It is also possible to classify factors affecting macroeconomic balance and gasoline prices according to their function: demand, supply, physical balancing and behaviour of related markets. Fig. 1 illustrates the major groups of factors affecting gasoline prices.
Figure 1. Groups of factors affecting energy prices (Gasoline Prices, 2005)
Among this variety of factors, there are several indicators that are core determinants of energy prices. According to (Annual Energy Outlook 2010, 2010), these are OPEC and non-OPEC supply characteristics (volumes and prices), spot prices reflecting major economic and geopolitical events, demand for crude oil products (classified as OECD and non-OECD demand), the behaviour of financial markets and the changes in inventories (which reflect the difference between supply and demand, and can be used to forecast expected fluctuations of gasoline prices). Among these factors the first one (the reaction of gasoline prices to economic and geopolitical changes) is non-quantifiable, while other factors can be included into the model (Toggins, 2008). The goal of this paper is to consider each of these factors as part of economic model and to determine the indicators of gasoline price changes using this model.
2. Data analysis
2.1. OPEC supply
Organization of the Petroleum Exporting Countries (OPEC) and its decisions directly affect gasoline prices due to the existence of OPEC quotas. Figure 2 shows the dynamics of changes of OPEC targeted production (expressed in year to year change) and crude oil prices. It should be noted that the growth of crude oil prices was witnessed when OPEC targets were reduced.
Figure 2. OPEC production targets vs WTI crude oil prices
Correlation coefficient for OPEC production targets and WTI crude oil prices is 0.62, which means that there is a notable relationship between these two variables (Toggins, 2008), and this factor should be used for forecasting changes of gasoline prices. Statistical data for sections 2.1-2.5 was taken from the website of US Energy Information Administration (2011).
OPEC oil production constitutes about 60% of total world oil production (US Energy Information Administration, 2011); however, such non-OPEC regions as North America and former USSR countries can also have an influence on oil prices (Gasoline Prices, 2005). Production of non-OPEC regions (listed as change in millions of barrels per day) and its dynamics compared to WTI crude oil prices per barrel during the 2001-2011 period is shown on Figure 3.
Figure 3. Non-OPEC production and WTI crude oil prices
The impact of non-OPEC production on gasoline prices is ambiguous: there are periods when increase of non-OPEC supply resulted in reduction of crude oil prices, and at the same time there are periods when market reaction was quite different. Most likely, non-OPEC production has a medium to low influence on gasoline prices. Correlation coefficient for non-OPEC production is -0.32, which means that non-OPEC supply slightly reduces prices. This factor can be used for predicting the dynamics of world crude oil prices, with notification that it’s a minor element in price formation (correlation coefficients with absolute value between 0.3 and 0.5 represent weak relationships) (Toggins, 2008).
2.3. World demand for gasoline (OECD and total)
OECD (Organization for Economic Cooperation and Development) includes USA, large part of Europe and a number of other developed economies. This economic group is the major world oil consuming entity, and it is essential that the prices for crude oil/gasoline are determined by the demand for oil products, which is, in its turn, driven by OECD oil consumption. Figure 4. illustrates the dynamics of crude oil prices (per barrel, in 2009 dollar value) and OECD consumption of liquid fuels (expressed in % of change).
Figure 4. OECD consumption and WTI crude oil prices
Statistical analysis of correlation between the OECD oil consumption data and WTI crude oil prices showed that correlation between these values was -0.17, which is a not significant value for including this factor in the analysis. Although certain patterns between OECD consumption and oil prices can be witnessed, the dynamics of OECD does not significantly affect oil prices.
The absence of significant relationship between oil prices and use of gasoline in developed countries can be explained by the great number of taxes and regulations existing in these countries (US Energy Information Administration, 2011). These regulations tend to reduce the growth of gasoline consumption even during economic recovery, which can be seen at Fig. 4.
Let us analyze the impact of overall consumption of liquid fuels and the state of world economy on WTI world prices (Fig. 5).
Figure 5. Total demand for liquid fuels, world GDP and WTI crude oil prices
Table 1 contains correlation analysis for these three factors.
World consumption of liquid fuels World GDP WTI crude oil prices
World consumption of liquid fuels 1
World GDP 0.81 1
WTI crude oil prices 0.042 0.158 1
Table 1. Correlation between WTI crude oil prices, world consumption and world GDP
Results of analysis prove that although there is a strong correlation between world GRP and world oil consumption (correlation coefficient is 0.81), WTI crude oil prices are not significantly related to any of above-mentioned factors. This result shows that within gasoline market the power of suppliers is rather high (positive correlation with prices) and the power of buyers is low, since the changes of the demand do not significantly affect prices (Carbaugh, 2008).
2.4. Financial markets
Trade contracts related to oil products, gasoline and crude oil can also affect the dynamics of gasoline market. It is possible to see that the interest to crude oil future has significantly increased since the 2000s, with certain fluctuations due to economic recession and recovery. Fig. 6 illustrates the growth of market interest to crude oil futures, expressed in the number of contracts per year.
Figure 6. Interest to crude oil futures (expressed in thousands of contracts)
Correlation analysis of interest to crude oil future and the dynamics of WTI crude oil prices has shown that the relationship between these two factors is not statistically significant (correlation ratio is 0.12).
Inventories are likely to be the leading indicators of changes for crude oil and, consequently, gasoline prices: when inventories increase, supply is growing and prices are likely to fall, and vise versa, when supplies are decreasing, this indicates the growth of demand and the expected increase of prices (Gasoline Prices, 2005). It is most likely that inventories will be strongly related to the change of oil price of future, and, consequently, will have a weaker relationship with the factual crude oil prices. Fig. 7 illustrates the changes of OECD inventory and dynamics of WTI future spread.
Figure 7. OECD inventory and WTI 12-1 futures price spread
Statistical analysis of these factors and WTI crude oil prices has shown that there is a relationship of medium strength between these factors. Table 2 shows the results of the analysis.
OECD inventory change WTI 12-1 futures price spread WTI crude oil prices
OECD inventory change 1
WTI 12-1 futures price spread 0.75 1
WTI crude oil prices -0.62 -0.48 1
Table 2. Correlation between WTI prices, OECD inventory change and WTI spread change
Thus, the suggestion was confirmed, and OECD inventories have a inverse relationship to WTI future price spread, and, what is more important, have a slightly weaker inverse relationship to WTI crude oil prices. The values of OECD inventories can be used to determine a forecast for crude oil and gasoline price dynamics, while previously mentioned factors can be used to correct this forecast and to establish an effective model for predicting gasoline price changes in the short run.
Analysis of factors affecting gasoline prices has shown that among the variety of factors affecting these prices, there are several constituents which can be successfully quantified and used as a basis to forecast future values of crude oil and gasoline prices. The factors suggested for analysis were world demand and supply (OPEC and non-OPEC) for crude oil products, behaviour of financial markets and OECD inventories. Statistical analysis has shown that key factors which can be used to forecast gasoline prices are OPEC and non-OPEC supply and OECD inventories. It is recommended to determine trends of future gasoline prices using OECD inventory indicators, and to refine these values using oil supply characteristics. It is also recommended to incorporate important economic and geopolitical events into the model; however, the nature of adjustments depends on the character of the events and it is not possible to generate an algorithm for these adjustments. At the same time, the proposed model is likely to yield reliable results in the short run.
Annual Energy Outlook 2010. (2010). Government Printing Office.
Carbaugh, R.J. (2008). International Economics. Cengage Learning.
Gasoline Prices. (2005). Hearing Before the Committee on Energy and Natural Resources US Senate. Diane Publishing.
Toggins, W.N. (2008). New Econometric Modeling Research. Nova Publishers.
US Energy Information Administration. (2011). Available from http://www.eia.gov/