320 44610 privatesector P U B L I C P O L I C Y F O R T H E NUMBER NOTE 2008 Oil Price Risks JUNE Robert Bacon and Measuring the Vulnerability of Oil Importers Masami Kojima Oil prices more than tripled between January 2004 and March 2008. PRESIDENCY Robert Bacon The effects can be hard on countries with large net oil imports (rbacon@worldbank.org) VICE is a consultant, and relative to income. This Note sets out a measure of vulnerability to oil Masami Kojima price shocks and breaks it down into its components. That allows (mkojima@worldbank.org) a lead energy specialist, cross-country benchmarking and helps to show how changes in such in the World Bank's Oil, factors as energy efficiency and the real exchange rate can make DEVELOPMENT Gas, and Mining Policy countries more vulnerable or less so. Division. SECTOR One measure of the importance of oil to an econ- The results may highlight countries that are omy is its "vulnerability," defined as the ratio of experiencing atypical changes in vulnerability net imports of crude oil and oil products to GDP. and point to factors causing those changes. The PRIVATE A rise in oil prices both increases the value to the findings could call for more detailed analysis of economy of any domestic oil production and why a country might be departing markedly AND decreases GDP according to the consumption of from the global trend of countries with compa- oil and oil products. For countries that consume rable levels of economic development and more than they produce, a change in the balance resource endowments and what steps could FINANCIAL of these two effects--the value of net oil reduce its vulnerability. imports--is a measure of the adjustment that will have to be made when oil prices rise (in the Measuring vulnerability absence of other, offsetting exogenous shocks). Data for constructing a simplified measure of The adjustment will have to be made by deflating vulnerability are available for a large number of GROUP the economy to restore the balance of payments countries between 1996 and 2006 from the U.S. or running down foreign exchange reserves. Department of Energy's Energy Information BANK The vulnerability of an economy can be Administration (EIA) and from World Bank related to a number of factors through an databases. For the 123 countries that were net oil accounting identity. Such a decomposition importers in 1996 and 2006, the annual average analysis allows a study of how vulnerability to price of three marker crudes--Brent, West WORLD increases in oil prices changes over time and Texas Intermediate, and Dubai Fateh--was mul- how much changes in individual factors con- tiplied by the volume of net oil imports. This THE tribute to the overall changes in vulnerability. underestimates vulnerability for two reasons: all O I L P R I C E R I S K S M E A S U R I N G T H E V U L N E R A B I L I T Y O F O I L I M P O R T E R S refined products other than fuel oil are more 1. The crude oil price in U.S. dollars expensive than crude oil, and the cost of trans- 2. The ratio of the volume of net oil imports to porting oil to the importing country is excluded. domestic oil consumption (import depend- In 1996, when the average price of crude oil ence) was US$20.40 a barrel, 85 percent of countries 3. The ratio of domestic oil consumption to had a vulnerability of less than 5 percent (table total domestic energy consumption (oil 1). By 2006, when the average price was share) US$64.30, only 50 percent did. This dramatic 4. The ratio of total domestic energy consump- change is strongly correlated with the rise in the tion to real GDP in local currency (energy 2 oil price, but it also depends on changes in the intensity) use of imported oil relative to GDP as well as on 5. The ratio of real GDP in local currency to cur- movements in the real exchange rate. rent GDP in local currency In 2007 the average oil price rose by another 6. The ratio of current GDP in local currency to 10 percent. GDP for many countries would also current GDP in U.S. dollars. have risen somewhat, so that vulnerability would This identity is valuable because it separates not have changed greatly in 2007 as a whole. But the movements in vulnerability into components the rapid increase in oil prices in the second half that have direct relevance to understanding of 2007 and the first four months of 2008 to those movements. The last two factors bring in around US$100 a barrel would, if sustained, rep- the movements of domestic prices and the resent a 50 percent increase since 2006. This oil exchange rate, so that together they represent price rise would represent a very large change the effect of the real exchange rate. This identity relative to GDP even if partially offset by growth is then manipulated to arrive at a formula that of GDP and a favorable exchange rate move- allows addition of each of the five factors (with ment against the dollar. Countries with a vul- factors 5 and 6 combined) to give the percentage nerability of greater than 10 percent in 2006 difference in vulnerability between any two years. would likely face a further import shock equiva- To illustrate the use of this identity, attention lent to 5 percent or more of GDP. is focused on a group of countries whose vulner- ability between 1996 and 2006 rose faster than Decomposing changes in vulnerability the nominal oil price, that is, by more than a fac- Changes in vulnerability can be linked to several tor of 3.15. For these countries, where growth in factors, including the exchange rate, through an GDP offset some of the oil price rise, the increase identity that forms the basis for a decomposition in the use of oil was such that the oil import bill analysis that allocates the change in vulnerability relative to GDP rose more than the oil price. to changes in the different factors in the identity Thirty-three countries fall into this category, of (for a detailed explanation, see Ang 2005). which 10 are very small island economies. For 18 The ratio of net oil imports in U.S. dollars to of the larger countries the changes in vulnera- current GDP in U.S. dollars at a moment in time bility are decomposed into components related is equal to the product of six factors: to the factors shown above (table 2; box 1).1 The oil price effect results in the largest increase in vulnerability for the economies Table Net-oil-importing countries by vulnerability, 1996 and 2006 where the initial ratio of oil imports to GDP was 1 Vulnerabilitya 1996 2006 highest. For five countries this effect was equal Less than 2.5% 79 19 to or greater than 5 percent of GDP. 2.5­5% 26 42 In five countries that are oil producers, oil 5­7.5% 8 23 production did not keep up with oil consump- 7.5­10% 6 18 tion, with the result that the volume of oil 10­15% 3 12 imports rose faster than GDP. This was particu- >15% 1 9 larly important for Albania, Australia, and a. Value of net oil imports/GDP. China, where half or more of the increase in vul- Source: Authors' calculations. nerability was due to rising import dependence. This effect may have been unavoidable because this increase and whether it suggests a need for of the nature of oil reserves in these countries. active policies to reduce energy use. In 10 countries the share of oil in total energy In nine countries the real exchange rate use increased, while in 6 it decreased. Further moved to offset some of the impact of the dollar analysis could indicate whether the oil share price rise in imports. But in seven countries the increased because of relative growth in the movement worsened oil vulnerability. In transport sector where there were no substitute Burundi, The Gambia, and Malawi this effect fuels or because oil increasingly became a fuel was equivalent to another 1 percent of GDP. The of choice for the power sector. real exchange rate effect reflects the impact of 3 Aggregate energy intensity is determined by a number of domestic policies and does not rep- three broad factors: the income elasticities of resent just a strengthening or weakening of the demand for energy at the sector level, the shares of U.S. dollar. Although exchange rates against the GDP of sectors with different levels of energy inten- dollar have been changing recently, in 2004­06 sity, and the levels of energy efficiency in each sec- Madagascar and Malawi experienced substan- tor.Energyintensityfellinsevencountriesbetween tial nominal depreciation against the dollar, 1996 and 2006, reflecting increasing energy effi- and Uruguay was the only country in the sample ciency, an increasing share of less energy-intensive to experience a substantial appreciation. sectors in GDP, or income elasticities of demand The oil share and energy intensity factors for energy less than unity. Energy intensity rose in reflect country choices about the use of energy. eight countries; in Guyana and Togo it increased In Benin, Ethiopia, Madagascar, Namibia, and very sharply, while in four more African countries Togo both factors contributed to the increase in it increased the level of vulnerability by more than oil vulnerability, suggesting that these countries 1 percent of GDP. Further analysis would be may need to reappraise their energy policies in needed to see which of the factors contributed to case oil prices rise even further. Decomposition of changes in vulnerability between 1996 and 2006 Table (percentage of current GDP in U.S. dollars except where otherwise specified) 2 Import Energy Oil price dependence Oil share intensity Real exchange Vulnerability Change in Energy intensitya Country effect effect effect effect rate effect in 1996 vulnerability 1996 2006 Albania 2.5 3.2 1.5 0.4 1.4 0.5 5.4 27.9 22.8 Australia 0.7 0.4 0.1 0.0 0.2 0.3 0.8 12.3 12.1 Bangladesh 2.1 0.1 0.1 0.2 0.2 0.9 2.2 10.2 11.1 Benin 4.9 1.4 0.5 1.4 1.4 1.8 6.7 8.4 11.5 Burundi 5.4 0.0 0.2 0.0 1.1 2.2 6.4 9.5 9.6 China 1.4 1.8 0.0 0.2 0.3 0.3 2.7 40.9 35.5 Ethiopia 3.0 0.0 0.4 1.0 0.0 1.0 4.3 5.7 8.3 Gambia, The 6.4 0.0 0.1 0.7 0.9 2.9 6.7 8.9 7.9 Ghana 5.0 0.5 1.7 1.1 0.7 2.2 5.4 29.3 22.7 Guyana 17.2 0.0 0.6 5.2 1.5 7.1 20.2 21.1 29.8 Madagascar 4.4 0.0 0.6 1.2 0.0 1.6 6.1 6.8 9.1 Malawi 4.0 0.0 0.4 0.1 1.3 1.6 4.9 12.5 12.3 Namibia 4.6 0.0 0.5 1.1 0.8 1.9 5.4 10.4 13.8 New Zealand 1.9 0.5 0.3 0.4 0.2 0.8 2.0 16.9 12.9 Paraguay 4.3 0.0 0.9 0.1 0.4 1.6 5.4 50.7 49.1 Togo 8.8 0.0 2.3 6.9 1.2 2.1 16.8 10.1 25.0 Uganda 1.9 0.0 0.0 0.0 0.2 0.9 2.1 5.4 5.3 Uruguay 2.9 0.1 0.4 0.3 0.5 1.2 3.3 6.6 7.3 a. Thousands of British thermal units (Btu) for every U.S. dollar of GDP in 2000 prices based on market exchange rates. Source: Authorsí calculations. O I L P R I C E R I S K S M E A S U R I N G T H E V U L N E R A B I L I T Y O F O I L I M P O R T E R S Box Methodology for decomposing changes in vulnerability 1 Using an accounting identity to decompose changes in vulnerability over time allows the separation of the pure oil price change effect from other changes that can affect vulnerability. For the analysis described in this Note, the following identity is used: viewpoint [ONP$/GDPC$] [ONP$/ON] [ON/OC] [OC/EC] [EC/GDPRL] [GDPRL/GDPCL] [GDPCL/GDPC$] is an open forum to ONP$ volume of net oil imports per year in barrels times annual crude price in U.S. dollars GDPC$ value of GDP in current U.S. dollars encourage dissemination of ON volume of net oil imports per year in barrels public policy innovations for OC volume of oil consumption per year in barrels private sector­led and EC total primary energy consumption per year in quadrillion Btu market-based solutions for GDPRL value of GDP in constant local currency development. The views GDPCL value of GDP in current local currency published are those of the authors and should not be To analyze the change in vulnerability over a number of years, based on the identity, a log mean Divisia index (LMDI) is used. The attributed to the World change in a country's vulnerability (D ) between any two years can be decomposed into the sum of the effects of the change in Bank or any other affiliated P (the oil price effect, P ); the change in (the share of imported oil in total oil consumption, the import dependence effect, ); organizations. Nor do any of the change in (the share of oil in total energy, the oil share effect, ); the change in (the ratio of energy consumed to real the conclusions represent GDP, the energy intensity effect, ); the change in (the ratio of GDP in constant to current prices, the inverse of the price official policy of the World deflator effect, ); and the change in (the ratio of current GDP valued in local currency to current GDP valued in U.S. dollars, Bank or of its Executive the exchange rate effect, ). The sum of the change in and the change in is the real exchange rate effect ( ). Directors or the countries they represent. D ( ) ­ (0) + + + + To order additional copies The effects, in turn, can be calculated from the following formula using the log mean Divisia index: contact Suzanne Smith, managing editor, [ ( ) ­ (0)] { log [ ( ) / (0) ] / ln [ ( ) / (0) ] } Room F 4K-206, The World Bank, with other factors calculated analogously. 1818 H Street, NW, Washington, DC 20433. Conclusion poses the change in vulnerability for a larger number of oil Telephone: This illustration of the decomposition of vulner- importers and exporters. 001 202 458 7281 ability indicates that it can be a useful tool in 1. Data for Cambodia from the International Energy Fax: focusing attention on movements in the econ- Agency differ substantially from the EIA data through 001 202 522 3480 omy that will make a country more vulnerable to 2005 and would place it in this group with changes in vul- Email: oil price shocks. It highlights the role of changes nerability greater than the oil price change. This Note uses ssmith7@worldbank.org in import dependence, in the oil share in energy, EIA data because they extend through 2006. and in the energy intensity of GDP, all of which Produced by Grammarians, may be susceptible to instruments of government Reference Inc. policy. For any country, more detailed analysis is Ang, B. W. 2005. "The LMDI Approach to Decompo- required to understand the determinants of the sition Analysis: A Practical Guide." Energy Policy 33: 867­71. Printed on recycled paper level and changes in vulnerability so that appro- priate policies can be designed to reduce it. Notes The results presented here are drawn from an ongoing study on policy responses to high oil prices that decom- T h i s N o t e i s a v a i l a b l e o n l i n e : h t t p : / / r r u . w o r l d b a n k . o r g / P u b l i c P o l i c y J o u r n a l