Historic, Archive Document Do not assume content reflects current scientific knowledge, policies, or practices. • USDA • • • Electronic Report from the Economic Research Service United States Department www.ers.usda.gov of Agriculture The Soybean Processing Decision Technical Bulletin Number Exercising a Real Option 1897 on Processing Margins Gerald Plato Abstract The gross soybean processing margin (the gross return per bushel of soybeans processed) is the main decision variable that processors use in deciding when and if to make binding commitments to process soybeans on future dates. Understanding how processors choose processing margins for future processing dates from among those available on successive days may help to resolve the ongoing concern about the level of competitiveness in processing agricultural commodities. Processing returns are treated as being equivalent to the returns to a call option. This approach provides the opportunity to simulate processor choice of processing margin by evaluating the incentive of waiting for a larger processing margin versus the incentive of locking in the currently available processing margin for a future date. The approach captures the irreversibility of the decision to process soybeans. Once the decision is made to process soybeans it cannot be economically reversed because of the contractual penalties involved. Processing margins selected using evaluations of these incentives explained variation in soybean crush, whereas spot margins for the corresponding processing dates did not. Keywords: real options, option exercise, processing decision, processing margins, futures prices, soybean crush Acknowledgments I thank Betsey Kuhn for continuously supporting this report after reading the first draft. I also thank Noel Blisard, Mark Denbaly. Bruce Gardner, Abebayehu Tegene, and George Wang for many valuable comments on and discussions about earlier drafts. List of Figures 1 Histogram of chosen gross processing margins .27 2 Time line of chosen futures gross processing margins .27 3 Soybeans crushed .27 4 Percent crushing capacity used .27 5 Capacity minus crush.27 6 Spot gross processing margin .27 List of Tables 1 Price movement combinations of asset 1 and asset 2 for up to four price moves .28 2 Equation for calculating call option prices on the price difference between asset 1 and asset 2.30 3 Dickey Fuller unit root tests.32 4 Average percent processing capacity used for the months in which the futures contract combinations expire.33 5 Percent processing capacity used for high- and low-capacity months.33 6 Estimated relationships between the chosen futures gross processing margins and soybean crush, percent capacity used, and capacity not used.34 7 Average futures and spot gross processing margins for highl¬ and low-capacity months .36 8 Estimated relationships between spot gross processing margins and soybean crush, percent capacity used, and capacity not used.37 iii Summary The soybean processing decision is a complex decision concerning when and if to commit to employing soybean-processing resources on future dates. Understanding this decision may help to resolve the ongoing concern about the level of competition in processing agricultural commodities. The decision when and if to process soybeans for a future date can be postponed until the date arrives. Further postponement is not possible when a future date arrives. In this situation, it must be decided to use the processing resources or idle them and forever lose the date's processing capacity. In addition, deciding to commit or pledging to use processing resources on a future date is economically irreversible because of the penalties involved in not fulfilling the commitment. The key decision variable concerning when and if to commit processing resources on a future date is the soybean gross processing margin (GPM) for that date. GPM is the soybean meal plus soybean oil revenue per bushel of soybeans processed minus the soybean (per bushel) price. GPM can be "locked in" for a future date by committing to employ soybean processing resources on the date. The commitment is made by committing to deliver soybean oil and meal on the date at agreed-to prices and committing to accept delivery of soybeans on the date at an agreed price. Soybean processors each day can calculate the GPM on future dates using the observed futures prices of soybeans, soybean meal, and soybean oil on future dates. Each day, a processor must decide whether a GPM for a future date is high enough to accept or too low to accept. If the GPM is high enough, the processor will commit soybean-processing resources for that date. If the GPM is too low to accept, the processor keeps the processing resources available for a future date and a higher GPM. The GPM that soybean processors receive from processing soybeans can be viewed as the returns to a call option. An option provides the right but the not the obligation to make a decision. Typically the right has an expiration date. Once the decision is made (the right exercised), it is irreversible. A call option provides the right to buy (receive) an asset by paying an exercise price. Using this right is called exercising the option. If an option involves the returns of a physical asset, it is called a real option. Soybean processors in effect have a real call option on their processing resources. They can receive returns equal to the GPM by paying variable processing cost, their exercise price, or they can choose not to produce and wait for a larger GPM. Viewing the returns to soybean processing as the returns to a call option provides opportunities to examine incentives faced by soybean processors, to simulate soybean processor decisions, and to measure processing returns. All are important in examining the level of competitiveness in soybean processing. A recently developed option pricing model in conjunction with futures prices for soybeans, IV soybean meal, and soybean oil was used to simulate decisions made by soybean processors. The futures prices were used to calculate daily GPMs that are available for future dates and the option-pricing model was used to choose the GPM levels instead of a soybean processor. A GPM was chosen for a processing date when the estimated value of waiting for a larger margin first fell to zero. The option-pricing model provided estimates of the value of waiting. The futures GPMs chosen by the option-pricing model explained a significant amount of the variation in monthly soybean crush. A similar examination using spot GPMs did not. In addition, the examination indicated that the curve of the relationship between crush and the chosen futures GPMs becomes flatter as the soybean-processing capacity limit is approached, as suggested by economic theory. That is, as the capacity limit is approached, it takes successively larger increments of GPM to commit a given increment in the amount crushed. The examination also showed the importance of controlling for the effect of processing capacity on the soybean processing decision when estimating the effect of GPM on the amount crushed. Capacity was controlled for by using percent capacity used and capacity not used as alternative dependent variables to be explained by the chosen futures and spot GPMs. The two dependent variables also offer ways to correct for a unit root in monthly crush. The correction is due to soybean-processing industry investment that increases capacity when tight and industry abandonment that reduces capacity when in excess. The procedures developed in this bulletin may be helpful in understanding the meat processor (packer) behavior involved in selecting forward prices for purchasing cattle, hogs, and sheep from farmers reported under the Mandatory Livestock Reporting Act of 1999. Forward selling prices available on the dates of the reported forward purchase prices can be used along with the reported purchase prices to estimate forward GPMs. These estimated GPMs then could be examined using the procedures developed here. v The Soybean Processing Decision Exercising a Real Option on Processing Margins Gerald Plato Introduction The decision to process soybeans is treated in this report as the equivalent of a decision to exercise a call option. Such an approach provides a way to explore the incentive to process soybeans and a way to explore the measurement of processing margins. The determination and measurement of processing margins are important in the debate about the level of competition in food processing. Much of the debate focuses on the concern that processor contracting may be reducing the level of competition. However, contracting is frequently the way that processors insure adequate commodity supply and the way they price margin prior to processing. A major problem inhibiting progress in this debate is that the timing of the decisions to process as well as the prices involved have been proprietary information, making it difficult to measure and examine processing margins. This report examines historical soybean processing margin data using an appropriate option valuation model to determine if progress in understanding and measuring processing margins can be made without using proprietary information.' The prices used in this bulletin are futures-contract prices for soybeans and soybean products. Forward-contract prices for soybeans and product, if available, could also have been used. In addition, futures and forward contract prices for other commodities and products can be used to examine the determination and measurement of processing margins. The examination that follows shows that frequently there is an economic incentive to buy soybeans and to sell product with futures contracts prior to processing. It also shows that the soybean processing margin when this economic incentive first occurs is a better measure of the processing margin used by soybean processors than is the margin measured by spot prices at the time of processing.1 2 1 The Mandatory Livestock Reporting Act of 1999 requires meatpackers (processors) to report the forward contract prices paid to cattle, hog, and sheep producers. These new price data will offer better measurement of processing margins. The new data will also beg for an explanation of the timing of meatpacker decisions to buy livestock from farmers and hence provide opportunities to explore the notion that meatpackers view their processing margins like returns to call options. 2 The farm to retail margin reported by the Economic Research Service, which includes the processing margin, has been measured using spot prices. Using the reported forward purchase prices required by the Mandatory Livestock Reporting Act of 1999 will improve the measurement of the farm to retail margin for cattle, hogs, and sheep. The processing and the option exercising decisions are equivalent and, consequently, result in the same returns when three conditions are met: • the option-exercise price equals variable processing cost, • the soybean processing and the option exercise decisions are based on the same gross processing margin (GPM), and • processors treat the decision to commit processing resources in a manner similar to the decision to exercise a call option on the GPM, and the processing commitment is made (the option is exercised) when the time value of waiting to decide falls to zero. The GPM in this report is valued using the futures prices for soybeans, soybean meal, and soybean oil. It is the price of (revenue from) the meal plus oil per bushel of soybeans processed minus the price (cost) of a bushel of soybeans. Variable processing cost is the marginal cost of converting one bushel of soybeans into meal and oil. It is the cost that is avoided by deciding not to process soybeans. The first condition imposes no restrictions on a call option. Exercise price can be set at any level. Soybean processors often use the futures market to price processing returns. Consequently, the second condition imposes no unusual restrictions on the soybean processing decision. The validity of the third condition is an unknown that is examined here. Soybean Processing as a Real Option The large and growing literature that treats the decision to invest as equivalent to the decision to exercise a call option provides help in understanding the decision to process soybeans (Amram and Kulatilala; Hubbard; and Trigeorgis). Options on investment opportunities in physical assets and on the use of existing physical assets—buildings, equipment, and land—are known as real options. The exercise price of a call option on an investment opportunity is the cost of the investment (the amount invested), and the gross option return is the discounted expected value of the investment returns. This type of call option is exercised (the investment is made) when the gross return rises above the exercise price sufficiently to compensate for loss of the flexibility to delay. Flexibility to delay would have no value if an investment decision could be undone without cost. The soybean processing decision is examined, in part, by examining how much the GPM must rise above variable processing costs to compensate for the loss of the flexibility to delay. Similarly, the flexibility to delay processing decisions would have no value if unfulfilled processing commitments could be canceled or disregarded without cost and replaced with "commitments" that offer a processor a higher margin. 2 Examining investment decisions as a decision to exercise or hold a call option has been shown to be a useful approach for a public entity or a conservation trust in estimating the amount of compensation to offer landowners for selling some of their investment options (development rights) (Wiebe, Tegene, and Kuhn). The investment options in that case are conservation easements that restrict development (investment) choices. Examining the investment decision this way is similar to evaluating how much to offer a soybean processor to sell processing capacity (resources). Uncommitted capacity provides a processor the right but not the obligation (a real option) to process soybeans. An approach used in the real option’s literature is to consider the per period (for example, annual) production capacities provided by an investment as offering a series of call options on the capacity (McLaughlin and Taggart). These call options are called production options by McLaughlin and Taggart and are contained within (imbedded in) an investment option.3 Each production option has a gross return equal to the price of the output minus variable costs or zero depending on whether the price of the output turns out to be larger or smaller than variable costs, respectively. The sum of the discounted estimated values of these per period call options estimates the expected returns to capacity (capital). The production options examined by McLaughlin and Taggart are similar to the call options on the GPM examined in this bulletin. Investment options with their imbedded production and other options can be valued using option valuation models developed for exchange-traded options or valued using option valuation models that are derived to account for special conditions of the investment decision (Tegene, Wiebe, and Kuhn). The valuation requirements are an appropriate option valuation model and data that tracks the returns to an investment (Amram and Kulatilaka). The tracking assets in this bulletin are soybeans, soybean meal, and soybean oil, and the production resources are soybean-processing capital. McLaughlin and Taggart specified their production options as European call options and suggested that these options could be valued using the Black-Scholes option valuation model (Black and Scholes). This option valuation model is not appropriate for examining the soybean GPM, because it does not allow for less than zero gross returns, that is, for an inverse crush where product revenue is less than the price of soybeans. An option valuation model that allows negative GPM outcomes is used in this report to examine the soybean processing decision. 3 Production options, abandonment options, altemative-capital-use options, and follow-on investment options imbedded within an investment option are frequently used in the literature of real options (Trigeorgis). Imbedded options introduce decisionmaking following a decision to invest. The subsequent decisionmaking increases the expected value of potential investments by cutting losses from bad outcomes and by increasing returns from taking advantage of new opportunities. A production option offers the ability to cut losses from bad outcomes. 3