting extent of domestic price controls and local currency pricing.
  - Implication: purchasing power parity does not hold; domestic tradable prices depend on exchange rate, transport costs, foreign prices, and limited pass-through factors.

### Non-tradable sector and money-market disequilibrium
- Non-tradable price determined by money-market disequilibrium:
  - p_t^N = φ [ m_t^s − m_t^d ]
  - m_t^s: logarithm of broad money (money supply).
  - m_t^d: real money demand.
  - φ: parameter mapping money-market disequilibrium into nontradable prices.
- Real money demand specification (log-linearized):
  - m_t^d = η y_t − μ E(π_t)
  - Assumption of adaptive expectations: E(π_t) = Δp_{t−1}, leading to
  - m_t^d = η y_t − μ Δp_{t−1}
  - Interpretation:
    - η y_t: real income effect — an increase in real income raises money demand.
    - −μ E(π_t): expected inflation raises the opportunity cost of holding money and lowers real money demand.
  - Monetary policy: m_t^s is the operational target for the central bank, but in practice is closely related to government financing needs.

### Log-linear price level as a function of fundamentals
- Combined price-level expression after substituting tradable and non-tradable components:
  - p_t = α φ [ m_t^s − η y_t + μ Δp_{t−1} ] + (1−α) (1−β + βγ) ( e_t + tc_t + p_t^* )
- Compact functional form:
  - p_t = f( m_t^s, y_t, Δp_{t−1}, e_t, tc_t, p_t^* )
- Key channels captured:
  - Monetary channel: higher m_t^s relative to m_t^d raises p_t via p_t^N.
  - External channel: exchange rate depreciation (e_t), higher foreign prices (p_t^*), or higher transport costs (tc_t) raise tradable prices and hence p_t.
  - Expectation channel: higher past inflation (Δp_{t−1}) raises expected inflation, lowers money demand, increases excess money supply, and raises current prices.
  - Real income channel: higher y_t increases money demand; if money supply is unchanged, this tends to lower p_t through reduced non-tradable prices (domestic supply shock).

### Comparative and policy-relevant implications
- Inflation sources identified:
  - Monetary-origin inflation: increases in money supply in excess of real money demand.
  - External-origin inflation: exchange rate depreciation, higher foreign prices, higher transport costs, and limited pass-through dynamics (β, γ).
  - Expectation-driven persistence: past inflation transmits to current inflation via expected inflation and money demand responses.
  - Real activity effects: higher real output can exert downward pressure on prices by raising money demand.
- Policy levers and considerations:
  - Monetary policy can target m_t^s to stabilize p_t, but effectiveness depends on the link between money supply and government financing needs.
  - Exchange-rate management and addressing transport costs can mitigate external price pressure transmission.
  - Managing inflation expectations is crucial due to the μ Δp_{t−1} channel that amplifies price persistence.

*IMF Working Paper — Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea (section 3).*

### 4. Empirical estimates

### 4. Empirical estimates

### A. Vector autoregression (VARX) model
- Model specification and variables:
  - VARX described by A·Y(L) = C·X(L) + ε (equation 9).
  - Endogenous vector Y includes: base money (BM), broad money (M2, 푚푡푠), nominal effective exchange rate (NEER, 푒푡), petrol price at the pump (P@P, 훽훾(푒푡+푡푐푡+푝푡∗)), real output (푦푡), and CPI inflation (푝푡).
  - Exogenous vector X contains: Baltic Dry index (푡푐푡) and global non-fuel commodity price index ((1−훽)푝푡∗).
- Data and frequency:
  - VAR estimated using quarterly data over the period 2004Q1 to 2021Q4.
  - Base money and broad money measured in GNF millions (BCRG); CPI and CPI components from INS; NEER from IMF Information Notice System; real GDP in constant GNF (higher-frequency proxies from 2010 onwards; cubic interpolation before 2010). Baltic Dry index from Haver Analytics; global non-fuel commodity prices from IMF database.
  - All variables measured by the annual log difference.
- Identification (sign restrictions, Table 2):
  - Expansionary monetary policy (MP) shock: contemporaneous increase in BM, M2, depreciation of NEER, and increase in CPI.
  - Demand shock: contemporaneous increase in output, M2, inflation, and appreciation of NEER.
  - Regulated petrol price (P@P) shock: increase in domestic petrol price, increase in inflation, fall in M2; conditioned on petrol price rising in foreign currency terms and constrained to not be driven by NEER depreciation.
  - Foreign exchange (FX) shock: contemporaneous fall in value of GNF, drop in M2, increase in inflation; constrained so change in domestic petrol price does not exceed depreciation of exchange rate.
  - Exogenous shocks: non-fuel commodity prices and global transport prices.
- Key impulse response insights:
  - Output growth does not react significantly to most shocks except petrol price, which provokes a lagged contraction.
  - Demand shock: output expansion → NEER appreciation → inflation quickly reverts.
  - Monetary policy shock: similar base money growth change as petrol shock but larger change in M2 and stronger NEER response → sharper, shorter-lived inflation increase.
  - Quantitative response: a one standard deviation shock to base money shifts the average inflation rate by about 1 percentage point one year after the shock. This is largely driven by the average change in the NEER of 3 percent over that horizon given a roughly 40 percent import share in the CPI basket.

### B. Decomposing inflation dynamics
- Variance decomposition (short-term horizon):
  - Identified shocks explain about 80 percent of the variation in broad money and inflation, and 60 percent of the NEER.
- Broad money (M2) drivers (approximate shares):
  - Monetary shock: about 30 percent.
  - Demand shocks: about 15 percent.
  - Non-fuel commodity price shocks: about 15 percent.
  - Petrol price shocks: about 10 percent.
  - Transport costs: negligible role.
  - Exchange rate shocks: about 5 percent.
- NEER drivers (approximate shares):
  - Monetary policy shocks: about 30 percent.
  - Domestic demand shocks: about 20 percent.
  - Exchange rate shocks: about 10 percent.
  - Residual: large (external non-identified factors, e.g., global risk aversion).
  - Petrol price and transport costs: limited role.
- Inflation drivers (short horizon, approximate shares):
  - Non-fuel commodity prices: 20 percent.
  - Transport costs: almost 10 percent.
  - Petrol price shocks (domestically regulated): 15 to 20 percent depending on horizon.
  - Exchange rate shocks: about 5 percent.
  - Domestic demand shocks: 15 to 20 percent depending on horizon.
  - Monetary policy shocks: 15 to 20 percent depending on horizon.
  - Over longer horizons: residual grows, external unaccounted factors become more important.
- Historical decomposition and episodes:
  - 2005 episode: loose monetary policy contributed to strong depreciation and high inflation.
  - 2021–22 episode: contractionary monetary policy reduced broad money growth and helped contain international price pressures through NEER appreciation despite transport costs and non-fuel commodity prices rising to historical highs.
  - In both episodes, the monetary shock contributed about a five-percentage point deviation of inflation from its deterministic path.
  - 2012–16 disinflation largely explained by an external factor: decline in non-fuel commodity prices (inflation slowed from more than 20 to below 10 percent). Domestic petrol price adjustments in that period pushed inflation up by 1 to 2 percentage points.
  - The 20 percent increase in the petrol price in June 2022 should put upward pressure on inflation in 2023.
- Interpretation:
  - External factors play a large role in Guinea’s inflation dynamics, but there is a robust relationship among monetary aggregates, the exchange rate, and inflation, indicating an important role for monetary policy in steering inflation through its effect on M2 and NEER.

### C. Exchange-rate pass through using local projections
- Purpose and specification:
  - Estimate extent to which exchange rate changes pass-through to domestic inflation irrespective of shock origin.
  - Local projection specification (Jordà, 2005) adapted from Carrière-Swallow and others (2021) (equation 10).
  - Dependent variable: cumulative change in CPI over horizons h = {0,...,24} months.
  - Regressors include monthly changes in NEER (∆푒푡), output gap (푔푎푝푡), and lags of price changes.
- Data:
  - Monthly frequency from September 2010 to December 2021.
  - CPI food and CPI non-food components from INS included for sub-sample analysis.
- Estimated pass-through results (impulse-response point estimates):
  - Overall CPI: a one-percent depreciation of the NEER is followed by:
    - about 0.3 percent increase in CPI after six months,
    - 0.5 percent after one year,
    - 0.8 percent after two years.
  - Cumulative implication: a one-percent depreciation is associated with a rate of inflation roughly 0.4 percent higher over the subsequent two years.
  - Food CPI: a one-percent depreciation is followed by a 0.9 percent rise in food prices after one year.
- Comparative and interpretive points:
  - Estimated 12-month pass-through coefficient of 0.5 is higher than the coefficient of about 0.3 in Carrière-Swallow and others (2021) for 30 emerging economies over 2000-19.
  - The 0.5 estimate is slightly higher than Guinea’s imported goods share in the CPI basket (0.4), suggesting complete pass-through to domestic price of imported goods and some second-round effects to domestically produced goods over the second year.
  - Pass-through to food appears stronger than to other goods; under assumption that all food is tradable, this would be consistent with near-complete pass-through after one year.
  - Low precision of estimates prevents rejecting the null that pass-through to food and non-food inflation are the same.
- Policy implication:
  - Nominal NEER changes have a strong incidence on domestic prices, leading to limited real exchange rate adjustment because of strong domestic price responses — this can weaken the expenditure switching channel that would otherwise help domestic demand and output respond to adverse external shocks.
  - Presence of significant foreign currency debt in Guinea may further limit the effectiveness of exchange-rate flexibility for expenditure switching.

*Italic: IMF Working Papers — Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea (section 4. Empirical estimates).*

### 5. Conclusion

### 5. Conclusion

### Challenges for central bankers in LICs and fragile states
- Central bankers in low-income countries and fragile states face severe challenges in delivering price stability.
- Traditional transmission mechanisms of monetary policy are often hampered and pressure to succumb to fiscal objectives is strong.
- Compared to richer countries, inflation is often high, persistent, volatile, and driven by external developments.
- Many LICs and fragile states opt to “tie their hands” by pegging their currency and importing inflation rates from the anchor nation, at the cost of foregoing a potentially powerful policy tool to shield the domestic economy from external shocks.

### Main findings for Guinea
- Guinea is a low-income country that has recently suffered repeated shocks—including health related events and military coups—in the past two decades.
- Inflation in Guinea over the past two decades has been high and remarkably persistent, with the largest autoregressive coefficient among Sub-Saharan African countries.
- External factors such as international non-fuel commodity prices, transport costs, and oil prices explain most inflation dynamics.
- Analysis reveals a robust relationship between monetary aggregates, the exchange rate, and inflation dynamics, suggesting that monetary policy—when implemented resolutely—has been effective in Guinea.

### Transmission channels and pass-through
- Monetary policy transmits largely via the exchange rate channel.
- Inflation is closely associated with money growth, which in turn drives exchange rate movements.
- Estimates point to strong pass-through of exchange rate movements to domestic prices—stronger for food than for other items—with some evidence of second-round effects on domestic goods over the two-year horizon.
- The strong transmission via the exchange rate was evident in 2021–22, when contractionary monetary policy helped appreciate the exchange rate.
- This exchange rate appreciation almost offset the international price pressures from global supply shocks in the post-Covid environment, which the paper estimates would have otherwise added 6 percentage points to an already high inflation rate.

### Central bank performance and operational context
- Recent reforms to modernize the central bank’s policy framework have likely contributed to the exchange rate outcome.
- The exchange rate has become more flexible, with the BCRG implementing a simple and transparent FX intervention rule.
- Despite broadly good performance amidst shocks, the BCRG has struggled in meeting its inflation objective over the past five years.
- Analysis shows that if implemented strictly, tight monetary policy can contribute to attaining lower inflation in the near-term.
- Continuing to strengthen the operational and financial independence of the central bank would reduce the scope for fiscal dominance and strengthen the BCRG’s ability to steer monetary conditions in pursuit of its price stability mandate.

### Policy implications and medium-term strategy
- Over the medium-term the BCRG may choose to transition from the current base money growth framework to an inflation targeting framework.
- Given Guinea’s dependence on commodity exports and potential for economic diversification, moving towards an inflation targeting regime could contribute to price stability and macroeconomic stability.
- Empirical evidence cited indicates that regimes with more flexible exchange rates and strong nominal anchors have delivered lower and less persistent inflation without causing increased inflation volatility in SSA countries.
- A next step should involve gradually making the exchange rate more flexible and allowing it to act increasingly as a shock absorber, while maintaining close control over monetary conditions through active operations.

*Source: IMF Working Paper — Inflation and Monetary Policy in a Low-Income and Fragile State: The Case of Guinea, 5. Conclusion*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023084.pdf_