How Do Ethanol Ending Stocks Affect Ethanol Price

  Iowa Grain Quality Initiative Logo
Decision Innovation Solutions Logo

By Sampath Jayasinghe and Jing Tang
Decision Innovation Solutions, 11107 Aurora Avenue, Urbandale, IA 50322
www.decision-innovation.com/

December 6th, 2016

According to the most recent Weekly Ethanol Plant Production report from the Energy Information Administration (EIA), U.S. ethanol production reached 1.017 million barrels per day for the week ending November 11. Correspondingly, weekly ending stocks of fuel ethanol are at 18.609 million barrels as published by EIA’s Weekly Supply Estimates report. Figure 1 illustrates weekly U.S. plant production of fuel ethanol (thousand barrels per day) and weekly U.S. ending stocks of fuel ethanol (thousand barrels) from January 2013 to November 11, 2016.

Weekly U.S. Ethanol Production and ending StocksThe historical highest record of production was seen in July and August 2016. Monthly average production in July was 1.003 million barrels per day and monthly average production in August was 1.024 million barrels per day. The average production rate in September and October fell mainly due to a slowdown in blending because of regular repair and maintenance work done each fall at production plants. Ethanol production has rebounded back above one million barrels per day in the last three consecutive weeks ending November 11, 2016.

Meanwhile, weekly ending stocks have consistently fallen to a 12-month low as shown in Figure 1. This is mainly due to gains in both domestic usage and exports. As we write this report (November 23, 2016), ethanol futures price of December 2016 contracts increased by approximately 6 percent since the EIA’s ending stocks numbers were released November 16. **Note that ethanol futures are thinly traded compared to other commodities in the futures market. In this month’s article, we look at the relationship between the monthly ending stocks and the monthly average wholesale (rack) ethanol price. U.S. bioenergy statistics released on November 11, 2016 by the Economic Research Service of the United States Department of Agriculture (USDA-ERS) are used in this analysis.

Relationship between ethanol ending stocks and wholesale ethanol pricesFigure 2 shows the monthly ending stocks of ethanol and the average monthly wholesale ethanol price in Omaha, Neb. from January 2010 to August 2016. Note that USDA’s bioenergy statistics accessed on November 23, 2016 reports monthly ending stocks only up to August 2016. In general, we expect ethanol prices should be negatively correlated with the ending stocks. Relatively lower stocks are an indicator of the tightness of ethanol market condition. We find the correlation coefficient between the monthly ethanol ending stocks and the average monthly wholesale price of ethanol to be -0.42 from January 2010 to August 2016.

As seen in Figure 2, wholesale ethanol prices have moved the opposite direction to the ending stocks with few significant exceptions. Notice in the chart, in March-April 2014, wholesale ethanol prices temporarily moved along with ending stocks due to serious winter-related rail transportation impediments. The severe winter in 2014 was coupled with a strong demand for rail transportation from the newly emerging and booming crude oil industry in North Dakota. Ethanol transportation was hampered by the increased competition from crude oil transportation. This tightened ethanol supply distribution to the rest of the United States and spiked the price temporarily. Also note the relative unresponsive period of ethanol price to the declining ending stocks during May-Oct 2015. This can mainly be attributed to the sharp decline in crude oil and gasoline prices. Ethanol price remained relatively flat in spite of a sharp decline in ending stocks, because ethanol price should maintain reasonable competitiveness with gasoline prices to support blending economics. 

To understand the relationship between ethanol prices and ethanol ending stocks, we begin by doing simple linear correlation among three major determinants of ethanol price — monthly average corn price, monthly average wholesale gasoline price, and monthly ethanol ending stocks. All the relevant data are collected from the USDA’s bioenergy statistics.

The linear correlation between monthly average wholesale ethanol price and corn price is 0.76, and the correlation between monthly average wholesale ethanol price and monthly average wholesale gasoline price is 0.82 for the period from January 2010 to August 2016. Again note the correlation between monthly average wholesale ethanol price and the monthly ethanol ending stocks is -0.42. As expected, gasoline and corn price are positively correlated with ethanol price, and ending stocks are negatively correlated with lesser magnitude compared to gasoline and corn price. The value of ethanol is expected to strongly and positively correlate with gasoline price as the value of blending ethanol is determined by the value of petroleum-crude oil and gasoline as discussed in a previous AgMRC article. A strong positive correlation is expected between ethanol and corn price, as corn is the major source of feedstock in production of ethanol in the United States.

Second, we did a regression analysis to investigate the determinants of ethanol price. A Generalized Linear Model (GLM) was fitted on ethanol price using the above three covariates: gasoline price, corn price, and ethanol ending stocks using R statistical software. Specifically, we estimated causal effects of ethanol price from gasoline price, corn price, and ethanol ending stocks. There are a few advantages of GLM over simple linear regression, such as it does not need to transform the response variable (i.e., ethanol price) and independent variables (i.e., gasoline price, corn price, etc.) to have a normal distribution.

The R-squared is 0.77 indicating that 77 percent of ethanol price variation can be explained by gasoline price, corn price, and ethanol ending stocks. R-squared is a statistical measure of how close the data are to the fitted regression line. The results show the parameter estimates for gasoline price, corn price, and ethanol ending stocks are all statistically significant (p< 0.001). Gasoline price and corn price have positive impact on ethanol price, whereas ethanol ending stocks have negative impact. Gasoline price matters most, followed by corn price and ethanol ending stocks. The estimated parameter for gasoline price indicates that $1 increase in gasoline price will increase ethanol price by $0.32. The estimated parameter for corn price indicates that $1 increase in corn price will increase ethanol price by $0.14. Finally, the estimated parameter for ethanol ending stocks shows a one million gallon increase in ethanol ending stocks will decrease ethanol price by $0.0016.

Concluding Remarks

The major finding of this analysis is that gasoline price and corn price have statistically significant positive relationships with ethanol price, whereas ethanol ending stocks have a statistically significant negative relationship. Gasoline price matters most, followed by corn price and ethanol ending stocks. Ethanol ending stocks have been influential in determining ethanol price but not to a great degree. There are a few limitations of this analysis. There are so many other factors that affect the price of ethanol, such as domestic ethanol consumption and transportation costs. These factors are not taken into account in our regression analysis.

References

U.S. Energy Information Administration (EIA), various. Weekly Petroleum Status Report.

USDA-ERS (U.S. Department of Agriculture, Economic Research Service), 2016. Bioenergy Statistics Data. Downloaded November 15, 2016.

McCullagh, P., J.A. Nelder, 1989. Generalized Linear Models, Second Edition, CRC Press LLC, Florida, USA.

R Core Team, 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.

**We thank Dave Miller and Sue Retka-Schill for helpful initial discussion on this topic.