## _wp0532

## Source details

**Canonical URL:** [_wp0532](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0532.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0532.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0532.pdf.json)

---

### I. Introduction — purpose and key hypothesis
- Presents a simultaneous equations model (SEM) of the world crude oil and natural gas markets.
- Key empirical findings:
  - Price elasticities are low; income elasticities are always significant and high.
  - Extreme inelasticity of demand and supply implies small excess demand or supply may require large price movements to clear markets.
  - Both interest rates and the U.S. dollar nominal effective exchange rate (NEER) influence crude prices inversely: rising interest rates and an appreciating NEER tend to depress crude prices; declining interest rates and a depreciating NEER have the opposite effect.
- Policy implication stated: coordination of monetary policy should explicitly include crude price volatility in the inflation target.

### II. Basic data (1918–2004) — production and price patterns
- Crude oil production:
  - 2004: 84 million barrels a day (mbd).
  - Average growth rates: 1.3 percent per year during 1974-2004; 7.2 percent per year during 1918-73.
  - OPEC output: 34 mbd in 2004 (above 31 mbd peak in 1973).
  - Non-OPEC production: 50 mbd in 2004; growth rates 2.2 percent a year during 1974-2004 and 5.8 percent during 1918-73.
- Natural gas production:
  - 2004: 2,719 billion cubic meters a year (bcm/y); 1973: 1,233 bcm/y.
  - Average growth rates: 7.3 percent per year during 1918-73; 2.6 percent per year during 1974-2004.
- Crude oil prices:
  - Nominal price rose to $36.7/bl in 1981 and exceeded $55/bl in 2004.
  - Nominal price falls to $9-$10/bl in 1986 and 1999 noted.
  - 2000-2004: demand shock, short-term fixed supply capacity, output increased by only 2 percent a year during 2000-2004.
- Natural gas prices:
  - Increased at 13.3 percent a year during 1974-2004 to an all-time peak of 521 cents/tcf in 2004.
  - Real prices rose at 8.1 percent a year during 1974-2004 compared with a decline at 0.3 percent a year during 1918-73.
  - Relative price of natural gas to crude oil appreciated at 2 percent in 1918-73 and 6.1 percent in 1974-2004.
- Notes on measurement:
  - Nominal oil prices measured in U.S. dollars per barrel; real prices in constant 1950 prices (deflator: consumer price index for the seven industrial countries).
  - Natural gas prices measured in U.S. cents per thousand cubic feet (tcf) at U.S. wellheads; real prices in 1950 prices.

### III. SEM specification and identification (model structure)
- Endogenous variables (logarithms):
  - y1 = crude oil output (mbd)
  - y2 = crude oil nominal price (US$/barrel)
  - y3 = natural gas output (bcm/y)
  - y4 = natural gas nominal price (US cents per cubic feet)
- Structural equations (as specified in text):
  - Crude oil demand includes γ11 z1, γ14 z4, δ12 y2.
  - Crude oil supply includes expected nominal price e y2, natural gas output y3, price-shock dummy z2.
  - Natural gas demand includes γ31 z1, γ34 y4.
  - Natural gas supply includes expected price e y4, crude oil output y1, price-shock dummy z3.
- Expectation formulation:
  - e y2 = y2 + η2; e y4 = y4 + η4 (forecast errors η2 and η4 uncorrelated with the information set).
- Identification and estimation:
  - Model identified; lag structure optimal at three or four periods.
  - Estimated by two-stage least squares for short-run estimates; reestimated in an error correction model (ECM) for robustness.
  - Long-run elasticities estimated via ECM and cointegration analysis.

### IV. Estimation results — short-run elasticities and interpretation
- Main interpretation: both demand and supply responses to price in the short run are extremely inelastic; income (real GDP of the seven industrial countries) is a significant driver.
- Short-run elasticities (two-stage least squares), periods: 1918-2004; 1918-73; 1974-2004 (exact values and t-ratios)
  - Crude oil demand
    - Price elasticity δ12: -0.05 (-2.18); -0.05 (-0.94); -0.003 ( -0.13)
    - Income elasticity γ11: 0.54 ( 5.03); 0.43 (3.43); 1.49 ( 3.23)
  - Crude oil supply
    - Price elasticity δ22: -0.04 (-1.97); -0.01 (-0.33); -0.05 (-2.11)
    - Gas output elasticity δ23: 0.38 (4.15); 0.42 (4.24); 1.04 (3.58)
  - Natural gas demand
    - Price elasticity δ34: -0.01 (-0.21); -0.17 (-0.78); -0.01 (-0.38)
    - Income elasticity γ31: 0.77 ( 7.06); 0.70 (4.63); 0.80 (2.94)
  - Natural gas supply
    - Price elasticity δ44: 0.09 (0.82); -0.32 (-1.15); 0.05 (1.49)
    - Oil output elasticity δ41: 0.88 (3.73); 1.04 (3.60); 0.35 (3.24)
- Short-run elasticities in the ECM (exact values and t-ratios)
  - Crude oil demand
    - Price elasticity δ12: -0.02 (-0.78); 0.04 (1.03); -0.01 (-0.64)
    - Income elasticity γ11: 0.36 (3.60); 0.30 (2.92); 1.41 (4.30)
  - Crude oil supply
    - Price elasticity δ22: -0.02 (-1.13); -0.01 (-0.32); -0.03 (-1.31)
    - Gas output elasticity δ23: 0.41 (4.52); 0.46 (4.74); 0.84 (3.03)
  - Natural gas demand
    - Price elasticity δ34: 0.006 (0.17); 0.0004 (0.003); 0.008 (0.46)
    - Income elasticity γ31: 0.73 (7.96); 0.74 (6.48); 0.85 (3.83)
  - Natural gas supply
    - Price elasticity δ44: -0.03 (-0.63); -0.05 (-0.34); 0.01 (0.77)
    - Oil output elasticity δ41: 0.53 (4.95); 0.62 (4.00); 0.33 (3.04)
- Key quantitative interpretations:
  - Crude oil demand: short-run price elasticity very low (e.g., -0.05 for 1918-2004); income elasticity rises substantially in 1974-2004 to 1.49 (two-stage least squares).
  - Crude oil supply: short-run price elasticity low and sometimes negative; natural gas output has a significant positive short-run elasticity on oil supply (e.g., δ23 = 0.38 for 1918-2004; 1.04 in 1974-2004).
  - Natural gas demand: short-run price elasticity near zero and nonsignificant; income elasticities significant (e.g., 0.77 for 1918-2004).
  - Natural gas supply: short-run price elasticities low and nonsignificant; crude oil output significantly affects natural gas supply (e.g., δ41 = 0.88 for 1918-2004).
- Mechanisms:
  - Inelastic short-run demand and supply reflect fixed capital, existing contracts, capacity constraints, OPEC quotas, and strategic producer behavior.
  - Large income elasticities in 1974-2004 help explain oil price volatility following world growth swings.
  - Strong interdependence between crude oil and natural gas production: correlation coefficient between crude oil and natural gas production over 1918-2004 was 0.996.

### V. Long-run elasticities (ECM and cointegration results)
- Vector error correction method — Crude oil (exact values and t-ratios)
  - Demand price elasticity: -1.59 (-4.02) for 1918-2004; -2.73 (-3.52) for 1918-73; -0.12 (-1.86) for 1974-2004.
  - Demand income elasticity: 3.48 (5.60) for 1918-2004; 3.43 (5.46) for 1918-73; 0.62 (7.47) for 1974-2004.
  - Supply price elasticity: 0.16 (2.46) for 1918-2004; 0.44 (3.50) for 1918-73; 0.23 (3.96) for 1974-2004.
  - Supply natural gas elasticity: 0.75 (13.21) for 1918-2004; 0.72 (13.27) for 1918-73; 0.24 (2.67) for 1974-2004.
- Vector error correction method — Natural gas (exact values and t-ratios)
  - Demand price elasticity: -0.45 (-4.742) for 1918-2004; -2.79 (-3.50) for 1918-73; -0.06 (-1.67) for 1974-2004.
  - Demand income elasticity: 2.35 (14.49) for 1918-2004; 4.37 (6.69) for 1918-73; 0.96 (9.66) for 1974-2004.
  - Supply price elasticity: 0.16 (5.51) for 1918-2004; 0.16 (1.42) for 1918-73; 0.21 (6.51) for 1974-2004.
  - Supply crude oil elasticity: 1.00 (30.63) for 1918-2004; 0.98 (18.31) for 1918-73; 0.72 (4.46) for 1974-2004.
- Cointegration method — Crude oil (exact values and t-ratios)
  - Demand long-run price elasticity: -0.27 (-1.89) for 1918-2004; -0.32 (-4.99) for 1918-73; -0.26 (-3.09) for 1974-2004.
  - Demand long-run income elasticity: 1.53 (6.74) for 1918-2004; 1.93 (21.44) for 1918-73; 0.80 (5.65) for 1974-2004.
  - Supply long-run price elasticity: 0.12 (2.39) for 1918-2004; 0.46 (46.0) for 1918-73; 0.25 (3.79) for 1974-2004.
  - Supply long-run natural gas elasticity: 0.78 (16.98) for 1918-2004; 0.73 (16.59) for 1918-73; 0.45 (18.0) for 1974-2004.
- Cointegration method — Natural gas (exact values and t-ratios)
  - Demand long-run price elasticity: -0.48 (-2.40) for 1918-2004; -0.37 (-2.00) for 1918-73; -0.01 (-1.11) for 1974-2004.
  - Demand long-run income elasticity: 1.78 (5.17) for 1918-2004; 2.35 (16.71) for 1918-73; 0.88 (23.73) for 1974-2004.
  - Supply long-run price elasticity: 0.34 (5.50) for 1918-2004; 0.26 (1.73) for 1918-73; 0.19 (7.12) for 1974-2004.
  - Supply long-run crude oil elasticity: 0.83 (11.52) for 1918-2004; 0.89 (12.33) for 1918-73; 0.74 (4.77) for 1974-2004.
- Key long-run interpretations:
  - Long-run crude oil demand price elasticities are significant and indicate a pronounced drop in demand elasticity in 1974-2004 relative to 1918-73.
  - Long-run income elasticities for crude oil: 1.53 (1918-2004), 1.93 (1918-73), 0.80 (1974-2004) — showing a structural fall after 1973.
  - Long-run crude oil supply price elasticity declined from 0.46 (1918-73) to 0.25 (1974-2004), consistent with a shift from competitive to market-maker structure.
  - Natural gas long-run price elasticity fell to -0.01 in 1974-2004 from -0.37 in 1918-73; income elasticity fell to 0.88 in 1974-2004 from 2.35 in 1918-73.
  - Stability of crude oil elasticity affecting natural gas (0.83 over 1918-2004) indicates technical interdependence of discoveries and development between crude oil and natural gas.

### VI. Monetary policy, exchange rates, and oil prices
- Evidence and model extensions:
  - Relations between crude price, the NEER, and interest rates examined using monthly, quarterly, and annual data for 1974-2004.
  - Rising interest rates and an appreciating NEER tend to depress crude prices; declining interest rates and a depreciating NEER tend to raise crude prices.
  - VAR analyses do not reject at least one cointegration relation between crude oil prices, dollar NEER, and U.S. interest rates.
  - Interest rate shocks generally affect oil prices negatively and significantly; NEER shocks have significantly negative effects during periods of large movements.
  - Impulse responses: crude prices fall in response to rising interest rates and NEER appreciation, and rise in response to falling interest rates and NEER depreciation.
- Observations 2000-04:
  - Period witnessed low interest rates and a depreciating U.S. dollar.
  - NEER of the dollar depreciated by 13 percent over 2000-04.
  - Low interest rates stimulate aggregate demand and energy demand; dollar depreciation makes oil imports cheaper in local currencies and may prompt exporters to tighten supply to seek higher dollar prices.
- Policy implication:
  - Vulnerability to monetary shocks suggests coordination of monetary policy among industrial countries should account for oil market volatility.
  - A Taylor-like rule should explicitly include crude price volatility in the inflation target.

### VII. SEM extension (role of the NEER), model fit, and multipliers
- SEM extended to incorporate NEER (reestimated for 1974-2004).
- Short-Run and Long-Run elasticities (SEM extended to include NEER) — exact values and t-ratios:
  - Crude oil demand
    - Price elasticity: -0.004 (-0.20) (Short-Run); -0.17 (-1.56) (Long-Run)
    - Income elasticity: 1.43 (4.07) (Short-Run); 0.63 (4.90) (Long-Run)
    - NEER elasticity: -0.08 (-2.21) (Short-Run); -0.26 (-1.34) (Long-Run)
  - Crude oil supply
    - Price elasticity: -0.03 (-1.13) (Short-Run); -0.08 (-0.88) (Long-Run)
    - Natural gas elasticity: 1.60 (4.72) (Short-Run); 1.09 (12.46) (Long-Run)
    - NEER elasticity: -0.12 (-2.28) (Short-Run); -0.90 (-3.91) (Long-Run)
  - Natural gas demand
    - Price elasticity: 0.10 (3.58) (Short-Run); -0.01 (-2.08) (Long-Run)
    - Income elasticity: 0.61 (3.29) (Short-Run); 0.85 (43.55) (Long-Run)
    - NEER elasticity: -0.12 (-4.02) (Short-Run); -0.14 ( -4.36) (Long-Run)
  - Natural gas supply
    - Price elasticity: 0.06 (1.85) (Short-Run); 0.10 (4.16) (Long-Run)
    - Crude oil elasticity: 0.41 (5.33) (Short-Run); 1.29 (9.37) (Long-Run)
    - NEER elasticity: 0.05 (2.34) (Short-Run); 0.21 (1.61) (Long-Run)
- Model fit and residuals:
  - Residual standard errors: 0.009 for crude oil output, 0.03 for crude oil price, 0.006 for natural gas output, and 0.02 for natural gas price.
  - Model fits quantities closely; price prediction less precise, particularly crude oil price during 1974-2004.
- Multipliers (long-run effects of a 1 unit percent change in exogenous variables):
  - A 1 percent increase in income leads to:
    - crude oil output +0.49 percent,
    - crude oil price +1.25 percent,
    - natural gas output +1.0 percent,
    - natural gas price +2.53 percent.
  - For the NEER: a 1 percent depreciation leads to an increase by 0.62 percent in the crude oil price (additional NEER multipliers reported in text).

### VIII. Policy recommendations and implications
- Monetary policy coordination:
  - Because crude oil and natural gas markets are highly vulnerable to monetary shocks (via interest rates and NEER), monetary policy coordination should explicitly include crude price volatility in the inflation target.
- Taxation and demand management:
  - High short-run price inelasticities imply small shifts in supply or demand can produce large price movements.
  - Taxation may be used in oil-importing countries to curb petroleum demand and appropriate part of producer surplus; inelastic demand may require high tax rates to influence quantities demanded.
- Structural change recognition:
  - Recognize the role of structural changes post-1973 (oil shocks) in raising income sensitivity and altering supply-demand dynamics.
- Model applications:
  - SEM can simulate dynamic impacts of exogenous changes, impulse responses to shocks, perform sensitivity analysis on structural parameters, and provide forecasts of prices given stability of elasticity estimates.

*Source: _wp0532 - References..............................................................................................................*

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

### _wp0532 - References..............................................................................................................

### I. Introduction — purpose and key hypothesis
- Presents a simultaneous equations model (SEM) of the world crude oil and natural gas markets.
- Key empirical findings summarized:
  - Price elasticities are low; income elasticities are always significant and high.
  - Extreme inelasticity of demand and supply implies small excess demand or supply may require large price movements to clear markets.
  - Both interest rates and the U.S. dollar nominal effective exchange rate (NEER) influence crude prices inversely: rising interest rates and an appreciating NEER tend to depress crude prices; declining interest rates and a depreciating NEER have the opposite effect.
- Policy implication stated: coordination of monetary policy should explicitly include crude price volatility in the inflation target.

### II. Basic data (1918–2004) — production and price patterns
- Crude oil production:
  - 2004: 84 million barrels a day (mbd).
  - Average growth rates: 1.3 percent per year during 1974-2004; 7.2 percent per year during 1918-73.
  - OPEC output: 34 mbd in 2004 (above 31 mbd peak in 1973).
  - Non-OPEC production: 50 mbd in 2004; growth rates 2.2 percent a year during 1974-2004 and 5.8 percent during 1918-73.
- Natural gas production:
  - 2004: 2,719 billion cubic meters a year (bcm/y); 1973: 1,233 bcm/y.
  - Average growth rates: 7.3 percent per year during 1918-73; 2.6 percent per year during 1974-2004.
- Crude oil prices:
  - Nominal price rose to $36.7/bl in 1981 and exceeded $55/bl in 2004.
  - Nominal price falls to $9-$10/bl in 1986 and 1999 noted.
  - 2000-2004: demand shock, short-term fixed supply capacity, output increased by only 2 percent a year during 2000-2004.
- Natural gas prices:
  - Increased at 13.3 percent a year during 1974-2004 to an all-time peak of 521 cents/tcf in 2004.
  - Real prices rose at 8.1 percent a year during 1974-2004 compared with a decline at 0.3 percent a year during 1918-73.
  - Relative price of natural gas to crude oil appreciated at 2 percent in 1918-73 and 6.1 percent in 1974-2004.
- Notes on measurement:
  - Nominal oil prices measured in U.S. dollars per barrel; real prices in constant 1950 prices (deflator: consumer price index for the seven industrial countries).
  - Natural gas prices measured in U.S. cents per thousand cubic feet (tcf) at U.S. wellheads; real prices in 1950 prices.

### III. SEM specification and identification (model structure)
- Four equations (endogenous variables in logarithms):
  - y1 = crude oil output (mbd)
  - y2 = crude oil nominal price (US$/barrel)
  - y3 = natural gas output (bcm/y)
  - y4 = natural gas nominal price (US cents per cubic feet)
- Equations (as specified):
  - Crude oil demand: y1 = γ11 z1 + γ14 z4 + δ12 y2 + ... (equation (1) as presented)
  - Crude oil supply: includes expected nominal price e y2, natural gas output y3, price-shock dummy z2 (equation (2))
  - Natural gas demand: y3 = γ31 z1 + γ34 y4 + ... (equation (3))
  - Natural gas supply: includes expected price e y4, crude oil output y1, price-shock dummy z3 (equation (4))
- Expectation formulation:
  - e y2 = y2 + η2; e y4 = y4 + η4 (forecast errors η2 and η4 uncorrelated with the information set)
- Identification and estimation approach:
  - Model is identified; lag structure optimal at three or four periods.
  - Estimated by two-stage least squares for short-run estimates; reestimated in an error correction model (ECM) for robustness.
  - Long-run elasticities estimated via ECM and cointegration analysis.

### IV. Estimation results — short-run elasticities and interpretation
- Main interpretation: both demand and supply responses to price in the short run are extremely inelastic; income (real GDP of the seven industrial countries) is a significant driver.

- Short-run elasticities (two-stage least squares), exact values and t-ratios (t-ratio in parenthesis) — Periods: 1918-2004; 1918-73; 1974-2004

  - Crude oil demand
    - Price elasticity δ12: -0.05 (-2.18); -0.05 (-0.94); -0.003 ( -0.13)
    - Income elasticity γ11: 0.54  ( 5.03); 0.43 (3.43); 1.49  ( 3.23)

  - Crude oil supply
    - Price elasticity δ22: -0.04 (-1.97); -0.01 (-0.33); -0.05 (-2.11)
    - Gas output elasticity δ23: 0.38 (4.15); 0.42  (4.24); 1.04 (3.58)

  - Natural gas demand
    - Price elasticity δ34: -0.01 (-0.21); -0.17 (-0.78); -0.01 (-0.38)
    - Income elasticity γ31: 0.77 ( 7.06); 0.70  (4.63); 0.80  (2.94)

  - Natural gas supply
    - Price elasticity δ44: 0.09 (0.82); -0.32  (-1.15); 0.05 (1.49)
    - Oil output elasticity δ41: 0.88 (3.73); 1.04  (3.60); 0.35 (3.24)

- Short-run elasticities in the ECM (exact values and t-ratios)

  - Crude oil demand
    - Price elasticity δ12: -0.02 (-0.78); 0.04 (1.03); -0.01 (-0.64)
    - Income elasticity γ11: 0.36 (3.60); 0.30 (2.92); 1.41 (4.30)

  - Crude oil supply
    - Price elasticity δ22: -0.02 (-1.13); -0.01 (-0.32); -0.03 (-1.31)
    - Gas output elasticity δ23: 0.41 (4.52); 0.46 (4.74); 0.84 (3.03)

  - Natural gas demand
    - Price elasticity δ34: 0.006 (0.17); 0.0004  (0.003); 0.008 (0.46)
    - Income elasticity γ31: 0.73 (7.96); 0.74 (6.48); 0.85  (3.83)

  - Natural gas supply
    - Price elasticity δ44: -0.03 (-0.63); -0.05 (-0.34); 0.01 (0.77)
    - Oil output elasticity δ41: 0.53 (4.95); 0.62 (4.00); 0.33 (3.04)

- Key quantitative interpretations:
  - Crude oil demand: short-run price elasticity is very low (e.g., -0.05 for 1918-2004) and unchanged across subperiods; income elasticity rises substantially in 1974-2004 to 1.49 (two-stage least squares).
  - Crude oil supply: short-run price elasticity low and sometimes negative (e.g., -0.04 for 1918-2004); natural gas output has a significant positive short-run elasticity on oil supply (e.g., δ23 = 0.38 for 1918-2004; 1.04 in 1974-2004).
  - Natural gas demand: short-run price elasticity near zero and nonsignificant (e.g., -0.01 for 1918-2004); income elasticities significant (e.g., 0.77 for 1918-2004).
  - Natural gas supply: short-run price elasticities low and nonsignificant (e.g., 0.09 for 1918-2004); crude oil output significantly affects natural gas supply (e.g., δ41 = 0.88 for 1918-2004).

- Economic mechanisms and interpretations highlighted in text:
  - Inelastic short-run demand and supply reflect fixed capital, existing contracts, capacity constraints, OPEC quotas, and strategic producer behavior (e.g., withholding output to preserve price gains).
  - Large income elasticities in 1974-2004 help explain oil price volatility following world growth swings.
  - Strong interdependence between crude oil and natural gas production: correlation coefficient between crude oil and natural gas production over 1918-2004 was 0.996.

### V. Monetary policy, exchange rates, and oil prices
- Evidence and model extensions:
  - Section V examines relations between crude price, the NEER, and interest rates using monthly, quarterly, and annual data for 1974-2004.
  - Rising interest rates and an appreciating NEER tend to depress crude prices; declining interest rates and a depreciating NEER tend to raise crude prices.
  - Section VI extends the SEM by including the NEER as an additional exogenous variable and finds the NEER has a significant impact on market equilibrium; the SEM solution closely tracks actual data.

### VI. Policy recommendations and implications (as stated)
- Because crude oil and natural gas markets are highly vulnerable to monetary shocks (via interest rates and NEER), monetary policy coordination should explicitly include crude price volatility in the inflation target.
- High short-run price inelasticities imply:
  - Small shifts in supply or demand can produce large price movements.
  - Taxation may be used in oil-importing countries to curb petroleum demand and appropriate part of producer surplus; inelastic demand may require high tax rates to influence quantities demanded.
- Recognize the role of structural changes post-1973 (oil shocks) in raising income sensitivity and altering supply-demand dynamics.

*Source: _wp0532 - References..............................................................................................................*

### 0.88 in 1918-2004, 1.04 in 1918-73, and 0.35 in 1974-2004, implying interdependence in oil

### _wp0532 - 0.88 in 1918-2004, 1.04 in 1918-73, and 0.35 in 1974-2004, implying interdependence in oil

### Short-run elasticities and dynamics
- Crude oil:
  - Demand has low short-run price elasticities (no single short-run price numeric given in excerpt).
  - Income elasticity is high with a noticeable increase in 1974-2004 (specific short-run income levels not restated in excerpt).
  - Supply has a low short-run price elasticity.
- Natural gas:
  - Output elasticity increased to 0.84 in 1974-2004 from 0.46 in 1918-73, indicating a stronger relation between crude oil and natural gas.
  - Demand function in the ECM had short-run price elasticities that were low and insignificant.
  - Income elasticities in short-run ECM: 0.73 in 1918-2004, 0.74 in 1918-73, and 0.85 in 1974-2004.
  - Supply function in ECM had no significant short-run price elasticity.
  - Natural gas supply was significantly influenced by crude oil output with elasticities: 0.53 in 1918-2004, 0.62 in 1918-73, and 0.33 in 1974-2004, implying strong interdependence between crude oil and natural gas production.
- Interpretation:
  - Negative short-run price elasticity of natural gas supply in 1918-2004 is linked to demand-side effects and producer pricing/quantity strategies.
  - Prices played a weak role in influencing short-run supply for natural gas.

### Long-run demand and supply elasticities (ECM and cointegration results summarized)
- Crude oil (Vector error correction method):
  - Demand price elasticity: -1.59 (-4.02) for 1918-2004; -2.73 (-3.52) for 1918-73; -0.12 (-1.86) for 1974-2004.
  - Demand income elasticity: 3.48 (5.60) for 1918-2004; 3.43 (5.46) for 1918-73; 0.62 (7.47) for 1974-2004.
  - Supply price elasticity: 0.16 (2.46) for 1918-2004; 0.44 (3.50) for 1918-73; 0.23 (3.96) for 1974-2004.
  - Supply natural gas elasticity: 0.75 (13.21) for 1918-2004; 0.72 (13.27) for 1918-73; 0.24 (2.67) for 1974-2004.
- Natural gas (Vector error correction method):
  - Demand price elasticity: -0.45 (-4.742) for 1918-2004; -2.79 (-3.50) for 1918-73; -0.06 (-1.67) for 1974-2004.
  - Demand income elasticity: 2.35 (14.49) for 1918-2004; 4.37 (6.69) for 1918-73; 0.96 (9.66) for 1974-2004.
  - Supply price elasticity: 0.16 (5.51) for 1918-2004; 0.16 (1.42) for 1918-73; 0.21 (6.51) for 1974-2004.
  - Supply crude oil elasticity: 1.00 (30.63) for 1918-2004; 0.98 (18.31) for 1918-73; 0.72 (4.46) for 1974-2004.
- Crude oil (Cointegration method):
  - Demand long-run price elasticity: -0.27 (-1.89) for 1918-2004; -0.32 (-4.99) for 1918-73; -0.26 (-3.09) for 1974-2004.
  - Demand long-run income elasticity: 1.53 (6.74) for 1918-2004; 1.93 (21.44) for 1918-73; 0.80 (5.65) for 1974-2004.
  - Supply long-run price elasticity: 0.12 (2.39) for 1918-2004; 0.46 (46.0) for 1918-73; 0.25 (3.79) for 1974-2004.
  - Supply long-run natural gas elasticity: 0.78 (16.98) for 1918-2004; 0.73 (16.59) for 1918-73; 0.45 (18.0) for 1974-2004.
- Natural gas (Cointegration method):
  - Demand long-run price elasticity: -0.48 (-2.40) for 1918-2004; -0.37 (-2.00) for 1918-73; -0.01 (-1.11) for 1974-2004.
  - Demand long-run income elasticity: 1.78 (5.17) for 1918-2004; 2.35 (16.71) for 1918-73; 0.88 (23.73) for 1974-2004.
  - Supply long-run price elasticity: 0.34 (5.50) for 1918-2004; 0.26 (1.73) for 1918-73; 0.19 (7.12) for 1974-2004.
  - Supply long-run crude oil elasticity: 0.83 (11.52) for 1918-2004; 0.89 (12.33) for 1918-73; 0.74 (4.77) for 1974-2004.
- Key long-run interpretations:
  - Long-run crude oil demand price elasticities are significant and indicate a pronounced drop in demand elasticity in 1974-2004 relative to 1918-73.
  - Long-run income elasticities for crude oil: 1.53 (1918-2004), 1.93 (1918-73), 0.80 (1974-2004) — showing a structural fall after 1973.
  - Long-run crude oil supply price elasticity declined from 0.46 (1918-73) to 0.25 (1974-2004), consistent with a shift from competitive to market-maker structure.
  - Long-run natural gas output elasticity fell to 0.45 (1973-2004) from 0.73 (1918-73), indicating oil producers restraining supply in later period.
  - Natural gas long-run price elasticity fell dramatically to -0.01 in 1974-2004 from -0.37 in 1918-73; income elasticity fell to 0.88 in 1974-2004 from 2.35 in 1918-73.
  - Stability of crude oil elasticity affecting natural gas (0.83 over 1918-2004) indicates technical interdependence of discoveries and development between crude oil and natural gas.

### Comparison with other studies and implications for market equilibrium
- Agreement with literature:
  - Findings align with empirical research showing low demand price elasticities in both short and long run.
  - Income elasticities generally close to or higher than unity in many studies; this paper finds similar patterns historically.
- Equilibrium and price volatility implications:
  - Price inelasticity of both demand and supply explains high volatility in equilibrium prices.
  - Short-run: a reduction in crude oil supply likely causes a large short-run increase in crude oil prices; an increase in supply likely causes a large short-run drop in prices.
  - Supply price inelasticity implies that demand shocks (slower world economy, warmer winter) cause sharp price falls; demand increases cause large price rises.
  - Interaction between crude oil and natural gas supplies: an increase in natural gas supply likely accompanies increased crude oil supply and a significant fall in crude oil prices; reduced natural gas supply likely reduces crude oil supply and causes a significant increase in crude oil prices.

### Relation between oil prices, interest rates, and the U.S. dollar NEER
- Observations 2000-04:
  - Period witnessed low interest rates and a depreciating U.S. dollar.
  - NEER of the dollar depreciated by 13 percent over 2000-04 (additional note: dollar depreciated by 48 percent against the euro from October 2000 through November 2004 is stated in footnote).
  - Low interest rates stimulate aggregate demand and energy demand; dollar depreciation makes oil imports cheaper in local currencies and may prompt exporters to tighten supply to seek higher dollar prices.
- VAR and cointegration findings:
  - VAR analyses do not reject existence of at least one cointegration relation between crude oil prices, dollar NEER, and U.S. interest rates.
  - Interest rate shocks generally affect oil prices negatively and significantly for most sample periods.
  - NEER shocks have significantly negative effects essentially during periods of large movements in interest and exchange rates.
  - Impulse responses show crude prices fall in response to rising interest rates and NEER appreciation, and rise in response to falling interest rates and NEER depreciation.
  - A positive random shock to crude price increases price on impact but is quickly dampened by negative effects from interest rate and NEER shocks.
- Policy implication:
  - Vulnerability to monetary shocks suggests coordination of monetary policy among industrial countries should account for oil market volatility.
  - A Taylor-like rule should explicitly include crude price volatility in the inflation target.

### SEM extension and solution (role of NEER and model performance)
- Role of NEER:
  - SEM extended to incorporate NEER as an exogenous variable and reestimated for 1974-2004.
  - An appreciation of the NEER has a significant negative short-term impact on crude oil price and on demand and supply for crude oil; it also has a negative short-term effect on demand for natural gas.
  - Long-run effect of NEER is generally negative on both demand and supply of crude oil and on demand for natural gas.
  - Interpretation: an appreciation of NEER would most likely depress oil and natural gas markets; depreciation would have opposite effect.
- Solution and fit:
  - SEM written as ttt ByCz u+= with reduced form 1ˆˆˆ tt yBCz −=−.
  - Residual standard errors: 0.009 for crude oil output, 0.03 for crude oil price, 0.006 for natural gas output, and 0.02 for natural gas price.
  - Model fits quantities closely; price prediction is less precise, particularly crude oil price during 1974-2004.
  - Multipliers (long-run effects of a 1 unit percent change in exogenous variables):
    - A 1 percent increase in income leads to:
      - crude oil output +0.49 percent,
      - crude oil price +1.25 percent,
      - natural gas output +1.0 percent,
      - natural gas price +2.53 percent.
    - For the NEER: a 1 percent depreciation leads to an increase by 0.62 percent in the crude oil price (excerpt truncates further NEER multipliers).

*Source: IMF working paper content provided in the supplied PDF excerpt.*

### 0.83 percent in the natural gas price.

### _wp0532 - 0.83 percent in the natural gas price.

### Estimated Elasticities (SEM Extended to Include the NEER)
- Table 5: World Crude Oil and Natural Gas: The SEM Extended to Include the NEER
  - Short-Run and Long-Run Elasticities (1/ t-ratio in parenthesis)

  - Crude oil demand
    - Price elasticity: -0.004 (-0.20) (Short-Run); -0.17 (-1.56) (Long-Run)
    - Income elasticity: 1.43 (4.07) (Short-Run); 0.63 (4.90) (Long-Run)
    - NEER elasticity: -0.08 (-2.21) (Short-Run); -0.26 (-1.34) (Long-Run)

  - Crude oil supply
    - Price elasticity: -0.03 (-1.13) (Short-Run); -0.08 (- 0.88) (Long-Run)
    - Natural gas elasticity: 1.60 (4.72) (Short-Run); 1.09 (12.46) (Long-Run)
    - NEER elasticity: -0.12 (-2.28) (Short-Run); -0.90 (-3.91) (Long-Run)

  - Natural gas demand
    - Price elasticity: 0.10 (3.58) (Short-Run); -0.01 (-2.08) (Long-Run)
    - Income elasticity: 0.61 (3.29) (Short-Run); 0.85 (43.55) (Long-Run)
    - NEER elasticity: -0.12 (-4.02) (Short-Run); -0.14 ( -4.36) (Long-Run)

  - Natural gas supply
    - Price elasticity: 0.06 (1.85) (Short-Run); 0.10 (4.16) (Long-Run)
    - Crude oil elasticity: 0.41 (5.33) (Short-Run); 1.29 (9.37) (Long-Run)
    - NEER elasticity: 0.05 (2.34) (Short-Run); 0.21 (1.61) (Long-Run)

### Model Fit and Residuals
- Figure 4 (Actual and Fitted Values and Residuals of the SEM) presents time series plots for 1980 1990 2000 of:
  - Crude oil output (lq) — actual and fitted, with scaled residual r:lq
  - Crude oil price (lp) — actual and fitted, with scaled residual r:lp
  - Natural gas output (lg) — actual and fitted, with scaled residual r:lg
  - Natural gas price (lp g) — actual and fitted, with scaled residual r:lp g

### Key Conclusions and Findings
- Time span studied: 1918-2004, with particular attention to pre- and post-1973 oil shock market equilibria.
- Demand effects
  - Oil shocks caused demand for crude oil and natural gas to decelerate sharply in 1973-2004.
  - Demand for crude oil underwent a deep structural change in 1973-2004.
  - Upward jumps in oil prices and high energy taxation in oil-importing countries contributed to a significant reduction of demand elasticities.
  - Long-run demand for oil was compressed to a nonelastic region, with demand for liquid fuels increasingly limited to transportation.
  - Income elasticities were high for both crude oil and natural gas demand.

- Supply effects
  - Long-run supply price elasticity for crude oil fell sharply after the oil shock, reflecting a change from a competitive to a market-maker structure.
  - Long-run supply price elasticity for natural gas rose in the vector error correction model, reflecting a flexible supply response to a switch in demand toward natural gas.

- Price inelasticities and volatility
  - Demand and supply for both crude oil and natural gas were highly price inelastic in the short run.
  - Short-run inelasticities led to excessive volatility in oil and natural gas markets.

- Monetary and exchange rate effects
  - Changes in interest rates and in the dollar’s NEER could have a significant impact on oil prices.
  - Falling interest rates and a depreciating NEER could lead to a surge in oil prices, and inversely.
  - Oil and natural gas prices would continue to be under upward pressure so long interest rates keep falling and the dollar’s NEER keeps depreciating.
  - Policy implication: Coordination of monetary policy among industrial countries should include volatility of oil markets separately in the inflation target.

### Model Applications and Policy Uses
- The SEM (structural econometric model) can:
  - Simulate over time the impact of a change in an exogenous variable via dynamic multipliers.
  - Simulate the impact of random shocks to demand and supply of crude oil and natural gas via impulse response functions (requiring an identification scheme to define shocks).
  - Perform sensitivity analysis on one or more structural parameters because the model is estimated in structural form.
  - Provide forecasts of prices; given the stability and reliability of the elasticity estimates, forecasts using SEM could provide realistic and relevant information for market traders.

### Sources of Data Used in the Analysis (Appendix)
- Arthur Andersen/Cambridge Energy Research Associates, Natural Gas Trends, 1985-2003.
- Arthur Andersen/Cambridge Energy Research Associates, World Oil Trends, 1985-2003.
- Cambridge Energy Research Associates: World Oil Watch, various reports, 1995-2003.
- Commodity Research Bureau, Commodity Yearbook, John Wiley & Sons, Inc. 1955, 1965, 1975, 1985, and 1996.
- International Energy Agency: Oil Market Report, various reports, 1995-2004.
- International Energy Agency, Energy Statistics and Balances of Non-OECD Countries, various publications, 1985-2004.
- International Monetary Fund, International Financial Statistics, 1950-2004.
- Mitchell, B. R., International Historical Statistics, the Americas and Europe, 1750-1993, Fourth Edition, Macmillan, 1998.
- Oil and Gas Journal, Worldwide Crude Oil and Gas Production, 2003.
- Oil and Gas Journal, Database, 2003.
- OPEC, Annual Statistical Bulletin, various publications, 1980-2003.
- Petroleum Economist, various publications, 1985-2004.
- United Nations, World Energy Supplies, 1950-1974, New York, 1976.
- U.S. Bureau of Mines, The Petroleum Data Book, 1950.
- U.S. Department of Commerce, Bureau of the Census, Historical Statistics of the United States, 1975.
- U.S. Department of Energy, Twentieth Century Petroleum Statistics, 1990.

*Source: _wp0532 - 0.83 percent in the natural gas price.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp0532.pdf_