Commodity Storage and the Market Effects of Biofuel Policies

Abstract

US legislation passed in 2007 (RFS2) increased by about 1.3 billion bushels the net amount of corn required to be processed annually into ethanol for motor-fuel use. Using modern time-series methods, we estimate that corn prices were about 30 percent higher between 2006 and 2014 than they would have been but for RFS2 and if pre-2006 trends had continued. We estimate a permanent corn demand increase of 1.3 billion bushels increased the long-run price by 31% (90% confidence interval is [5%,95%]). Our identification strategy is unique in the literature because it enables estimation of the effects of transitory shocks, such as weather, separately from the effects of persistent shocks, such as the ethanol mandate. Keywords:Ethanol; agriculture; energy policy; VAR; partial identification The difference between the RFS2 and RFS mandates is approximately 5.5 billion gallons (bgal) of ethanol annually in the years 2010-2012, which corresponds to about 1.3 billion bushels of corn after accounting for feed by-products. We take these 1.3 billion bushels as the permanent increase in corn demand in the RFS2, and we estimate that it caused a 31 percent long-run increase in corn prices.Interpreting our estimates correctly requires a precise statement about counterfactual ethanol production. In our counterfactual, ethanol production would have increased from its actual 2005 value of 3.9bgal to 8.8bgal in 2014. This path represents a continuation of the trend observed between 2002 and 2005 and is slightly below the path predicted by the USDA in early 2006. We argue it is a reasonable counterfactual path in the absence of the RFS2. We also point that our long-run estimate is scalable. To estimate the effect of increasing corn demand permanently by any multiple of 1.3 billion bushels, a reader could multiply our estimate by that multiple.Previous studies have also found that the increase in corn-ethanol production affected corn We focus our analysis on commodity inventory dynamics, which distinguishes our work from the extant literature. This approach enables us to estimate the price effects of persistent shocks to supply or demand separately from the effects of transitory shocks in a market for a storable commodity. This distinction is important in our context because persistent shocks have larger price effects than transitory shocks. The market can respond to a transitory shock, such as poor growing season weather, by drawing down inventory. This action mitigates the price effect. A persistent shock, such as an increase in current and expected future demand, cannot be mitigated by drawing down inventory.To identify these two types of shocks, we exploit their differential effects on inventory levels and on the term structure of futures prices. For example, all else equal, a sudden increase in this year's consumption demand reduces available inventories and raises spot prices relative to futures prices. This is a transitory shock. In contrast, a predicted increase in next year's consumption demand generates an increase in i...

Did the Renewable Fuel Standard Shift Market Expectations of the Price of Ethanol?

“…Figure 10, however, shows no evidence of an increase in the 12-month corn futures price in 2005 or in 2006 (except toward the end of the year), consistent with our economic reasoning. This result adds to the evidence that corn price expectations (and hence the storage demand for corn) do not move proportionately with ethanol price expectations (and storage demand), calling into question the interpretation in Wright (2014) and Carter, Rausser and Smith (2016).…”

“…The changes to U.S. biofuel policies after 2005 are a case in point. It is well understood that the response of the price of ethanol (and the price of related commodities such as corn) to shifts in biofuel policies operates in part through market expectations, yet it has proved difficult to measure these expectations to date and to directly confront these views with the data (see Wright 2014;Baumeister and Kilian 2014;Carter, Rausser and Smith 2016).…”

“…This problem is well known. For example, Carter, Rausser and Smith (2016) deal with this data limitation by specifying a model of the corn market based on annual data. Our approach is different.…”

“…Finally, based on an econometric model of the corn market, Carter, Rausser and Smith (2016) concluded that expectation shifts associated with biofuel policies caused a surge in corn prices.…”

“…They may also respond to shifts in the supply of gasoline triggered by refinery outages, to cost shocks associated with shifts in the global price of crude oil, and with shifts in the domestic demand for gasoline driven by the U.S. business cycle or by changes in consumer behavior (see, e.g., Coglianese, Davis, Kilian and Stock 2016). Although there have been many analyses of these markets in isolation, only recently attention has been given to the close economic relationship between energy and agricultural markets (see, e.g., Salvo and Huse 2011;Mallory, Irwin, and Hayes 2012;Baumeister and Kilian 2014;Carter, Rausser and Smith 2016;Bielen, Newell and Pizer 2016).…”