## wpiea2025121-print-pdf

## Source details

**Canonical URL:** [wpiea2025121-print-pdf](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025121-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025121-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025121-print-pdf.pdf.json)

---

### Introduction — context and objectives
- Tonga is a small open, import-dependent economy with a trade deficit averaging about 50 percent of GDP in the past decade.
- Remittances are around 40 percent of GDP and, together with budget support, fund the large trade deficit.
- Excess liquidity in the banking system exceeds 30 percent of GDP.
- NRBT primary objectives:
  - maintain internal price stability by keeping inflation below a reference rate (currently of 5 percent), and
  - maintain an adequate level of foreign reserves of a minimum of 3 months (and optimally around 7.5 months) of imports.
- The Tongan Pa’anga is pegged to a basket of currencies; NRBT can adjust exchange rates by 5 percent each month if necessary.
- NRBT policy rate has been constant at zero percent since 2012; NRBT has relied more on the exchange rate channel due to weak interest rate transmission.
- To protect the peg and foreign reserves, NRBT employs capital controls; the Foreign Exchange Control Act of 2018 is cited.
- Foreign reserves rose to over 10 months of imports in 2020; the interbank market has been inactive since 2010. NRBT has not issued notes or conducted open-market operations since 2009.
- Recent shocks: a natural disaster every 2 years since 2014 (including tropical cyclones and the 2022 volcanic eruption and tsunami) and the COVID-19 pandemic; one domestic bank has shown pronounced deterioration motivating a bank-failure scenario analysis.
- Purpose: customize a Quarterly Projection Model (QPM) to Tonga to produce medium-term quarterly projections of output, inflation, interest rate, and exchange rate and evaluate three scenarios: (1) bank failure, (2) natural disaster, and (3) negative external shock.

### The QPM and its calibration to Tonga
- Model structure (canonical QPM):
  - Four main equations: aggregate demand, aggregate supply, uncovered interest parity (UIP), and a Taylor rule.
  - Aggregate demand links the output gap, monetary conditions, foreign output gap, and aggregate demand shocks. Monetary condition index (MCI) is a function of the interest rate gap and the exchange rate gap.
  - Aggregate supply links inflation to lagged inflation, expected inflation, real marginal cost, and supply shocks; real marginal cost depends on output gap and real exchange rate gap.
  - UIP links nominal exchange rate to expected future exchange rate, domestic and foreign nominal interest rates, a country risk premium, and an exchange rate shock.
  - Taylor rule specifies interest rate dynamics as a function of lagged interest rates, a neutral interest rate, expected inflation deviation from target, the output gap, and policy shocks.
- Adaptations for Tonga:
  - Exchange rate acts as the nominal anchor; monetary policy conducted via exchange rate interventions rather than inflation targeting.
  - Monetary transmission is limited: small economy, high excess liquidity, high openness, low financial market development, heavy reliance on remittances.
  - Key model features: MCI gives more weight to the exchange rate gap than the interest rate gap; high persistence of core inflation; volatile passthrough for energy and food prices; UIP modified for managed float with p1 = 0.2 and h2 = 0.5; strong persistence in domestic interest rates and low responsiveness to output and inflation gaps relative to the 5 percent reference.
- Institutional/historical inputs included in calibration:
  - Excess liquidity >30 percent of GDP; NRBT policy rate constant at zero percent since 2012; capital controls (Foreign Exchange Control Act of 2018); foreign reserves >10 months in 2020; interbank market inactive since 2010; no NRBT notes issued since 2009.

- QPM selected calibrated parameter values (as specified):
  - Aggregate demand:
    - b1 (Degree of output persistence): 0.4
    - b2 (Monetary policy passthrough): 0.3
    - b3 (Impact of external demand on domestic output): 0.3
    - b4 (Weight of real interest rate and real exchange rate in MCI): 0.2
    - b5 (Persistence of the credit premium): 0.7
  - Aggregate supply:
    - a1 (Inflation persistence): 0.97
    - a2 (Policy passthrough — impact of real marginal cost on inflation): 0.05
    - a3 (Weight in real marginal cost equation, output gap vs exchange rate gap): 0.15
  - Uncovered Interest Parity (UIP):
    - p1 (Deviation from canonical UIP): 0.2
    - h2 (Dirty float/managed float/peg indicator): 0.5
  - Exchange rate intervention rule:
    - f1 (Contribution of past exchange rate target): 0.5
    - f2 (Contribution of deviation from inflation target): 0.25
    - f3 (Contribution of the output gap): 0.1
  - Taylor rule:
    - g1 (Interest rate persistence): 0.98
    - g2 (Contribution of deviation from inflation target): 0.1

- Steady states used in the model:
  - Domestic variables:
    - Inflation target: 5
    - Real interest rate: 3
    - Trend real exchange rate: -1
    - Potential output: 2
    - Relative price of food inflation to CPI: 0
    - Relative price excluding food and energy to CPI: 0
  - Foreign variables:
    - US inflation target: 2
    - US real interest rate: 0.75
    - Rest of the world food price: 0
    - Rest of the world oil price: 5

### Scenario analysis — overview and experimental design
- Three notional experiments relative to baseline:
  1. Bank failure (persistent credit premium shock)
  2. Natural disaster (shocks to potential output and inflation dynamics)
  3. Negative external shock (shocks to external conditions)
- Experiments are qualitative proxies to assess impulse response shapes rather than to mimic precise historical magnitudes.

### Scenario 1 — Bank failure: setup, shocks, mechanics, and outcomes
- Rationale: a large domestic bank with high non-performing loans and declining liquidity could fail, affecting credit conditions and country risk.
- Assumed aggregate impact: a 1.5 percent decrease in output relative to the baseline over the period 2025Q1 – 2026Q4.
- Shock specification to the Credit Premium, εt_CRPRP:
  - 1.5 percent (150 basis pts) in 2025Q1,
  - 0.5 percent in 2025Q2,
  - 0.25 percent in 2025Q3,
  - followed by a permanent increase of 0.75 percent to the Credit Premium from 2025Q4 onward (through end of simulation 2026Q4).
- Additional effect: increase in country risk (prrpmt) due to potential government-backed bank failure.
- Model mechanics:
  - MCI augmented by credit premium: MCI = b4 * (real interest rate gap + CreditPremium) + (1−b4)*(−real exchange rate gap). An increase in MCI is restrictive.
  - Credit premium evolution: CRPRPt = b5 * CRPRPt−1 + (1−b5) * (prrpmt − prrpmt−1) + εt_CRPRP + AAAAACRPRPt, with b5 = 0.7.
- Impulse response highlights (alternative minus baseline):
  - Initial change in the MCI gap of 0.3 percent in 2025Q1.
  - Larger MCI difference in 2025Q2, decline in 2025Q3 reflecting shock sequence, then increased MCI gap from 2025Q4 onward due to the permanent 0.75 percent credit premium increase.
  - Output gap becomes more negative; combined shocks amount to a permanent loss over 2025Q1–2026Q4 of 1.5 percent of output relative to baseline.
- Predicted real-economy and price effects relative to baseline:
  - Interest rates: mild reduction relative to baseline due to low policy passthrough and negative output gap.
  - Exchange rate: Pa’anga depreciates; imported energy prices in local currency initially more expensive.
  - Inflation: alternative scenario inflation decreases slightly relative to baseline (consistent with calibrated low passthrough from real marginal cost).
  - Core inflation: highly persistent as calibrated.

### Scenario 2 — Natural disaster: setup, shocks, mechanics, and outcomes
- Context: Tonga frequently experiences natural disasters that damage capital formation; aid inflows often mitigate output effects.
- Modeled as a transitory downward shift in potential output; estimated impact: 4.6 percent decrease in output relative to the baseline over 2025Q1 – 2026Q4.
- Specific shock assumptions:
  - A positive shock to the output gap, εt_y, of 5 percent in 2025Q1 due to a decline in potential output εt_y* of 5 percent.
  - Staggered annual inflation increases: 3 percent in 2025Q1, 1.5 percent in 2025Q2, 0.75 percent in 2025Q2 and 0.375 percent in 2025Q4. (As specified in the source.)
- Model responses and transmission:
  - Output gap: shock leads to an increase in the output gap (given the specification).
  - Monetary condition index: output gap negatively related to MCI; MCI depends on interest rate gap via the Taylor rule.
  - Interest rates: model predicts a mild reduction in interest rates relative to the baseline (but source also notes: "In effect, interest rates go up with respect to the baseline" reflecting a higher path at times, potentially due to default risk).
  - Exchange rate: Pa’anga appreciates slightly relative to baseline; imported energy prices in local currency initially less expensive.
  - Inflation: alternative scenario inflation increases slightly relative to baseline; core inflation remains persistent.
- Additional parameter and historical notes:
  - Domestic food comprises about 15 percent of the CPI basket.
  - Historical pattern: local food prices tend to rise by more than 10 percent in the first quarter after a natural disaster, followed by further price hikes along the supply cycle.

### Scenario 3 — Negative external shock: specification and outcomes
- Shock components and timing:
  - 1. One-time shock to world consumer price index: 1.5 percent in 2025Q1.
  - 2. Prolonged shock to the world output gap: −0.25 percentage points per quarter for four quarters commencing in 2025Q2.
  - 3. Shock to the exchange rate (TOP/USD): appreciation over 2025Q1–2025Q3 followed by depreciation commencing 2025Q4.
- Table of shock values (as specified):
  - 2025Q1: Global Inflation Shock 1.5; Global Output Gap Shock 0; Exchange Rate Shock (TOP/USD) −0.5
  - 2025Q2: Global Inflation Shock 0; Global Output Gap Shock −0.25; Exchange Rate Shock (TOP/USD) −0.2
  - 2025Q3: Global Inflation Shock 0; Global Output Gap Shock −0.25; Exchange Rate Shock (TOP/USD) −0.1
  - 2025Q4: Global Inflation Shock 0; Global Output Gap Shock −0.25; Exchange Rate Shock (TOP/USD) 0.3
  - 2026Q1: Global Inflation Shock 0; Global Output Gap Shock −0.25; Exchange Rate Shock (TOP/USD) 0.5
- Transmission and outcomes:
  - Domestic inflation:
    - Alternative path slightly higher than baseline at shock time; higher global inflation increases real marginal cost leading to higher food and energy inflation.
    - Non-core components (food and energy) make up 56.5 percent of the CPI basket; high pass-through from global inflation to domestic non-core inflation.
    - Pa’anga appreciation in first three periods reduces inflation relative to baseline; subsequent depreciation increases inflation relative to baseline.
    - The one-time 2025Q1 global inflation shock produces a moderate increase in domestic inflation relative to baseline for the first three periods; output gap begins to increase in 2025Q4 contributing to higher inflationary pressures.
  - Output gap:
    - Global output gap drives Tonga’s output gap; repeated −0.25 percent shocks from 2025Q2 decrease Tonga’s output gap relative to baseline.
    - Pa’anga depreciation during 2025Q4–2026Q1 increases the output gap relative to baseline.
  - Exchange rate and MCI:
    - Initial Pa’anga appreciation (2025Q1–2025Q3) makes the MCI less negative and decreases the domestic output gap relative to baseline.
    - From 2025Q4 onwards, Pa’anga depreciation makes the MCI more negative and causes the output gap to increase relative to baseline.
  - Real GDP growth and exchange rate path:
    - Real GDP growth relative to baseline: initial decrease until 2025Q3, then increase in 2025Q4 mainly driven by the exchange rate shock.
    - Weaker depreciation of the Pa’anga relative to baseline up to end-2026; larger appreciation under the alternative compared to baseline after peak in 2026Q1.
    - Summary: "Overall, the Tongan Pa’anga is anticipating lesser depreciation for the first four quarters, followed by stronger appreciation in 2026. High global inflation increases the real marginal cost, which results in higher domestic food and energy inflation."

### Conclusion — comparative impacts and policy relevance
- Comparative findings:
  - Bank failure scenario: main impact is to reduce the output gap and raise the MCI, primarily via the credit risk premium shock; aggregate loss of 1.5 percent of output over 2025Q1–2026Q4 relative to baseline.
  - Natural disaster scenario: shocks lead to increased domestic interest rates and MCI relative to baseline, and the speed of adjustment to the baseline is around four periods; estimated 4.6 percent decrease in output relative to baseline over 2025Q1–2026Q4.
  - Negative external shock scenario: wide-ranging impacts; rest-of-world inflation and output gap shocks directly affect domestic equivalents and cause complex interactions between the Pa’anga, MCI, output gap, and inflation.
- Methodological and policy relevance:
  - Results are conditional on the model calibration.
  - The QPM is useful for understanding shock transmission in Tonga and for guiding monetary policy discussions at NRBT given limited interest-rate pass-through and strong exchange-rate policy prominence.

*Source: IMF Working Paper — A Quarterly Projection Model for Tonga (Introduction and model calibration and scenarios) — wpiea2025121-print-pdf*

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

### Introduction

### Context and key vulnerabilities
- Tonga is a small open, import-dependent economy with a trade deficit averaging about 50 percent of GDP in the past decade.  
- Remittances are around 40 percent of GDP and, together with budget support, fund the large trade deficit.  
- Excess liquidity in the banking system exceeds 30 percent of GDP.  
- The National Reserve Bank of Tonga (NRBT) primary objectives:
  - maintain internal price stability by keeping inflation below a reference rate (currently of 5 percent), and
  - maintain an adequate level of foreign reserves of a minimum of 3 months (and optimally around 7.5 months) of imports.  
- The Tongan Pa’anga is pegged to a basket of currencies; NRBT can adjust exchange rates by 5 percent each month if necessary.  
- NRBT relies more on the exchange rate channel due to a weak interest rate transmission: the NRBT policy rate has been constant at zero percent since 2012.  
- To protect the peg and foreign reserves, NRBT employs capital controls; the Foreign Exchange Control Act of 2018 is cited.  
- Foreign reserves rose to over 10 months of imports in 2020; the interbank market has been inactive since 2010. NRBT has not issued notes or conducted open-market operations since 2009.  

### Recent shock history motivating the analysis
- Tonga has experienced a natural disaster every 2 years since 2014 (including tropical cyclones and the 2022 volcanic eruption and tsunami) and the COVID-19 pandemic, contributing to non-performing loans and weakened bank positions.  
- One domestic bank has shown pronounced deterioration, motivating a bank-failure scenario analysis.

### Purpose of the paper
- Customize a Quarterly Projection Model (QPM) to Tonga to produce medium-term quarterly projections of output, inflation, interest rate, and exchange rate.  
- Use the calibrated QPM to evaluate three scenarios: (1) bank failure, (2) natural disaster, and (3) negative external shock.

### Prior forecasting practice in Tonga
- NRBT previously used Excel-based sectoral forecasts (Real, External, Monetary); MOF produced fiscal forecasts; forecasts were produced separately with limited sectoral links and reconciled through discussion.  
- Medium-term forecasts were not produced systematically and relied mainly on staff expert judgment.

### _IMF working paper — A Quarterly Projection Model for Tonga_

---

### The QPM and its Calibration to Tonga

### Model structure (canonical QPM summary)
- Four main equations: aggregate demand, aggregate supply, uncovered interest parity (UIP), and a Taylor rule.  
- Aggregate demand links the output gap, monetary conditions, foreign output gap, and aggregate demand shocks.  
  - Output gap equation (symbolically referenced in the source).  
  - Monetary condition index (MCI) is a function of the interest rate gap and the exchange rate gap.  
- Aggregate supply links inflation to lagged inflation, expected inflation, real marginal cost, and supply shocks; real marginal cost depends on output gap and real exchange rate gap.  
- UIP links nominal exchange rate to expected future exchange rate, domestic and foreign nominal interest rates, a country risk premium, and an exchange rate shock.  
- Taylor rule specifies interest rate dynamics as a function of lagged interest rates, a neutral interest rate, expected inflation deviation from target, the output gap, and policy shocks.

### Adaptations for Tonga
- Exchange rate acts as the nominal anchor; Tonga is not an inflation-targeting country and conducts monetary policy via exchange rate interventions.  
- Monetary transmission is limited: small economy, high excess liquidity, high openness, low financial market development, heavy reliance on remittances.  
- Model features for Tonga:
  - Aggregate demand: moderate output-gap persistence; MCI gives more weight to exchange rate gap than interest rate gap; relatively low policy passthrough; low impact of external demand on domestic output.  
  - Aggregate supply: high persistence of core inflation; low persistence for domestic food and energy prices; sticky nominal wages and prices for core inflation; volatile passthrough for energy and food prices.  
  - UIP is modified to reflect managed float and exchange rate interventions; deviations from canonical UIP parameterized by p1 and h2. For Tonga, p1 = 0.2 and h2 = 0.5.  
  - Taylor rule: strong persistence in domestic interest rates; low responsiveness to output gap and inflation gap relative to the reference rate of 5 percent (model assumes this is the target).  
  - Interest rates are not sensitive to world interest rates due to capital controls and FX outflow restrictions.

### Key institutional and historical policy details included in calibration
- Excess liquidity >30 percent of GDP; NRBT policy rate constant at zero percent since 2012; capital controls introduced; Foreign Exchange Control Act of 2018 cited; foreign reserves >10 months in 2020; interbank market inactive since 2010; no NRBT notes issued since 2009.

### QPM Calibration for Tonga (selected calibrated parameter values)
- Aggregate demand equation:
  - b1 (Degree of output persistence): 0.4
  - b2 (Monetary policy passthrough): 0.3
  - b3 (Impact of external demand on domestic output): 0.3
  - b4 (Weight of real interest rate and real exchange rate in MCI): 0.2
  - b5 (Persistence of the credit premium): 0.7
- Aggregate supply equation:
  - a1 (Inflation persistence): 0.97
  - a2 (Policy passthrough — impact of real marginal cost on inflation): 0.05
  - a3 (Weight in real marginal cost equation, output gap vs exchange rate gap): 0.15
- Uncovered Interest Parity (UIP):
  - p1 (Deviation from canonical UIP): 0.2
  - h2 (Dirty float/managed float/peg indicator): 0.5
- Exchange rate intervention rule:
  - f1 (Contribution of past exchange rate target): 0.5
  - f2 (Contribution of deviation from inflation target): 0.25
  - f3 (Contribution of the output gap): 0.1
- Taylor rule:
  - g1 (Interest rate persistence): 0.98
  - g2 (Contribution of deviation from inflation target): 0.1

### Steady states used in the model
- Domestic variables:
  - Inflation target: 5
  - Real interest rate: 3
  - Trend real exchange rate: -1
  - Potential output: 2
  - Relative price of food inflation to CPI: 0
  - Relative price excluding food and energy to CPI: 0
- Foreign variables:
  - US inflation target: 2
  - US real interest rate: 0.75
  - Rest of the world food price: 0
  - Rest of the world oil price: 5

---

### Scenario analysis — Overview
- Three notional experiments implemented in the calibrated QPM to study impact relative to baseline:
  1. Bank failure (persistent credit premium shock)
  2. Natural disaster (shocks to potential output and inflation dynamics)
  3. Negative external shock (shocks to external conditions)
- The experiments are qualitative proxies to assess impulse response shapes rather than to mimic precise historical magnitudes.

### 1. A Bank Failure — setup and key assumptions
- Rationale: one large domestic bank with high non-performing loans and declining liquidity could fail, affecting credit conditions and country risk.  
- Assumed impact: a 1.5 percent decrease in output relative to the baseline over the period 2025Q1 – 2026Q4.  
- Shock specification (alternative scenario vs baseline):
  - Shock to the Credit Premium, εt_CRPRP (Credit Risk Premium), of:
    - 1.5 percent (150 basis pts) in 2025Q1,
    - 0.5 percent in 2025Q2,
    - 0.25 percent in 2025Q3,
    - followed by a permanent increase of 0.75 percent to the Credit Premium from 2025Q4 onward (through end of simulation 2026Q4).  
  - The exercise also captures an increase in country risk (prrpmt) due to a potential government-backed bank failure.  
- Model mechanics and responses:
  - MCI augmented by a credit premium variable (MCI = b4 * (real interest rate gap + CreditPremium) + (1−b4)*(−real exchange rate gap)). An increase in MCI is restrictive.  
  - Credit premium evolution equation:
    - CRPRPt = b5 * CRPRPt−1 + (1−b5) * (prrpmt − prrpmt−1) + εt_CRPRP + AAAAACRPRPt
    - with b5 = 0.7 (persistence of the credit premium).  
  - The impulse response shows:
    - an initial change in the MCI (alternative minus baseline) of 0.3 percent in 2025Q1,
    - a larger difference in 2025Q2, then a decline in 2025Q3 reflecting the sequence of shocks, and
    - an increase in the MCI gap from 2025Q4 onward due to the permanent 0.75 percent increase in the credit premium.  
  - Output gap is negatively related to the MCI; following the bank failure, the alternative vs baseline output gap becomes more negative, with the combined shocks amounting to a permanent loss over 2025Q1–2026Q4 of 1.5 percent of output relative to the baseline.

---

*Source: IMF Working Paper — A Quarterly Projection Model for Tonga (Introduction and model calibration and scenarios) — wpiea2025121-print-pdf*

### 1.25 percent of real GDP.

### wpiea2025121-print-pdf - 1.25 percent of real GDP.

### 1. Bank failure scenario — key model mechanisms and outcomes
- Monetary policy framework for Tonga in the model:
  - Assumes monetary policy is conducted more in terms of exchange rates rather than interest rates.
  - Interest rate channel rendered ineffective by excess liquidity; NRBT uses the exchange rate channel to achieve objectives.
- Predicted effects relative to the baseline:
  - Interest rates: mild reduction in interest rates relative to the baseline due to prior assumptions (low impact), attributable to the negative impact of the shock on the output gap.
  - Exchange rate: depreciates (Pa’anga depreciation) because the negative impact of the output gap on domestic interest rates; imports become more expensive.
  - Imported energy prices in local currency: initially more expensive following Pa’anga depreciation.
  - Inflation: under the alternative scenario relative to the baseline is decreasing (though slightly), consistent with chosen parameter settings in the core inflation and real marginal cost index in the model (low impact by assumption).
  - Core inflation: calibrated as highly persistent, in line with stylized facts.

### 2. Natural disaster scenario — assumptions and modeled impacts
- Context:
  - Tonga prone to natural disasters; recent examples include tropical cyclone Gita in 2018, tropical cyclone Harold in 2020, and HT-HH volcanic eruption in 2022.
  - Natural disasters typically damage capital formation (infrastructure and buildings); large inflows of aid often mitigate actual output effects through recovery efforts.
- Shock specification and baseline/alternative scenario assumptions:
  - The natural disaster shock modeled as a transitory downward shift in potential output using the QPM.
  - Estimated impact: 4.6 percent decrease in output relative to the baseline over the period 2025Q1 – 2026Q4.
  - Specific shocks for the exercise:
    - A positive shock to the output gap, 휀휀푡푡푦푦, of 5 percent in 2025Q1 due to a decline in potential output 휀휀푡푡푦푦� of 5 percent.
    - A staggered increase in annual inflation: 3 percent in 2025Q1, 1.5 percent in 2025Q2, 0.75 percent in 2025Q2 and 0.375 percent in 2025Q4.
  - (Model equations provided in the source specify relationships for the output gap, monetary condition index, and interest rate processes.)
- Model responses and transmission:
  - Output gap: shock leads to an increase in the output gap (given the specification).
  - Monetary condition index: output gap negatively related to the monetary condition index, which depends on the interest rate gap via the Taylor rule.
  - Interest rates:
    - For Tonga, the model predicts a mild reduction in interest rates relative to the baseline due to prior assumptions (low impact).
    - Following the natural disaster shock, interest rates decline by less than the baseline would otherwise show; this is attributed to the positive impact of the shock on the output gap. The source states: "In effect, interest rates go up with respect to the baseline."
    - The path of the alternative scenario is analogous to the baseline but with a higher interest rate, potentially reflecting increased default risk as businesses struggle after a natural disaster.
  - Exchange rate:
    - Alternative vs baseline: Pa’anga appreciates slightly due to the positive impact of the output gap on domestic interest rates; imports become less expensive, supporting recovery (in line with NRBT monetary policy practice).
    - Appreciation of the Pa’anga results in imported energy prices in local currency being initially less expensive.
  - Inflation:
    - Inflation under the alternative relative to the baseline is increasing (though slightly), consistent with scenario 1 (bank failure) dynamics.
- Additional parameter notes:
  - Domestic food comprises about 15 percent of the CPI basket.
  - History suggests local food prices tend to go up by more than 10 percent in the first quarter after a natural disaster followed by other price hikes from affected supply cycle.

### 3. Negative external shock — specification and impacts
- Shock representation (three exogenous components):
  1. A one-time shock to the world consumer price index of 1.5 percent in 2025Q1.
  2. A prolonged shock to the world output gap of -0.25 percentage points per quarter for four quarters commencing in 2025Q2.
  3. A shock to the exchange rate of Tonga (TOP/USD): appreciation over three periods from 2025Q1 to 2025Q3 followed by depreciation over two periods commencing 2025Q4.
- Table of assumptions (as given):
  - 2025Q1: Global Inflation Shock 1.5; Global Output Gap Shock 0; Exchange Rate Shock (TOP/USD) -0.5
  - 2025Q2: Global Inflation Shock 0; Global Output Gap Shock -0.25; Exchange Rate Shock (TOP/USD) -0.2
  - 2025Q3: Global Inflation Shock 0; Global Output Gap Shock -0.25; Exchange Rate Shock (TOP/USD) -0.1
  - 2025Q4: Global Inflation Shock 0; Global Output Gap Shock -0.25; Exchange Rate Shock (TOP/USD) 0.3
  - 2026Q1: Global Inflation Shock 0; Global Output Gap Shock -0.25; Exchange Rate Shock (TOP/USD) 0.5
- Transmission and model outcomes:
  - Domestic inflation:
    - General path under the alternative does not change but becomes slightly higher than the baseline at the time of the shock.
    - Higher global inflation increases real marginal cost, leading to higher food and energy inflation in Tonga.
    - Tonga’s non-core components of inflation (food and energy) make up 56.5 percent of the CPI basket; movements in global inflation tend to have high pass-through to domestic inflation.
    - The appreciation shock reduces inflation relative to the baseline for the first three periods after the shock; subsequent depreciation increases inflation relative to the baseline.
    - The one-time 2025Q1 global inflation shock produces a moderate increase in domestic inflation relative to the baseline for the first three periods, reflecting a lower output gap relative to the baseline; the output gap then begins to increase in 2025Q4, contributing to higher inflationary pressures.
  - Output gap:
    - The global output gap is a key determinant of Tonga’s output gap.
    - The negative -0.25 percent shock to the global output gap (repeated over four quarters from 2025Q2) decreases Tonga’s output gap relative to the baseline.
    - Owing to Pa’anga depreciation during 2025Q4-2026Q1, the output gap increases relative to the baseline.
  - Exchange rate and monetary condition index:
    - Appreciation of the Pa’anga for three periods from 2025Q1 makes the monetary condition index less negative under the alternative scenario; inverse relationship between monetary condition index and output gap causes a decrease in the domestic output gap relative to the baseline.
    - From 2025Q4 onwards, Pa’anga depreciation makes the monetary condition index more negative relative to the baseline and causes the output gap to increase relative to the baseline.
  - Real GDP growth and exchange rate path:
    - Overall effect on real GDP growth rate relative to the baseline: initial decrease from the beginning of the projection period until 2025Q3, then an increase in 2025Q4 driven mainly by the exchange rate shock.
    - There is a weaker depreciation of the Tongan Pa’anga relative to the baseline projection for the periods up to the end of 2026.
    - The appreciation of the Pa’anga is larger under the alternative compared to the baseline after the peak in 2026Q1.
    - Summary phrasing from the source: "Overall, the Tongan Pa’anga is anticipating lesser depreciation for the first four quarters, followed by stronger appreciation in 2026. High global inflation increases the real marginal cost, which results in higher domestic food and energy inflation."

### 4. Conclusion — comparative impacts and model usefulness
- Comparative findings:
  - Bank failure scenario:
    - Main impact was to reduce the output gap and raise the monetary condition index, primarily through the shock to the credit risk premium.
  - Natural disaster scenario:
    - Shocks led to an increase in domestic interest rates, increasing the monetary condition index relative to the baseline, and helped reduce the impact of the initial output gap shock in subsequent periods.
    - The speed of adjustment to the baseline was around four periods.
  - Negative external shock scenario:
    - Impacts are wide-ranging; shocks to rest-of-world inflation and output gap have direct impacts on domestic equivalents.
    - Nature of the exchange rate shock results in the alternative crossing the baseline for domestic output gap, monetary condition index, and core inflation; reflects strong relationship between domestic output gap and real exchange rate through the monetary condition index.
- Methodological and policy relevance:
  - Results are conditional on the calibration of the model.
  - The exercise confirms the usefulness of the QPM for understanding the impact of various shocks and guiding monetary policy discussions at the NRBT.

*IMF Working Paper: A Quarterly Projection Model for Tonga, Working Paper No. WP/2025/121*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025121-print-pdf.pdf_