## _wp12112

## Source details

**Canonical URL:** [_wp12112](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12112.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12112.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12112.pdf.json)

---

### I. Introduction
- Research questions:
  - Assess whether commodity price shocks significantly influence fiscal outcomes, inducing fiscal uncertainty (i.e., is there a prima facie case for hedging against commodity price shocks?).
  - Evaluate whether commodity price hedging, based on derivative instruments, is likely to yield significant benefits in practice.
- Context and motivation:
  - 2007–08 food and fuel price boom, possible second global food-price shock, and observed commodity price volatility increased interest in fiscal exposure to commodity shocks, budget uncertainty, pro-cyclical fiscal policy, and threats to debt sustainability—especially for LICs.
- Key empirical focus:
  - Impact of commodity export and import price shocks on revenue/GDP, expenditure/GDP, social expenditure/GDP, fiscal balance/GDP, and public debt/GDP across groups: LICs, MICs, commodity exporters, commodity importers.
- Hedging feasibility issues highlighted:
  - Role of policy buffers (exchange-rate regimes, reserves), availability/cost of hedging instruments for a narrow set of commodities, basis risk (world vs country-specific prices), volume shocks, and co-movements of export and import prices.

### II. Stylized Facts
- Importance and concentration:
  - Commodity exporters/importers defined where commodity exports/imports ≥ 20 percent of GDP.
  - Top 3 commodities account on average for over 40 percent of total commodity exports for developing countries.
  - Crude oil dominates; other key commodities: copper, fish, coal, iron.
- Volatility:
  - Developing-country average annual growth rate of commodity prices exhibits over most periods a standard deviation on the order of at least 2–3 percentage points.
  - For individual countries, panel standard deviation of commodity export and import prices ~10 percentage points.
- Simple correlations (change in ratios, 1990–2010):
  - Revenue / GDP correlated with commodity export prices: 0.53; commodity import prices: -0.26; top-3 export prices: 0.56; non-top-3 export prices: 0.21.
  - Expenditure / GDP correlated with commodity export prices: 0.29; import prices: -0.32.
  - Fiscal balance / GDP correlated with commodity export prices: 0.39; import prices: 0.36.
  - General Government Debt / GDP correlated with commodity export prices: -0.53; import prices: 0.40.
- Caution: univariate correlations are preliminary and should be interpreted cautiously.

### III. Methodology
- Data and sample:
  - Unbalanced annual panel, 1990–2010, up to 116 countries depending on variable.
- Benchmark fiscal exposure equation:
  - Dependent variables ∆Y_{i,t}: percentage point change in (i) total revenue / GDP; (ii) total expenditure / GDP; (iii) social expenditure / GDP; (iv) fiscal balance / GDP; (v) public debt / GDP.
  - p^{x}_{i,t}, p^{m}_{i,t}: country-specific, time-varying commodity export and import price indices (weights = actual share of each commodity).
  - x_{i,t} = commodity exports / GDP; h_{i,t} = commodity imports / GDP (lagged).
  - Long-run effect = short-run impact / (1 - γ).
  - Forecast error variance decomposition used to attribute variance to fiscal own effect, export price effect, import price effect, and interactions.
- Interaction variables and hypotheses:
  - E_{i,t} exchange-rate regime: E_{i,t}=1 for fixed regimes (Ilzetzki et al., 2009 classification); floating may dampen revenue impacts but may magnify external-debt burden if debt is FX-denominated.
  - R_{i,t} reserves relative to imports: competing hypotheses on buffering via exchange-rate stabilization vs smoothing consumption.
- Estimation:
  - Four specifications: (i) no interactions; (ii) exchange-rate interactions; (iii) reserve interactions; (iv) both interactions.
  - Tests for asymmetry: positive shocks > mean + 1.5 × standard deviation; negative shocks < mean − 1.5 × standard deviation.
  - Dynamic model: Arellano–Bond difference GMM; robustness: fixed-effects static specification.
- Sub-samples and narrow-commodity (top-3 vs non-top-3) analysis to assess hedging relevance.

### IV. Results — Overview
- Cross-country panels indicate large fiscal exposure to commodity-price shocks driven by:
  - automatic stabilizers on revenue; and
  - positive and significant expenditure responses, including social spending.
- Responses are relatively high in LICs, particularly LIC commodity exporters and importers.
- In LICs, expenditure typically responds more strongly than revenue, causing deficits and debt to rise in response to import-price increases (LIC commodity importers) and export-price increases (LIC commodity exporters).
- Effects larger under fixed exchange-rate regimes and persistent over time (significant lagged dependent variables).

### IV.A Commodity export prices — Key quantitative findings
- Total revenue:
  - In commodity exporters, a 10 percent increase in commodity export prices ⇒ short-run tax revenue increase of 0.53 to 0.61 percentage points of GDP.
  - Table 4 short-run impacts: LICs 0.28; & MICs 0.25; LIC & MIC Commodity Exporters 0.40; Commodity Exporters 0.53; LIC Commodity Importers 0.08.
  - Long-run impacts (example floating vs fixed): floating total revenue long-run: 0.48 0.46 0.67 0.81 0.13; fixed: 0.77 0.59 0.73 0.94 0.13 (values correspond to sample columns).
- Total expenditure:
  - 10 percent export-price increase ⇒ short-run public spending increase 0.33 to 0.97 percentage points of GDP; highest in LIC commodity exporters.
  - Table 4 short-run expenditure (fixed regime examples): 0.40 0.49 0.68 0.97.
- Social expenditure:
  - LIC and MIC commodity exporters: 10 percent export-price increase ⇒ social expenditure increase 0.45 percentage points of GDP (aggregate statement); Table 4 short-run examples show 0.07–0.78 depending on subgroup and regime.
- Fiscal balance:
  - Full sample: fiscal balance increases with export-price increases.
  - LIC commodity exporters: 10 percent export-price increase ⇒ deterioration of fiscal balance by 0.22 percentage points of GDP (expenditure rises faster than revenue).
  - Table 4 short-run fiscal deficit examples: floating: -0.07 0.03 0.16 0.18; fixed: -0.02 0.17 0.21 0.26.
- Public debt:
  - Full sample: debt response to export-price increases generally insignificant.
  - In LICs (including LIC commodity exporters) public debt rises significantly in response to export-price increases.
  - Table 4 short-run government debt examples: floating: 0.06 0.10 0.21; long-run floating: 0.09 0.13 0.31.
- Benchmark economy assumptions for quantitative estimates:
  - commodity export or import ratio equals sub-sample means: 23.8, 50.2, 17.9, 27.0, and 45.2 percent of GDP in, respectively, LICs and MICs, LIC and MIC commodity exporters, LICs, LIC commodity exporters, and LIC commodity importers;
  - exchange rate regime lies half-way between a floating and a fixed exchange rate regime.
- Illustrative historical example:
  - Nigeria: commodity-related revenue rose from 27 percent to 35 percent of GDP in 2000; total government debt grew from less than 14 percent in 1980 to more than 68 percent of GDP in 2000.

### IV.B Commodity import prices — Key quantitative findings
- Total revenue:
  - 10 percent increase in commodity import prices ⇒ short-run fiscal revenue increase of 0.15 to 0.59 percentage points of GDP; larger in LIC commodity importers.
  - Table 5 short-run total revenue examples (floating vs fixed): floating: 0.15 0.31 0.29 0.29 0.47; fixed: 0.25 0.31 0.38 0.29 0.59.
- Total expenditure and social expenditure:
  - Responses positive; especially large in LIC commodity importers: 10 percent import-price increase ⇒ total expenditure increase of 0.91 percentage points of GDP (average).
  - Table 5 short-run total expenditure examples: floating: 0.32 0.45; fixed: 0.46 0.50.
  - Social expenditure short-run examples include values up to 1.14 (floating) and 1.17 (fixed) for some subgroups.
- Fiscal deficit and debt in LIC commodity importers:
  - 10 percent import-price increase ⇒ average fiscal deficit increase of 0.22 percentage points of GDP.
- Dynamic buildup:
  - Responses generally build up over time (significant lagged dependent variables).

### IV.C Role of traditional buffers and hedging potential
- Floating exchange rates:
  - Partially dampen impacts in LICs; interaction coefficients generally significant; impacts larger under fixed regimes.
  - Example: 10 percent export-price increase ⇒ fiscal balance deterioration by 0.18 percentage points of GDP in LIC commodity exporters with a floating regime, and by 0.26 percentage points of GDP with a fixed regime.
- Foreign-exchange reserves:
  - No evidence that reserves buffer LICs against commodity price shocks; reserves × price interaction coefficients typically insignificant, including asymmetric specifications.
- Top-3 commodities:
  - Controlling for top-3 commodity prices, impacts of other commodity prices are on average statistically insignificant.
  - Coefficients on top-3 commodity prices often significant; non-top-3 coefficients generally insignificant—implying narrow-commodity hedging could capture most benefits.
- Variance decomposition:
  - In LICs, commodity export prices account for a large share (typically about 1/4) of changes in fiscal revenue, public expenditure, and government debt; effects stronger in LIC commodity exporters.
  - Commodity import prices and interactions have somewhat smaller effects.

### V.D Basis risk (definition, causes, empirical findings)
- Basis risk: difference between benchmark “world” commodity prices (used by hedging instruments) and country-specific commodity prices actually received/paid.
- Causes listed:
  - Differences in quality;
  - Barriers to trade (transportation costs);
  - Imperfect exchange-rate pass-through (firms with market power).
- Empirical approach:
  - Country-specific unit values for 40 commodities from UN COMTRADE; panel unit-root tests; panel regressions with country fixed effects; Pedroni (2000) FM-OLS where needed.
- Empirical conclusions:
  - Country-specific commodity prices are statistically significantly correlated with world commodity prices.
  - Correlation is significantly below unity: world price changes associated with smaller proportionate changes in country-specific prices.
  - World prices explain only a fraction of variation in country-specific prices.
  - For almost all commodities, unit-root null is rejected; where cointegration identified, long-run movements tend to be one-to-one.
  - Crude oil: tight link between country-specific and world prices.
- Caveat:
  - Unit value data may be noisy and could exaggerate basis risk.
- Implication:
  - Basis risk is present; hedging world-price fluctuations yields benefits but leaves exposure to idiosyncratic country-specific price movements; more heterogeneous commodities imply greater basis risk.

### V.E Commodity prices versus volumes
- Decomposition (1990–2010, Price change / Volume change, percent):
  - All Developing countries: 59.7 / 40.3
  - Non-fuel exporters: 62.0 / 38.0
  - Fuel exporters: 67.3 / 32.7
  - All commodity exporters: 60.5 / 39.5
  - All LICs: 63.0 / 37.0
  - (LIC) Fuel exporters: 52.1 / 47.9
- During 2000–10 (Price change / Volume change, percent), e.g.:
  - All Developing countries: 64.5 / 35.5
  - Fuel exporters: 70.0 / 30.0
  - (LIC) Fuel exporters: 57.3 / 42.7
- Variance decomposition (1990–2010, Pure Price / Pure Volume / Price×Volume, percent):
  - All Developing countries: 69.5 / 8.2 / 22.3
  - All LICs: 82.8 / 4.2 / 13.0
  - (LIC) Fuel exporters: 27.4 / 49.2 / 23.4
- Implication:
  - Changes in commodity export revenue are largely driven by price changes rather than volumes in most groups, supporting potential effectiveness of price-based hedging for revenue stabilization (with caveats on basis risk).

### V.F Co-movements in export and import prices
- Export and import prices typically positively correlated (Table 14):
  - MICs and LICs: 0.53
  - MIC and LIC Commodity Exporters: 0.61
  - LICs: 0.48
  - LIC Commodity Exporters: 0.67
  - LIC Commodity Importers: 0.68
- Empirical implications:
  - Co-movements can dampen or amplify fiscal impacts depending on signs/magnitudes.
  - Findings: import-price impacts often insignificant except in LIC commodity importers; in LIC commodity importers, export-price impacts typically insignificant.
  - Cases where both export and import price impacts significant (full sample): tax revenue, total revenue, and total expenditure respond positively to increases in both export and import prices.
  - Net effect: positive co-movements tend to amplify fiscal impacts.

### VI. Conclusions — Policy-relevant takeaways
- Commodity price shocks significantly affect fiscal outcomes, increasing budget uncertainty, encouraging pro-cyclical fiscal policy, and threatening debt sustainability — especially in LIC commodity exporters and importers.
- Mechanisms:
  - Tax revenue rises with commodity price increases (likely via trade taxes).
  - Expenditure (including social spending) rises with commodity price increases; LIC commodity exporters may treat price increases as permanent and increase spending accordingly; LIC commodity importers may increase social safety nets or subsidies when import prices rise.
  - Fiscal balance deteriorates in response to import price increases; response to export price increases is ambiguous, with LIC commodity exporters showing deterioration (expenditure rises faster than revenue).
- Traditional buffers provide limited protection:
  - Floating exchange rate partially offsets impacts; foreign-exchange reserves show no evidence of damping effects.
- Hedging potential and limitations:
  - Effective hedging instruments likely available at reasonable cost for only a few commodities, but hedging a narrow set of key commodities (top-3) may capture most potential benefits because top-3 commodities drive volatility.
  - Price hedging can largely stabilize export revenue because price shocks dominate volume shocks.
  - Hedging tied to benchmark world prices will leave countries exposed to basis risk from country-specific price movements, reducing overall hedging benefits.
- Co-movements:
  - Positive co-movements in export and import prices are more likely to amplify fiscal impacts.

*Content derived from _wp12112 - References and Appendices (source PDF).*

### References .............................................................................................................

### _wp12112 - References

### I. Introduction
- Research questions:
  - Assess whether commodity price shocks significantly influence fiscal outcomes, inducing fiscal uncertainty (i.e., is there a prima facie case for hedging against commodity price shocks?).
  - Evaluate whether commodity price hedging, based on derivative instruments, is likely to yield significant benefits in practice.
- Context and motivation:
  - The 2007–08 boom in food and fuel prices, indications of a possible second global food-price shock, and observed volatility in commodity prices have intensified interest in fiscal exposure to commodity shocks.
  - Concerns include increased budget uncertainty, pro-cyclical fiscal policy, and threats to debt sustainability—especially for low-income countries (LICs).
  - Debate on hedging considers market-based instruments and contingent official financing.
- Key empirical focus:
  - Impact of commodity export and import price shocks on fiscal revenue, expenditure, social expenditure, and public debt.
  - Analysis across several economic groupings: LICs, middle-income countries (MICs), commodity exporters, and commodity importers.
- Key conceptual issues for hedging feasibility:
  - Can policy buffers (floating exchange rate regimes, reserve assets) mitigate adverse impacts?
  - Availability and cost-effective coverage of hedging instruments for a narrow set of commodities.
  - Basis risk from benchmark world commodity prices versus country-specific prices.
  - Whether volume shocks alter the effectiveness of price hedging.
  - Co-movements of export and import prices and possible natural fiscal stabilization.

### II. Stylized Facts
A. The Importance of Commodity Trade
- Trends and definitions:
  - Since 1980, for developing countries as a whole, exports and imports have grown in absolute terms and relative to GDP.
  - Middle-income countries (MICs) are broadly more open than LICs.
  - “Commodity exporters” or “commodity importers” are defined as those countries where commodity exports or imports account for at least 20 percent of GDP.
  - Ratios to GDP are considered more relevant than ratios to total exports or imports.
- Concentration and specialization:
  - Both commodity exports and imports are significantly concentrated, with no clear trend toward increased or decreased specialization over time.
  - The share of the 3 most important commodities in total commodity exports or imports has remained roughly stable since the mid-1990’s.
  - The mean degree of specialization appears greatest in LICs.
  - For developing countries as a whole, the top 3 commodities account on average for over 40 percent of total commodity exports.
  - Crude oil dominates developing countries’ commodity exports and imports; other key commodities include copper, fish, coal, and iron.
- Volatility:
  - Commodity export and import prices are both very volatile.
  - For developing countries as a group, late 1990’s saw a sharp decrease in commodity prices, followed by a boom for most of the 2000’s, and a collapse in 2008–09.
  - Developing-country average annual growth rate of commodity prices exhibits over most time periods a standard deviation on the order of at least 2–3 percentage points.
  - For individual countries, commodity export and import prices display an overall panel standard deviation on the order of 10 percentage points.

B. Commodity Price Shocks and Fiscal Exposure
- Simple correlation evidence:
  - Data suggest strong correlations between world commodity prices and fiscal outcomes.
  - Commodity export prices have positive, relatively large correlations with revenue / GDP, and negative correlations with debt / GDP.
  - Correlations are even larger with respect to the prices of the 3 most important export commodities.
  - These preliminary univariate correlations (allowing for country fixed effects) suggest a prima facie case that hedging against commodity price volatility may smooth fiscal adjustment.
- Caution:
  - Univariate correlation results should be taken with caution.

### III. Methodology
- Dataset and scope:
  - Unbalanced annual panel covering the period 1990–2010 and (depending on the precise variable) up to 116 countries.
- Benchmark fiscal exposure equation (reduced-form cross-country panel regressions):
  - Dependent variable ∆Y_{i,t} is the percentage point change in each of the following (estimated separately):
    - (i) total revenue / GDP;
    - (ii) total expenditure / GDP;
    - (iii) social expenditure (on education and health) / GDP;
    - (iv) fiscal balance (surplus) / GDP;
    - (v) public debt / GDP.
  - p^{x}_{i,t} and p^{m}_{i,t} are country-specific, time-varying commodity (spot) export and import price indices, constructed using actual weights of each commodity in the country’s export or import basket and the world price of that commodity.
  - x_{i,t} = commodity exports / GDP and h_{i,t} = commodity imports / GDP are country-specific, time-varying weights lagged to reduce endogeneity concerns.
  - Long-run effect computed as short-run impact / (1 - γ).
  - Forecast error variance decomposition is obtained by taking the variance of both sides of the benchmark equation (averaged over time and across countries), yielding contributions from the fiscal variable’s own effect, pure commodity export price effect, pure commodity import price effect, and interactions.
- Economic priors:
  - In commodity exporters:
    - Expect commodity export prices to be positively associated with revenue (through income taxes and non-tax revenue like royalties and production sharing agreements).
    - Expect positive association with expenditure, including social expenditure; magnitude depends on perceived permanence of price changes and investment needs.
    - Less definite priors for fiscal balance and debt; normative presumption is fiscal balance should increase if price changes are seen as temporary.
  - In commodity importers:
    - Expect commodity import prices to be positively associated with revenue via trade taxes.
    - Expect positive association with expenditure, particularly social safety nets or food and fuel subsidies.
    - Less definite priors on fiscal balance and debt.
- Interaction variables and hypotheses:
  1. E_{i,t} (de facto exchange rate regime): E_{i,t} = 1 for fixed exchange rate regimes, 0 otherwise (classification based on Ilzetzki et al., 2009 approach).
     - Hypothesis: Floating exchange rates may weaken the impact of commodity-price changes on revenue by damping output impacts and allowing depreciation to increase trade-tax revenue; but may magnify impact on external debt service/debt burden if debt is foreign-currency denominated.
     - All intermediate regimes (e.g., crawling pegs) are treated as fixed exchange rate regimes.
  2. R_{i,t} (reserves relative to imports):
     - Two competing hypotheses:
       - Greater reserves may reduce the response of exchange rates to commodity price changes, reducing dampening from exchange rate movements.
       - Greater reserves may allow governments to smooth consumption in the face of negative shocks.
- Estimation and specifications:
  - Four specifications estimated: (i) neither exchange-rate nor reserve interactions; (ii) exchange-rate regime interactions; (iii) reserve interactions; (iv) both interactions.
  - Tests for asymmetric fiscal responses to positive versus negative commodity-price shocks:
    - Positive shocks defined as changes > mean + 1.5 × standard deviation.
    - Negative shocks defined as changes < mean − 1.5 × standard deviation.
  - Dynamic model estimated using the Arellano–Bond difference GMM.
  - Robustness: fixed-effects estimation of a static version excluding the lagged dependent variable was also conducted.
- Sample group estimations:
  - Responses are estimated for the following groups:
    - (i) LICs;
    - (ii) LIC commodity exporters;
    - (iii) LIC commodity importers;
    - (iv) MICs;
    - (v) MIC commodity exporters;
    - (vi) MIC commodity importers;
    - (vii) LICs and MICs;
    - (viii) LIC and MIC commodity exporters;
    - (ix) LIC and MIC commodity importers.
  - Focus on commodity exporters and importers motivated by hypothesis that impacts are easier to observe in countries heavily reliant on commodity trade.
- Narrow-commodity analysis for hedging relevance:
  - Re-estimate regressions replacing aggregate commodity price indices with sub-indices for:
    - the three most important commodities (the “top-3 commodities”);
    - all other commodities (the “non-top-3 commodities”).
  - Sub-indices are weighted by each commodity’s share in the country’s export or import basket.
- Additional notes:
  - Additional lags of the dependent variable were not found to be significant.
  - Owing to space constraints, results for some groups are omitted.

_Italicized source attribution: Content derived from _wp12112 - References (source PDF).

### Appendix I describes in greater detail the above variables and their sources. Appendix II lists

### _wp12112 - Appendix I describes in greater detail the above variables and their sources. Appendix II lists

### IV. RESULTS — overview
- Cross-country panel regressions indicate large fiscal exposure to commodity-price shocks due to:
  - automatic stabilizers on the revenue side; and
  - a positive and significant response of expenditure, including social spending, to commodity prices.
- Responses are relatively high in LICs, particularly in commodity exporters and importers.
- In these LIC groups, expenditure typically responds more strongly than revenue, producing significant positive responses of public deficits and debt to commodity import prices in LIC commodity importers, and to commodity export prices in LIC commodity exporters.
- Effects are larger under fixed exchange-rate regimes and persist over time (evidence of persistence via significant coefficients on lagged dependent fiscal variables).

### A. Commodity export prices and fiscal outcomes — key quantitative findings
- Total revenue:
  - In commodity exporters, a 10 percent increase in commodity export prices leads to an average short-run increase in tax revenue of 0.53 to 0.61 percentage points of GDP.
- Total expenditure:
  - A 10 percent increase in commodity export prices leads to an average increase in public spending of 0.33 to 0.97 percentage points of GDP, with the highest response in LIC commodity exporters.
- Social expenditure:
  - In LIC and MIC commodity exporters, a 10 percent increase in commodity export prices leads to an average increase in social expenditure of 0.45 percentage points of GDP.
- Fiscal balance:
  - In the full sample, fiscal balance increases in response to commodity export price increases.
  - In LIC commodity exporters, a 10 percent increase in commodity export prices leads to an average deterioration of the fiscal balance of 0.22 percentage points of GDP (expenditure rises faster than revenue).
- Public debt:
  - In the full sample, debt response to commodity export price increases is either statistically or economically insignificant.
  - In LICs (including LIC commodity exporters) public debt rises significantly in response to commodity export price increases.
- Short-run versus long-run:
  - Long-run effects are larger; coefficients on lagged dependent fiscal variables are generally significant, indicating persistence and accumulation of effects over time.
- Benchmark economy assumptions (for quantitative estimates unless otherwise stated):
  - commodity export or import ratio equals the mean value in the relevant sub-sample: 23.8, 50.2, 17.9, 27.0, and 45.2 percent of GDP in, respectively, LICs and MICs, LIC and MIC commodity exporters, LICs, LIC commodity exporters, and LIC commodity importers;
  - exchange rate regime lies half-way between a floating and a fixed exchange rate regime.

- Illustrative historical example:
  - Nigeria: commodity-related revenue rose from 27 percent to 35 percent of GDP in 2000; total government debt grew from less than 14 percent in 1980 to more than 68 percent of GDP in 2000.

### B. Commodity import prices and fiscal outcomes — key quantitative findings
- Total revenue:
  - A 10 percent increase in commodity import prices is associated with an average short-run increase in fiscal revenue of 0.15 to 0.59 percentage points of GDP; responses larger in LIC commodity importers.
- Total expenditure and social expenditure:
  - Responses positive; especially large in LIC commodity importers, where a 10 percent increase in commodity import prices is associated with an average increase in total expenditure of 0.91 percentage points of GDP.
- Fiscal deficit and debt in LIC commodity importers:
  - A 10 percent increase in commodity import prices leads to an average increase in the fiscal deficit of 0.22 percentage points of GDP.
- Dynamic buildup:
  - These responses generally build up over time.

### C. Role of traditional buffers and hedging potential
- Floating exchange rates:
  - Partially dampen the impact of commodity price shocks on LICs; interaction coefficients generally significant and imply bigger impacts under fixed regimes.
  - Example: a 10 percent increase in commodity export prices leads to a deterioration of the fiscal balance by 0.18 percentage points of GDP in LIC commodity exporters with a floating regime, and by 0.26 percentage points of GDP in LIC commodity exporters with a fixed regime.
- Foreign-exchange reserves:
  - No evidence that reserves provide buffering in response to commodity price shocks in LICs; coefficients on reserves–commodity price interaction terms are typically insignificant, even when both exchange rate regime and reserves are included.
  - Asymmetric specifications (reserves interacted separately with positive and negative shocks) yield similar insignificant results.
- Top-3 commodities versus others:
  - Controlling for prices of the top-3 commodities, the impact of other commodity prices is on average statistically insignificant.
  - Coefficients on top-3 commodity prices are often significant with expected signs; coefficients on non-top-3 commodity prices generally insignificant.
- Variance decomposition:
  - In LICs, commodity export prices account for a large share (typically, about ¼) of changes in fiscal revenue, public expenditure, and government debt; effects stronger in LIC commodity exporters.
  - Commodity import prices and interactions (prices × exchange rate regimes × reserves) have somewhat smaller effects.

### V. BROADER ISSUES — D. Basis risk
- Definition: basis risk is the difference between benchmark “world” commodity prices (used by hedging instruments) and country-specific commodity prices actually received or paid by a country.
- Causes of basis risk listed in the source:
  - Differences in quality (commodities not fully homogeneous).
  - Barriers to trade (e.g., transportation costs).
  - Imperfect exchange-rate pass-through (firms with market power may dampen or amplify exchange-rate impacts).
- Empirical approach:
  - Country-specific unit values for exports of 40 commodities obtained from UN COMTRADE.
  - Panel unit-root tests applied to country-specific commodity prices; where stationarity rejected, panel regression of country-specific prices on benchmark world price with country fixed effects; where unit-root not rejected, Pedroni (2000) heterogeneous-panel FM-OLS cointegration estimation used.
- Empirical conclusions (from Table 11 and Appendix IV):
  - Country-specific commodity prices are statistically significantly correlated with world commodity prices.
  - Correlation is significantly below unity: world price changes are associated with significantly smaller proportionate changes in country-specific prices.
  - World prices explain only a fraction of overall variation in country-specific prices.
  - For almost all commodities, unit-root null is rejected; where cointegration is identified, long-run movements tend to be one-to-one.
  - For crude oil, a tight link exists between country-specific and world prices (noted as an interesting case).
- Caveat:
  - Much analysis relies on unit value data, which may be noisy and could exaggerate basis risk.
- Implication:
  - Basis risk is present; hedging world-price fluctuations brings benefits but countries remain subject to significant idiosyncratic country-specific price fluctuations. More heterogeneous commodities imply greater basis risk.

### V. BROADER ISSUES — E. Commodity prices versus volumes
- Decomposition and variance analysis (Tables 12 and 13) indicate:
  - Changes in commodity export revenue are largely driven by changes in commodity prices rather than export volumes.
  - Variance decomposition shows commodity export revenue fluctuations largely reflect commodity price variance (with volume and covariance components smaller).

### V. BROADER ISSUES — F. Co-movements in export and import prices
- Commodity export and import prices are typically positively correlated (Table 14).
- Empirical implications:
  - Co-movements can either dampen or amplify fiscal impacts depending on sign/magnitude of separate effects.
  - Key findings: impact of commodity import prices often statistically insignificant except in LIC commodity importers; in LIC commodity importers, commodity export price impacts are typically insignificant.
  - Cases where both export and import price impacts are significant: tax revenue, total revenue, and total expenditure in the full sample — all respond positively to increases in both export and import prices.
  - Net effect: positive co-movements in export and import prices tend to amplify rather than dampen fiscal impacts.

### VI. CONCLUSIONS — policy-relevant takeaways
- Commodity price shocks significantly affect fiscal outcomes, increasing budget uncertainty, encouraging pro-cyclical fiscal policy, and threatening debt sustainability — especially in LIC commodity exporters and importers.
- Mechanisms:
  - Tax revenue rises with commodity price increases (likely via trade taxes).
  - Expenditure (including social spending) rises with commodity price increases; LIC commodity exporters may treat price increases as permanent and increase spending accordingly; LIC commodity importers may increase social safety nets or subsidies when import prices rise.
  - Fiscal balance deteriorates in response to import price increases; response to export price increases is ambiguous, with LIC commodity exporters showing a deterioration (expenditure rises faster than revenue).
- Traditional buffers provide limited protection:
  - A floating exchange rate partially offsets impacts; foreign-exchange reserves show no evidence of damping effects.
- Hedging potential:
  - Effective hedging instruments likely available at reasonable cost for only a few commodities, but hedging a narrow set of key commodities may realize most potential benefits because top-3 commodities tend to drive volatility.
  - Price hedging can largely stabilize export revenue because price shocks dominate volume shocks.
  - However, hedging tied to benchmark world prices will leave countries exposed to basis risk from country-specific price movements, reducing overall hedging benefits.
- Co-movements:
  - Positive co-movements in export and import prices are more likely to amplify fiscal impacts.

*Italic: Source — _wp12112 - Appendix I describes in greater detail the above variables and their sources. Appendix II lists (PDF chapter/section).*

### REFERENCES

### REFERENCES

### Bibliographic entries
- Arias, D., and Mario A. De Franco, 2011, “Integrating Central American and International Food Markets: An Analysis of Food Price Transmission in Honduras and Nicaragua,” LCSSD Occasional Paper Series on Food Prices.
- Arze del Granado, J., S. Gupta, and A. Hajdenberg, 2010, “Is Social Spending Procyclical?” IMF Working Paper 10/234 (Washington, D.C: IMF).
- Becker, T., O. Jeanne, P. Mauro, and R. Rancière, 2007, “Country Insurance: The Role of Domestic Policies,” IMF Occasional Paper No. 254 (Washington, DC: IMF).
- Borensztein, E., O. Jeanne, and D. Sandri, 2009, “Macro-Hedging for Commodity Exporters,” IMF Working Paper 09/299 (Washington, DC: IMF).
- Cespedes, L., and A. Velasco, 2011, “Was this time different? Fiscal Policy in Commodity Republics”, BIS Working Paper 365.
- Daniel, J., 2001, “Hedging Government Oil Price Risk,” IMF Working Paper 01/185 (Washington, DC: IMF).
- Dawe, D., 2010, “Cereal Price Transmission in Several Large Asian countries During the Global Food Crisis,” Asian Journal of Agriculture and Development, Vol. 6 (June).
- Food and Agriculture Organization of the United Nations and Organization for Economic Co-operation and Development, 2010, “Price Volatility and Price Transmission” in OECD-FAO Agricultural Outlook 2010-2019.
- Galí, J., and R. Perotti, 2003, “Fiscal policy and monetary integration in Europe,” Economic Policy 37: 533–572.
- Gilbert, C. L., 2011, “Grains Price Pass-Through, 2005-09,” in Safeguarding Food Security in Volatile Markets, ed. Adam Prakash (Rome: Food and Agriculture Organization of the United Nations).
- llzetzki, E., C. Reinhart, and K. Rogoff, 2009, “Exchange Rate Arrangements Entering the 21st Century: Which Anchor Will Hold?” mimeo (University of Maryland and Harvard University).
- Ilzetzki, E., and C.A. Végh, 2008, “Procyclical Fiscal Policy in Developing Countries: Truth or Fiction?” NBER Working Paper No 14191 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Kaminsky, G., 2010, “Terms Of Trade Shocks And Fiscal Cycles,” NBER Working Paper No 15780 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Kaminsky, G., C.M. Reinhart, and C.A. Végh, 2005, “When It Rains, It Pours: Procyclical Capital Flows and Macroeconomic Policies,” in NBER Macroeconomics Annual 2004, Vol. 19 (Cambridge, Massachusetts: National Bureau of Economic Research, MIT Press).
- Maliszewski, W., 2009, “Fiscal Policy Rules for Oil-Producing Countries: A Welfare-Based Assessment,” IMF Working Paper 09/126 (Washington, DC: IMF).
- Martin, W., and K. Anderson, 2011, “Export Restrictions and Price Insulation during Commodity Price Booms,” World Bank Policy Research Working Paper No 5645 (Washington, DC: World Bank).
- Martin, W., and K. Anderson, 2010, “Trade Distortions and Food Price Surges,” paper presented at 30th Anniversary Conference of the International Agricultural Trade Research Consortium (IATRC), Berkeley CA, December 12–14, 2010.
- Medina, L., 2010, “The Dynamic Effects of Commodity Prices on Fiscal Performance in Latin America” IMF Working Paper 10/192 (Washington, DC: IMF).
- Minot, N., 2010, “Transmission of World Food Price Changes to African Markets and its Effect on Household Welfare,” Michigan State University, Department of Agricultural, Food, and Resource Economics: Food Security Collaborative Working Papers No 58563.
- Organization for Economic Co-operation and Development, 2010, “Developments in Commodity Price Volatility,” document presented to the 52nd Session of the Working Party on Agriculture Policies and Markers.
- Ossowski, R., M. Villafuerte, P. Medas, and T. Thomas, 2008, Managing the Oil Revenue Boom: The Role of Fiscal Institutions, IMF Occasional Paper No. 260.
- Pedroni, P., 2000, “Fully-Modified OLS for Heterogeneous Cointegrated Panels,” Advances in Econometrics (15): 93–130.
- Villafuerte, M., P. Lopez-Murphy, and R. Ossowski, 2010, “Riding the Roller Coaster: Fiscal Policies of Nonrenewable Resource Exporters in Latin America and the Caribbean,” IMF WP 10/251 (Washington, DC: IMF).
- World Bank, 2009, Eastern Africa: A Study of the Regional Maize Market and Marketing Costs (Washington, DC: World Bank).
- Wright, B., 2011, “Grain Price Volatility and Its Economic Interpretation,” presented at the World Bank’s Development Economics Vice Presidency Lecture Series on March 11, 2011 in Washington, DC.

### Table 1. TOP 10 COMMODITIES: EXPORTS AS A SHARE OF TOTAL COMMODITY EXPORTS, 2000–08 (AGGREGATE VALUES)
- Source: IFS; WEO; World Bank; Commodity Prices Database; and COMTRADE/WITS.
- Columns: Rank | LICs Commodity Exporters | MICs Commodity Exporters | LICs & MICs | LIC Commodity Exporters | MIC Commodity Exporters
- 1: Crude Oil 22.3 | Crude Oil 31.8 | Crude Oil 28.9 | Crude Oil 26.4 | Crude Oil 33.0
- 2: Gold 11.2 | Copper 10.1 | Copper 9.6 | Cocoa 10.6 | Copper 13.3
- 3: Cocoa 8.3 | Fish 7.2 | Fish 6.7 | Gold 8.9 | Fish 9.0
- 4: Copper 6.0 | Coal 5.1 | Coal 5.4 | Copper 6.6 | Coal 6.0
- 5: Coffee 5.8 | Iron 4.1 | Iron 4.2 | Fish 6.2 | Aluminum 3.8
- 6: Shrimp 5.7 | Aluminum 4.1 | Aluminum 4.0 | Shrimp 5.6 | Palm Oil 3.7
- 7: Fish 5.5 | Palm Oil 3.0 | Gold 3.3 | Aluminum 5.1 | Rubber 3.0
- 8: Aluminum 4.4 | Soybean Meal 2.5 | Palm Oil 2.8 | Coffee 4.5 | Gold 2.4
- 9: Rubber 3.5 | Soybeans 2.3 | Soybean Meal 2.6 | Rice 4.1 | Hides 1.9
- 10: Cotton 3.4 | Gold 2.3 | Rubber 2.6 | Rubber 3.7 | Soybean Meal 1.7

### Table 2. TOP 10 COMMODITIES: IMPORTS AS A SHARE OF TOTAL COMMODITY IMPORTS, 2000–2008 (AGGREGATE VALUES)
- Columns: Rank | LICs Commodity Importers | MICs Commodity Importers | LICs & MICs | LIC Commodity Importers | MIC Commodity Importers
- 1: Crude Oil 54.2 | Crude Oil 51.4 | Crude Oil 51.2 | Crude Oil 56.4 | Crude Oil 49.0
- 2: Wheat 6.5 | Fish 7.1 | Fish 6.7 | Gold 5.1 | Fish 7.9
- 3: Rice 4.8 | Copper 5.7 | Copper 5.4 | Wheat 4.4 | Copper 6.4
- 4: Fish 4.2 | Iron 4.4 | Iron 4.4 | Fish 4.1 | Iron 6.1
- 5: Palm Oil 4.0 | Gold 3.7 | Gold 3.8 | Copper 3.8 | Aluminum 4.3
- 6: Gold 3.1 | Aluminum 3.6 | Aluminum 3.5 | Rice 3.7 | Soybeans 3.1
- 7: Sugar 2.4 | Coal 2.8 | Coal 2.7 | Hides 3.5 | Hides 2.6
- 8: Cotton 2.4 | Soybeans 2.3 | Soybeans 2.2 | Aluminum 2.7 | Coal 1.9
- 9: Aluminum 2.2 | Hides 2.1 | Wheat 2.2 | Soybean Meal [text truncated in source]
- 10: [table content continues beyond provided extract]

*Source: IFS; WEO; World Bank; Commodity Prices Database; and COMTRADE/WITS.*

### 2.1       Wheat       1.5

### _wp12112 - 2.1       Wheat       1.5

### Correlation analysis: fiscal outcomes and commodity prices (1990–2010)
- Change in: Revenue / GDP correlated with:
  - Expenditure / GDP: 0.26
  - Fiscal balance / GDP: 0.16
  - General Government Debt / GDP: -0.21
  - Commodity export prices: 0.53
  - Commodity import prices: -0.26
  - Top-3 commodity export prices: 0.56
  - Non-top-3 commodity export prices: 0.21
- Change in: Expenditure / GDP correlated with:
  - Fiscal balance / GDP: 0.27
  - General Government Debt / GDP: 0.19
  - Commodity export prices: 0.29
  - Commodity import prices: -0.32
  - Top-3 commodity export prices: 0.47
  - Non-top-3 commodity export prices: -0.16
- Change in: Fiscal balance / GDP correlated with:
  - General Government Debt / GDP: -0.09
  - Commodity export prices: 0.39
  - Commodity import prices: 0.36
  - Top-3 commodity export prices: 0.58
  - Non-top-3 commodity export prices: 0.18
- General Government Debt / GDP correlations with commodity prices:
  - Commodity export prices: -0.53
  - Commodity import prices: 0.40
  - Top-3 commodity export prices: -0.54
  - Non-top-3 commodity export prices: -0.13
- Commodity exporters (subset) correlations of Revenue / GDP:
  - Expenditure / GDP: 0.02
  - Fiscal balance / GDP: 0.18
  - General government: 0.36
  - Commodity export: 0.57
  - Commodity import: -0.10
  - Top-3 commodity: 0.61
  - Non-Top-3 commodity: 0.23
- Commodity importers (subset) correlations of Revenue / GDP:
  - Expenditure / GDP: 0.87
  - Fiscal balance / GDP: 0.23
  - General government debt / GDP: 0.31
  - Commodity export prices: 0.01
  - Commodity import prices: 0.64
  - Top-3 commodity export prices: 0.03
  - Non-Top-3 commodity export prices: 0.04

### Fiscal responses to a 10 percent increase in commodity export prices (percentage points of GDP) — Table 4
- Total revenue — Short-run impact:
  - LICs: 0.28
  - & MICs: 0.25
  - LIC & MIC Commodity Exporters: 0.40
  - Commodity Exporters: 0.53
  - LIC Commodity Importers: 0.08
- Total revenue — Short-run impact by exchange rate regime:
  - Floating exchange rate regime: 0.28 0.25 0.40 0.53 0.08
  - Fixed exchange rate regime: 0.45 0.33 0.43 0.61 0.08
- Total revenue — Long-run impact:
  - Floating exchange rate regime: 0.48 0.46 0.67 0.81 0.13
  - Fixed exchange rate regime: 0.77 0.59 0.73 0.94 0.13
- Total expenditure — Short-run impact:
  - Floating exchange rate regime: 0.33 0.37 0.35 0.84
  - Fixed exchange rate regime: 0.40 0.49 0.68 0.97
- Total expenditure — Long-run impact:
  - Floating exchange rate regime: 0.48 0.60 0.72 1.22
  - Fixed exchange rate regime: 0.58 0.80 1.40 1.42
- Social expenditure — Short-run impact:
  - Floating exchange rate: 0.07 0.42 0.55 0.38 0.12
  - Fixed exchange rate: 0.13 0.48 0.78 0.38 0.15
- Social expenditure — Long-run impact:
  - Floating exchange rate: 0.15 0.58 0.73 0.55 0.16
  - Fixed exchange rate: 0.27 0.67 1.03 0.55 0.20
- Fiscal deficit — Short-run impact:
  - Floating exchange rate regime: -0.07 0.03 0.16 0.18
  - Fixed exchange rate regime: -0.02 0.17 0.21 0.26
- Fiscal deficit — Long-run impact:
  - Floating exchange rate regime: -0.11 0.04 0.33 0.30
  - Fixed exchange rate regime: -0.03 0.21 0.44 0.43
- Government debt — Short-run impact:
  - Floating exchange rate regime: 0.06 0.10 0.21
  - Fixed exchange rate regime: 0.06 0.10 0.21
- Government debt — Long-run impact:
  - Floating exchange rate regime: 0.09 0.13 0.31
  - Fixed exchange rate regime: 0.09 0.13 0.31
- Source: Authors’ calculations, based on regression results (Tables 6–11). The short-term response of a fiscal variable is computed as a linear combination of the statistically significant coefficients, excluding that on the lagged dependent variable.

### Fiscal responses to a 10 percent increase in commodity import prices (percentage points of GDP) — Table 5
- Total revenue — Short-run impact:
  - Floating exchange rate: 0.15 0.31 0.29 0.29 0.47
  - Fixed exchange rate regime: 0.25 0.31 0.38 0.29 0.59
- Total revenue — Long-run impact:
  - Floating exchange rate: 0.26 0.57 0.50 0.45 0.78
  - Fixed exchange rate regime: 0.42 0.57 0.65 0.45 0.98
- Total expenditure — Short-run impact:
  - Floating exchange rate: 0.32 0.45
  - Fixed exchange rate regime: 0.46 0.50
- Total expenditure — Long-run impact:
  - Floating exchange rate: 0.46 0.91
  - Fixed exchange rate regime: 0.67 1.02 1.41
- Social expenditure — Short-run impact:
  - Floating exchange rate: 0.15 0.72 0.06 1.14
  - Fixed exchange rate: 0.20 0.06 0.95 0.06 1.17
- Social expenditure — Long-run impact:
  - Floating exchange rate: 0.31 0.96 0.09 1.49
  - Fixed exchange rate regime: 0.43 0.08 1.25 0.09 1.54
- Fiscal deficit — Short-run impact:
  - Floating exchange rate: 0.12 0.17
  - Fixed exchange rate regime: 0.12 0.22
- Fiscal deficit — Long-run impact:
  - Floating exchange rate: 0.17 0.36
  - Fixed exchange rate regime: 0.17 0.46 0.86
- Government debt — Short-run impact:
  - Floating exchange rate: 0.25
  - Fixed exchange rate regime: 0.25
- Government debt — Long-run impact:
  - Floating exchange rate: 0.32
  - Fixed exchange rate regime: 0.26 0.32 0.69
- Source: Authors’ calculations, based on regression results (Table 6 - Table 10). The responses are computed as linear combination of the statistically significant coefficients.

### Panel regression findings (Tables 6–10) — selected empirical highlights
- Change in central government total revenue / GDP (Table 6):
  - Lagged dependent variable ranges: 0.415***, 0.447***, 0.414***, 0.355**, 0.399***, 0.319***, 0.363***, 0.407***, 0.328***, 0.571** (standard errors reported in table).
  - Change in commodity export prices, weighted: 0.1188***; 0.1264***; 0.1904***; 0.2541***; 0.0403*.
  - Change in commodity import prices, weighted: 0.0774**; 0.1574**; 0.1272**; 0.1451**; 0.1547**.
  - Exchange rate regime (floating=0) coefficients: 0.247**, 0.547**, 0.108**, 0.202**, 0.277**.
  - Observations: 1,089; 704; 324; 660; 894 (panel sizes by specification). Number of countries: 93; 56; 32; 54; 76.
- Change in central government total expenditure / GDP (Table 7):
  - Lagged dependent variable examples: 0.317***, 0.386**, 0.511***, 0.315***, 0.321***, 0.165, 0.351**, 0.522***, 0.331***, 0.316***.
  - Change in commodity export price (weighted): 0.165***; 0.139***; 0.177***; 0.391***.
  - Change in commodity import price (weighted): 0.109***; 0.305; 0.223***; 0.187; 0.454***.
  - Exchange rate regime effects and interactions reported; Observations: 1,167; 719; 396; 218; 282. Number of countries: 97; 56; 33; 17; 22.
- Change in central government social expenditure / GDP (Table 8):
  - Lagged dependent variable examples: 0.525***; 0.277***; 0.244***; 0.312***; 0.238***; 0.471***; 0.244***; 0.171***; 0.162***; 0.397***.
  - Change in commodity export prices, weighted: coefficients reported (e.g., 0.124; 0.158*; 0.277**; 0.154**; 0.109).
  - Change in commodity import prices, weighted: 0.186; 0.434; 0.361***; 0.207; 0.483**.
  - Observations: 1,080; 643; 462; 794; 874. Number of countries: 112; 63; 48; 59; 88.
- Change in central government fiscal balance / GDP (Table 9):
  - Lagged dependent variable examples: 0.325***; 0.220***; 0.521***; 0.388***; 0.405***; 0.292*; -0.362*; 0.483***; 0.413***; 0.424***.
  - Change in commodity export prices, weighted: -0.037***; 0.017***; 0.079***; 0.085**; 0.101.
  - Change in commodity import prices, weighted: 0.059**; 0.209; 0.079**; -0.128; 0.079**.
  - Observations: 1,317; 656; 472; 794; 1069. Number of countries: 104; 52; 38; 59; 85.
- Change in central government gross debt (Table 10):
  - Lagged dependent variable examples: 0.241**; 0.329***; 0.210***; 0.309***; 0.327***; 0.089***; 0.368*; 0.0502**; 0.087**; 0.095**.
  - Change in commodity export prices, weighted: 0.156; 0.0333; 0.0503***; 0.107**; 0.442.
  - Change in commodity import prices, weighted: 0.0979*; 0.408; 0.125**; 0.491; 0.198**.
  - Observations: 906; 476; 382; 511; 752. Number of countries: 116; 61; 49; 60; 95.

### Basis risk: cross-country panel regression (Table 11)
- Dependent variable: Log Country-Specific Commodity Prices. Independent variable: Log World Commodity Prices.
- Commodity group averages and diagnostics:
  - Food:
    - Average β Coefficient: 0.50
    - % of commodities s.t. p-value for β=0 < 10%: 95
    - % of commodities s.t. p-value for β=1 < 10%: 100
    - Avg R2: 0.47
    - Avg Countries: 92
    - Number of commodities: 23
    - Avg Obs: 1,566
  - Agricultural Raw Materials:
    - Average β Coefficient: 1.46
    - % of commodities s.t. p-value for β=0 < 10%: 100
    - % of commodities s.t. p-value for β=1 < 10%: 100
    - Avg R2: 0.34
    - Avg Countries: 89
    - Number of commodities: 7
    - Avg Obs: 1,481
  - Metals:
    - Average β Coefficient: 0.74
    - % of commodities s.t. p-value for β=0 < 10%: 100
    - % of commodities s.t. p-value for β=1 < 10%: 75
    - Avg R2: 0.49
    - Avg Countries: 93
    - Number of commodities: 8
    - Avg Obs: 1,568
  - Fuel:
    - Average β Coefficient: 0.82
    - % of commodities s.t. p-value for β=0 < 10%: 100
    - % of commodities s.t. p-value for β=1 < 10%: 100
    - Avg R2: 0.49
    - Avg Countries: 83
    - Number of commodities: 2
    - Avg Obs: 1,868
- Notes: Methodology includes panel unit-root tests, panel regressions controlling for country fixed effects, and Pedroni (2000) heterogeneous-panel FM-OLS estimator where applicable. Results presented are averages for each commodity group.

*Source: _wp12112 - 2.1       Wheat       1.5 (PDF chapter/section)._

### Appendix III for results for each commodity.

### _wp12112 - Appendix III for results for each commodity

### Changes in commodity exports: relative contributions of price versus volume changes (Table 12)
- During 1990–2010 (Price change / Volume change, percent):
  - All Developing countries: 59.7 / 40.3
  - Non-fuel exporters: 62.0 / 38.0
  - Fuel exporters: 67.3 / 32.7
  - All commodity exporters: 60.5 / 39.5
  - All LICs: 63.0 / 37.0
  - (LIC) Non-fuel exporters: 60.7 / 39.3
  - (LIC) Fuel exporters: 52.1 / 47.9
  - (LIC) All commodity exporters: 61.0 / 39.0
  - All commodity exporters (duplicate line in source): 61.0 / 39.0

- During 2000–10 (Price change / Volume change, percent):
  - All Developing countries: 64.5 / 35.5
  - Non-fuel exporters: 60.9 / 39.1
  - Fuel exporters: 70.0 / 30.0
  - All commodity exporters: 66.4 / 33.6
  - All LICs: 64.0 / 36.0
  - (LIC) Non-fuel exporters: 59.3 / 40.7
  - (LIC) Fuel exporters: 57.3 / 42.7
  - (LIC) All commodity exporters: 58.7 / 41.3

### Variance decomposition of commodity exports: prices versus volumes (Table 13)
- Identity used: ln(Commodity Exports) ≡ ln(Commodity Price Index) + ln(Commodity Volumes); variances and covariances of RHS expressed as sum of individual variances and covariances.
- During 1990–2010 (Pure Price Effect / Pure Volume Effect / Price and Volume Correlation Effect, percent):
  - All Developing countries: 69.5 / 8.2 / 22.3
  - Non-fuel exporters: 72.3 / 6.2 / 21.4
  - Fuel exporters: 73.5 / 10.4 / 16.2
  - All commodity exporters: 75.9 / 6.8 / 17.3
  - All LICs: 82.8 / 4.2 / 13.0
  - (LIC) Non-fuel exporters: 69.9 / 8.2 / 21.9
  - (LIC) Fuel exporters: 27.4 / 49.2 / 23.4
  - (LIC) All commodity exporters: 72.2 / 7.4 / 20.4

- During 2000–10 (Pure Price Effect / Pure Volume Effect / Price and Volume Correlation Effect, percent):
  - All Developing countries: 65.2 / 5.6 / 29.3
  - Non-fuel exporters: 74.4 / 3.2 / 22.4
  - Fuel exporters: 77.5 / 3.0 / 19.5
  - All commodity exporters: 77.1 / 2.6 / 20.3
  - All LICs: 83.7 / 5.3 / 11.0
  - (LIC) Non-fuel exporters: 78.7 / 4.6 / 16.6
  - (LIC) Fuel exporters: 21.2 / 45.3 / 33.5
  - (LIC) All commodity exporters: 82.7 / 3.4 / 13.9

### Correlations between change in commodity export prices and change in commodity import prices, 1990–2010 (Table 14)
- Sample includes 85 MICs and 47 LICs. The coefficients control for country effects.
- Correlations:
  - MICs and LICs: 0.53
  - MIC and LIC Commodity Exporters: 0.61
  - LICs: 0.48
  - LIC Commodity Exporters: 0.67
  - LIC Commodity Importers: 0.68

### Figures and stylized facts (selected)
- Trade / GDP, 1981–2008 (mean values): Figures show time series for Exports / GDP and Imports / GDP across All Countries, LICs, MICs, Developing Countries, and Commodity Exporters (1981–2006 data points shown in figures).
- Commodity trade / GDP, 1988–2008 (mean values): Commodity Exports / GDP and Commodity Imports / GDP across country groups (1988–2008).
- Share of top 3 commodities in total commodity trade, 1988–2008 (mean values): Top-3 Commodity Exports / Total Commodity Exports and Top-3 Commodity Imports / Total Commodity Imports across country groups.
- Growth rates of commodity export and import prices, 1989–2010 (income effect, percent of GDP, mean values): Income effect calculated as the growth rate of commodity export (or import) prices multiplied by the share of commodity exports (or imports) in GDP; figures show Income Effect (Percent of GDP) and Growth of Commodity Export/Import Prices for All Countries, LICs, MICs, Developing Countries, Commodity Exporters.
- Short-run fiscal responses to a 10 percent increase in commodity prices:
  - Export prices (Arellano-Bond estimates, average short-run effects over 1990–2010, benchmark where commodity exports / GDP = 20 percent): figures report effects (percentage points of GDP) on Total revenue, Total expenditure, Social expenditure, Fiscal deficit for All LICs & MICs; LIC and MIC Commodity Exporters; LICs; LIC Commodity Exporters; LIC Commodity Importers (scale shown from -0.3 to 1.8 percentage points).
  - Import prices (Arellano-Bond estimates, average short-run effects over 1990–2010, benchmark where commodity imports / GDP = 20 percent): analogous figures for Total revenue, Total expenditure, Social expenditure, Fiscal deficit (scale shown from -0.3 to 1.8 percentage points).
- Contributions of commodity prices to changes in fiscal outcomes: variance decomposition figures show shares of Commodity export price effect, Commodity import price effect, Interaction effect, and Other factors for Fiscal Revenue, Public Expenditure, and Central Government Gross Debt for LICs and Commodity exporters (displayed as percentages from -20% to 100% or 0% to 100% depending on panel).

### Data description and definitions (Appendix I)
- Revenue: Government tax revenue, non tax revenue, and grants as share of GDP (percent). Source: IMF, World Economic Outlook.
- Social spending: Government education spending and Government health spending as share of GDP. Source: IMF Fiscal Affairs Department.
- Fiscal deficit: Government total expenditure minus government revenue (including grants) as share of GDP (percent). Source: IMF, World Economic Outlook.
- Debt: Government gross debt as a share of GDP (percent). Source: IMF, World Economic Outlook.
- Reserves: International Reserves in percent of prospective (subsequent year) import of goods and Services. IMF, World Economic Outlook; authors’ calculations.
- Exchange rate regime: floating = 0, fixed = 1 (based on the Exchange Rate Classification in Ilzetzki et al., 2009).
- Commodity exports and commodity imports: Source: UN COMTRADE. Commodity type lists provided in source for Food, Agricultural Raw Materials, Metals, Fuel (exact commodity lists in source).
- Commodity exporter: = 1 if average commodity exports to GDP over 2005–09 exceed 20 percent. Sources: IMF, World Economic Outlook.
- Commodity importer: = 1 if average commodity imports to GDP over 2005–09 exceed 20 percent. IMF, World Economic Outlook.
- Commodity Export Price Index: For each country and year, a weighted average of the growth rate of commodity prices is constructed, using weights given by the ratio of exports of the given commodity to total commodity exports.
- Commodity Import Price Index: For each country and year, a weighted average of the growth rate of commodity prices is constructed, using weights given by the ratio of imports of the given commodity to total commodity imports.
- Commodity Export Price Index (2005=100), Weighted by Total Commodity Exports to GDP: constructed by multiplying the growth rate of the commodity export price index by total commodity exports to GDP.
- Commodity Import Price Index (2005=100), Weighted by Total Commodity Imports to GDP: constructed by multiplying the growth rate of the commodity import price index by total commodity imports to GDP.

### Country groupings (Appendix II)
- Country list with Region and Income Group is provided in the source (multiple pages listing countries and their region and income classification).

*Source: _wp12112 - Appendix III for results for each commodity.*

### APPENDIX III: SUMMARY STATISTI CS

### APPENDIX III: SUMMARY STATISTI CS

### All LICs and MICS, 1990–2010
- Observations and summary statistics (Percent of GDP):
  - Revenue: Obs 3100; Mean 21.1; Std. Dev. 15.6; Min 6.9; Max 46.9
  - Tax: Obs 1358; Mean 15.2; Std. Dev. 17.9; Min 5.1; Max 38.0
  - Expenditure: Obs 3220; Mean 25.2; Std. Dev. 17.4; Min 12.2; Max 45.5
  - Social expenditure: Obs 2067; Mean 6.6; Std. Dev. 3.6; Min 5.0; Max 28.5
  - Fiscal balance: Obs 3640; Mean -3.0; Std. Dev. 14.1; Min -14.3; Max 7.6
  - Debt: Obs 1700; Mean 53.1; Std. Dev. 6.6; Min 11.0; Max 98.1
- Change in ratio to GDP:
  - Revenue: Obs 2914; Mean 0.2; Std. Dev. 11.1; Min -4.5; Max 2.8
  - Tax: Obs 1223; Mean 0.0; Std. Dev. 13.4; Min -3.3; Max 3.7
  - Expenditure: Obs 3026; Mean 0.2; Std. Dev. 11.4; Min -1.5; Max 4.1
  - Social expenditure: Obs 1875; Mean 0.1; Std. Dev. 1.2; Min -1.7; Max 1.2
  - Fiscal balance: Obs 3455; Mean 0.0; Std. Dev. 11.9; Min -5.0; Max 3.4
  - Debt: Obs 1547; Mean 0.6; Std. Dev. 6.7; Min -2.0; Max 4.6

### LIC and MIC Commodity Exporters, 1990–2010
- Observations and summary statistics (Percent of GDP):
  - Revenue: Obs 1600; Mean 23.3; Std. Dev. 16.5; Min 20.0; Max 46.9
  - Tax: Obs 752; Mean 15.1; Std. Dev. 16.3; Min 8.6; Max 38.0
  - Expenditure: Obs 1655; Mean 26.5; Std. Dev. 16.5; Min 18.7; Max 43.3
  - Social expenditure: Obs 1026; Mean 6.8; Std. Dev. 2.9; Min 9.9; Max 17.3
  - Fiscal balance: Obs 1886; Mean -3.0; Std. Dev. 16.0; Min -8.3; Max 7.6
  - Debt: Obs 836; Mean 49.1; Std. Dev. 8.2; Min 11.4; Max 77.3
- Change in ratio to GDP:
  - Revenue: Obs 1512; Mean 0.2; Std. Dev. 14.9; Min -4.5; Max 2.8
  - Tax: Obs 688; Mean 0.0; Std. Dev. 12.8; Min -3.3; Max 3.7
  - Expenditure: Obs 1567; Mean 0.1; Std. Dev. 15.0; Min -1.1; Max 2.2
  - Social expenditure: Obs 937; Mean 0.1; Std. Dev. 1.0; Min -0.4; Max 1.2
  - Fiscal balance: Obs 1797; Mean 0.0; Std. Dev. 6.5; Min -4.8; Max 7.6
  - Debt: Obs 760; Mean 0.3; Std. Dev. 6.2; Min -1.6; Max 4.7

### LICs, 1990–2010
- Observations and summary statistics (Ratio to GDP, percent):
  - Revenue: Obs 1275; Mean 17.8; Std. Dev. 14.1; Min 6.9; Max 30.9
  - Tax: Obs 446; Mean 13.9; Std. Dev. 18.1; Min 5.1; Max 28.7
  - Expenditure: Obs 1299; Mean 23.4; Std. Dev. 17.5; Min 12.2; Max 34.3
  - Social expenditure: Obs 984; Mean 6.5; Std. Dev. 4.2; Min 5.0; Max 16.1
  - Fiscal balance: Obs 1472; Mean -3.5; Std. Dev. 7.4; Min -14.3; Max 8.4
  - Debt: Obs 586; Mean 70.6; Std. Dev. 8.3; Min 11.0; Max 98.1
- Change in ratio to GDP:
  - Revenue: Obs 1187; Mean 0.2; Std. Dev. 4.0; Min -4.5; Max 2.0
  - Tax: Obs 392; Mean 0.2; Std. Dev. 11.9; Min -3.3; Max 1.6
  - Expenditure: Obs 1203; Mean 0.2; Std. Dev. 7.2; Min -1.5; Max 4.1
  - Social expenditure: Obs 900; Mean 0.0; Std. Dev. 1.1; Min -1.7; Max 1.7
  - Fiscal balance: Obs 1384; Mean 0.0; Std. Dev. 7.7; Min -5.0; Max 2.7
  - Debt: Obs 520; Mean 0.8; Std. Dev. 5.2; Min -2.0; Max 4.6

### LIC Commodity Exporters, 1990–2010
- Observations and summary statistics (Ratio to GDP, percent):
  - Revenue: Obs 371; Mean 19.1; Std. Dev. 8.9; Min 20.0; Max 30.9
  - Tax: Obs 147; Mean 13.5; Std. Dev. 4.1; Min 8.6; Max 28.7
  - Expenditure: Obs 382; Mean 26.8; Std. Dev. 15.0; Min 18.7; Max 34.3
  - Social expenditure: Obs 312; Mean 10.4; Std. Dev. 3.2; Min 9.9; Max 16.1
  - Fiscal balance: Obs 42; Mean -4.2; Std. Dev. 9.5; Min -8.3; Max 8.4
  - Debt: Obs 154; Mean 35.3; Std. Dev. 6.4; Min 11.4; Max 77.3
- Change in ratio to GDP:
  - Revenue: Obs 351; Mean 0.1; Std. Dev. 5.2; Min -4.5; Max 2.0
  - Tax: Obs 132; Mean 0.2; Std. Dev. 1.8; Min -3.3; Max 1.6
  - Expenditure: Obs 360; Mean 0.0; Std. Dev. 11.4; Min -1.1; Max 2.2
  - Social expenditure: Obs 290; Mean 0.1; Std. Dev. 1.1; Min -0.4; Max 0.8
  - Fiscal balance: Obs 40; Mean 0.0; Std. Dev. 8.7; Min -4.8; Max 2.7
  - Debt: Obs 131; Mean 1.4; Std. Dev. 7.0; Min -1.6; Max 3.9

### LIC Commodity Importers, 1990–2010
- Observations and summary statistics (Ratio to GDP, percent):
  - Revenue: Obs 788; Mean 19.4; Std. Dev. 14.9; Min 6.9; Max 25.2
  - Tax: Obs 289; Mean 13.7; Std. Dev. 4.1; Min 5.1; Max 20.9
  - Expenditure: Obs 816; Mean 21.0; Std. Dev. 19.6; Min 12.2; Max 29.2
  - Social expenditure: Obs 648; Mean 5.4; Std. Dev. 4.3; Min 5.0; Max 8.5
  - Fiscal balance: Obs 917; Mean -3.9; Std. Dev. 16.5; Min -14.3; Max 2.6
  - Debt: Obs 405; Mean 42.5; Std. Dev. 35.1; Min 11.0; Max 95.0
- Change in ratio to GDP:
  - Revenue: Obs 733; Mean 0.1; Std. Dev. 4.1; Min -4.5; Max 1.4
  - Tax: Obs 255; Mean 0.2; Std. Dev. 1.7; Min -3.3; Max 1.4
  - Expenditure: Obs 761; Mean 0.3; Std. Dev. 8.4; Min -1.5; Max 4.1
  - Social expenditure: Obs 595; Mean 0.2; Std. Dev. 1.4; Min -1.7; Max 1.7
  - Fiscal balance: Obs 864; Mean -0.1; Std. Dev. 4.1; Min -5.0; Max 1.4
  - Debt: Obs 357; Mean 1.2; Std. Dev. 3.3; Min -2.0; Max 4.6

### APPENDIX IV: ADDITIONAL RESULTS — TABLE 15. BASIS RISK: CROSS-COUNTRY STATIONARY PANEL REGRESSION ANALYSIS
- Dependent Variable: Log Country-Specific Commodity Prices. Independent Variable: Log World Commodity Prices.
- Method notes:
  - For each commodity, a panel unit-root test is carried out on country-specific commodity prices.
  - Where the unit-root null is rejected, a panel regression of country-specific prices on the benchmark world price is estimated, controlling for country fixed effects.
  - Where the unit-root null is not rejected (denoted by * after the commodity name), a cointegrating relationship is estimated using the Pedroni (2000) heterogeneous-panel FM-OLS estimator.
- Selected commodity results (β, SE(β), R2, Countries, Observations):
  - Bananas: 0.37; 0.06; 0.64; 82; 1,399
  - Barley: 0.64; 0.06; 0.59; 86; 1,499
  - Beef *: 0.56; 0.16; NA; 21; 456
  - Chicken: 0.47; 0.08; 0.61; 24; 429
  - Cocoa: 0.70; 0.04; 0.51; 84; 1,454
  - Coconut Oil: 0.54; 0.06; 0.43; 41; 719
  - Coffee: 0.54; 0.03; 0.64; 111; 1,899
  - Corn: 0.53; 0.07; 0.53; 99; 1,663
  - Fish: 0.24; 0.11; 0.70; 135; 2,238
  - Fish Meal: 0.53; 0.05; 0.31; 71; 1,201
  - Ground Nuts: 0.59; 0.07; 0.59; 64; 1,089
  - Olive Oil: 0.59; 0.08; 0.34; 59; 1,053
  - Orange: 0.16; 0.05; 0.34; 102; 1,812
  - Pork: 0.12; 0.14; 0.39; 12; 214
  - Rapeseed Oil: 0.56; 0.06; 0.42; 53; 892
  - Rice: 0.55; 0.05; 0.39; 106; 1,791
  - Shrimp: 0.10; 0.05; 0.47; 95; 1,635
  - Soybean Meal: 0.65; 0.07; 0.35; 68; 1,072
  - Soybean Oil: 0.68; 0.05; 0.37; 87; 1,436
  - Soybeans: 0.72; 0.07; 0.49; 65; 1,139
  - Sugar: 0.47; 0.21; 0.22; 117; 1,956
  - Sunflower Oil: 0.55; 0.04; 0.39; 83; 1,396
  - Tea: 0.35; 0.10; 0.58; 99; 1,720
  - Wheat: 0.73; 0.05; 0.42; 124; 2,048
  - Cotton: 0.69; 0.08; 0.39; 109; 1,836
  - Hard Log: 1.90; 0.36; 0.2; 85; 1,243
  - Hard Sawn: 1.56; 0.21; 0.22; 109; 1,820
  - Rubber: 0.63; 0.04; 0.45; 69; 1,206
  - Soft Log: -2.37; 0.59; 0.19; 61; 930
  - Soft Sawn: 5.60; 0.37; 0.31; 94; 1,599
  - Wool: 0.36; 0.05; 0.59; 68; 1,225
  - Aluminum: 0.71; 0.06; 0.64; 117; 2,018
  - Copper: 0.74; 0.03; 0.55; 109; 1,852
  - Gold: 1.13; 0.11; 0.39; 107; 1,622
  - Iron: 0.32; 0.06; 0.44; 78; 1,309
  - Lead: 0.75; 0.03; 0.4; 96; 1,578
  - Nickel: 0.56; 0.06; 0.59; 61; 1,114
  - Tin: 0.93; 0.06; 0.39; 60; 1,093
  - Zinc: 0.70; 0.05; 0.44; 83; 1,450
  - Coal: 0.68; 0.06; 0.43; 72; 1,536
  - Crude oil: 0.91; 0.03; 0.53; 90; 2,200

*Source: APPENDIX III and APPENDIX IV content as provided in the source PDF.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12112.pdf_