motivation for the explanatory variables. We also provide a description of and motivation for the
three estimation methods that are used. The last section summarizes the results.
Previous Work
Quantitative analysis of the effect of ethanol on gasoline prices and on the profitability of the
refinery industry has been largely neglected in the literature. Eidman (2005) points out that
ethanol largely acts as a fuel extender. He also shows that there has been a strong positive
correlation between ethanol and gasoline prices. Employing an international ethanol model
consisting of behavioral equations for production, consumption, and trade, Tokgoz and Elobeid
(2007) analyze the price linkage between ethanol and gasoline markets. They conclude that
ethanol is mainly used as an additive to gasoline and that the complementary effect of ethanol
dominates the substitution effect on gasoline prices. Szklo, Schaeffer, and Delgado (2007)
conclude that by replacing methyl tertiary butyl ether (MTBE), which is a traditional additive
used as an oxygenate to raise the octane number, ethanol blending will not reduce gasoline use
until flexible fuel vehicles become widely available. Vedenov et al. (2006) apply a continuous-
time option pricing method to calculate the decision threshold of switching to ethanol. Their
empirical analysis suggests that blending ethanol into gasoline would generate lower gasoline
price volatility and that switching from conventional gasoline to an ethanol blend is an
economically sound decision.
The “3:2:1 crack spread” is used as one of the significant indicators of refinery profitability. It is
a term used in the oil industry and futures trading as a proxy for the profitability of refineries.
Although there is some qualitative description of its determinants, formal quantitative analysis is
limited in the literature. Asche, Gjolberg, and Volker (2003) examine the price relationships
among crude oil and refined products. They find that the crude oil price is weakly exogenous and
that the spread is constant among some of the prices. Girma and Paulson (1998) examine the
crack spread of daily futures prices of crude oil and heating oil. Girma and Paulson (1999)
investigate the long-run relationship among crude oil, gasoline, and heating oil futures prices and
find the prices are co-integrated. They also find a stationary relation between crude oil and its
end products.