Authors John Mayes and Mike Leger
In the Puff Daddy hip hop ode to crass materialism, the “Benjamins” (or money in general), are all that really matters. But is that the case when it comes to refined product demand? Or put another way; is the price of gasoline, diesel and other products the main determinant of how much is consumed? Certainly price is an important component in the equation, just as it generally is for most every commodity. And as we’ve discussed in previous blogs, an important silver lining of the crude price collapse for refiners has been the stimulative impact of lower prices on demand. This should be particularly true in the U.S. for gasoline, where fuel taxes are low and the price the consumer sees for gasoline moves pretty well in tandem with crude prices; but, price is certainly not the only driver of product demand; and, in fact, the relationship can be very complex. It is also true that demand for different products each has its own supply and demand dynamics and relationships with price. In today’s blog, we explore those relationships not only for gasoline and diesel, but also propane, butane, and natural gas; and, in the process, provide an answer for PD on “what’s real in the field.”
Propane
We’ll start with the lightest products, beginning with C3. Two-thirds of U.S. propane and propylene production is derived from natural gas processing plants and only one-third actually comes from petroleum refineries, where it is generally considered a byproduct. With very low natural gas prices, most drilling in this segment is focused on wet gas wells, meaning wells which yield a high volume of natural gas liquids (NGLs), with propane as one of the primary components. U.S. production of propane from natural gas plants has increased from 586 MBPD in 2010 to 1,117 MBPD in 2015, while refining yields of propane and propylene have been relatively flat. Demand, however, has declined by a modest 37 MBPD over the same period.
This surging production, with falling demand, has sent propane prices dramatically lower over the last six years. In 2010, propane prices in Mont Belvieu averaged $1.17 per gallon, but this fell to only $0.45 per gallon in 2015. As can be seen in Figure 1, there is no correlation between propane prices and demand. A statistical regression analysis yields a positive slope for the pricing elasticity, meaning propane demand is more likely to increase when prices increase, but the relationship is very weak. Obviously, it can be seen that there is very little to no elasticity of propane prices to demand, at least in the short term. It remains to be seen if the current low prices will stimulate new demand sources in the future, but which may take years to develop. In lieu of increased U.S. demand, net propane exports have risen in recent years. In 2010, the U.S. had net propane imports of 13 MBPD; but by 2015, the U.S. was exporting over 500 MBPD.
Butane
Like propane, the bulk of C4 production is from natural gas plants. Because refineries consume virtually all of their isobutane and butylene production in alkylation units and most of their normal butane output in gasoline-blending, natural gas plants yielded 93% of the total C4 production in 2015. From 2010 to 2015, C4 production from gas plants increased by 277 MBPD, while output from refineries declined by 32 MBPD. Over the same time period, total C4 demand increased by 151 MBPD.
There is a substantively better correlation between price and demand for C4s than for any other petroleum product. Falling prices have, in fact, increased total butane demand, at least for recent history. While increased consumption has absorbed most of the incremental production, net exports have also increased. In 2010, the U.S. had net imports of 9 MBPD of C4s, while in 2015 net exports increased to 84 MBPD.
Gasoline
Gasoline is certainly the most important and also the most complex product to evaluate as there are numerous short-term and long-term variables impacting demand. Almost half of the total refined product demand in the U.S. is for motor gasoline, and most adults over the age of 16 make purchase decisions on a regular basis. Generational shifts in driving patterns, CAFE mandated mileage improvements in automobiles, increased telecommuting, and fuel prices changes all influence consumption rates, and it can be very difficult to isolate the pricing correlation. Since 2010, gasoline demand has been erratic. After beginning the period at nearly 9.0 million BPD, consumption fell to 8.7 million BPD by early 2013, but has since risen briskly to nearly 9.2 million BPD. Average U.S. retail gasoline prices began the period below $3.00 per gallon, but then averaged between $3.40 and $4.00 per gallon from 2Q2011 through 3Q2014. Retail prices ended 2015 slightly below $2.15 per gallon.
Because of the complexities involved, the direct correlation between price and demand is not as strong for gasoline as it has shown to be for butane; however, clearly the price declines which began in the second half of 2014 have played a major factor in “fueling” the strong increases in consumer driving and consumption of gasoline. 2015 showed the highest increase in gasoline demand in the U.S. (240 MBPD) since 1976 and the increase in vehicles miles traveled (VMT) of 3.6% was the biggest since 1989. Adding complexity to the situation is the fact that the upturn in demand actually began even before prices fell, starting in early 2013, as shown in Figure 3 above. This was likely the result of early CAFE mileage improvements which had the same effect as reductions in the gasoline price. Going forward, ever more stringent CAFE mileage improvements will continue to impact demand growth, as will the demographic factors that we’ve discussed. Also, the “velocity” of price movements, both up and down, will play a role. For instance, even if prices remain low, the stimulative impact will be lessened as consumers both “get accustomed” to a certain price level and limits are reached on how much growth in VMT is actually possible.
Diesel
Diesel demand in the U.S. has long been known to track economic growth and is less responsive to price movements than gasoline. An example of this can be seen in Figure 4. Diesel demand has been weakening in the last several months in spite of the lowest prices in this decade. This is consistent with the latest economic reports which indicate a weakening of U.S. economic growth. A similar slowing of economic growth was also seen in the second half of 2012 which resulted in a modest downturn in consumption through the first half of 2013. Since 2010, however, diesel demand has increased by 400 MBPD, the largest increase for any of the petroleum products.
While there is a poor short-term statistical relationship between diesel prices and demand, there is likely to be a better long-term elasticity. As the U.S. remains a net petroleum importer, lower petroleum prices should result in stronger economic growth, which in turn should lead to stronger diesel consumption growth. This impact today, however, is being offset by the contraction of the crude exploration and development industry resulting from the lower crude price environment.
Natural Gas
Like propane and diesel, there is little correlation between the price of natural gas and its rate of consumption. Demand has risen steadily since 2010 from slightly over 1.9 trillion cubic feet per month to 2.3 trillion cubic feet per month. During this same period, natural gas prices have gyrated wildly between $6 per million BTU and $2 per million BTU. Prices are more likely to remain on the low side in the near term due to the mild recent winter and massive overhang of potential reserves which could easily be developed.
Summary
Price/demand elasticity varies significantly between petroleum products. Butane has an excellent correlation with an R2 = .7938, but the propane, diesel, and natural gas correlations are very weak, in large part because consumers are not as directly impacted by price movements, and also because other factors (including weather, general economic activity, etc.) have more direct effects on demand. For gasoline, a strong correlation between price and demand certainly is evident, and it appears that price likely continues to be the dominant variable; however, there is significant complexity involved in how much gasoline U.S. consumers purchase, involving many other factors (CAFE, other governmental policies, demographics, changes in driving habits, etc.), and it is not just “about the Benjamins,” in fact, the relationship will change as these other factors become more or less important in the future. As a result of these conflicting market forces, the future trend of the premium/regular price spread is clouded.
At Turner, Mason & Company, we continually evaluate the effects of changing dynamics on petroleum demand (and other aspects of supply/demand balances). We use these analyses to forecast impacts on refiners and other industry participants. These analyses and forecasts are discussed in detail in our recently issued 2016 Crude and Refined Products Outlook, and are also used in our ongoing consulting engagements assisting individual clients and industry groups. Please feel free call or email us or visit our website if we can answer any questions or assist you in any way.