comparing returns for grain corn production under various ... · comparing returns for grain corn...

Comparing Returns for Grain Corn Productionunder Various Marketing Strategies

Ontario Corn Producers’ Association

University

Comparing Returns for Grain Corn ProductionVarious Marketing Strategies

Prepared for:

Ontario Corn Producers’ Association

Prepared by:

Richard J. Vyn University of Guelph, Ridgetown Campus

June 2009

Comparing Returns for Grain Corn Production Various Marketing Strategies

Comparing Returns for Grain Corn Production under Various Marketing Strategies

University of Guelph, Ridgetown Campus i | P a g e

Executive Summary

Introduction For corn producers, decisions regarding the marketing of their crop can be very difficult to make, as a significant amount of uncertainty surrounds these decisions. With considerable fluctuation in the price of corn, it can be challenging to determine the optimal timing of marketing decisions and the appropriate marketing tools to use to ensure that sales or contracts are made during periods of relatively higher prices. The development of a marketing plan or marketing strategy may assist with this decision making. However, the development of an appropriate strategy can be difficult, particularly due to a lack of information regarding the relative effectiveness of various marketing strategies. This report addresses this issue through the development and use of a simulation model to compare the returns from a set of specific marketing strategies for corn producers in Ontario. There have been a number of previous studies, primarily in the United States, that have examined the effectiveness of marketing strategies for various agricultural commodities. Many of these studies involved comparing the returns of alternative marketing strategies to a baseline strategy in which the entire crop is sold at harvest. There is no clear consensus in the results of these studies with respect to the effectiveness of alternative marketing strategies, such as pre-harvest strategies. In some cases, pre-harvest strategies were found to generate significantly higher prices than the baseline strategy, while in other cases, any differences in returns were not statistically significant. However, the results of these studies may not necessarily apply to producers in Ontario, as cash price determination differs due to the influence of the exchange rate. While one previous study has examined marketing strategies in Ontario, this study was conducted using data from 30 years ago, before options were available. Methods The simulation model developed for this report uses historical cash and futures price data to determine average prices generated by specific marketing strategies. Data for this model was derived from the Ontario Commodity Reports, which provide daily futures prices and cash prices at specific locations across Ontario from 1992 to the present time, as well as from the Chicago Board of Trade, which provides margin requirements and daily options data. The marketing strategies evaluated by this model were developed through consultation with marketing specialists and producers to ensure that these strategies are representative of those currently used by corn producers in Ontario. There are 18 strategies incorporated into the simulation model, which include the use of marketing tools such as cash sales (Strategies #1-4), forward contracts (#5-7), futures contracts (#8-10), options (#11-14), basis contracts (#15), and combinations of these tools (#16-18).


University of Guelph, Ridgetown Campus ii | P a g e

Prices for each strategy are selected from a range of dates, to allow for more flexibility and to avoid using a specific date on which prices may be relatively high or relatively low compared to the average price during that time period. For simplicity, there are five three-week time periods that are used across all strategies during which the marketing activities may occur: harvest (October 21 – November 10), early in the new year (January 5 – 25), spring (April 10 – 30), early summer (June 20 – July 10), and late summer (August 10 – 30). These time periods were selected in an attempt to reflect periods when a larger amount of corn is marketed. The tools used and the timing of marketing activities for each strategy are fixed across all years to allow for direct comparison of specific marketing strategies. However, multiple strategies are included for cash sales, forward contracts, futures contracts, and options, to allow for variation in the timing of the use of these tools. The returns generated by the simulation model take into account costs that are associated with the use of specific marketing tools. For strategies in which the entire crop is not sold by harvest, storage and inventory costs are deducted from the selling price. Strategies that involve the use of futures contracts factor in the interest costs on funds required for trading accounts and margin calls. Commission fees for futures contracts and options are also taken into account. The simulation model is run multiple times across all years to generate an average price per bushel for each strategy. The price for each strategy is compared to the baseline strategy (selling the entire crop at harvest) and t-tests are conducted to determine whether the differences in the average prices are statistically significant. In addition, the relative level of risk associated with each strategy is determined based on the standard deviation of the average price and on the probability that the price generated for a specific strategy would be less than the cost of production (COP). Results The results of the simulation model are provided in table 1. The strategies with the highest average prices across all years and the greatest price premiums over the baseline are Strategies #10, #8, #9, and #16, all of which incorporate short hedges using futures contracts. These strategies returned prices that are between 29.8 and 32.8 cents per bushel higher than the baseline strategy. The put option strategies (Strategies #11, #12, #13, and #17) also tended to have much greater returns than the baseline, generating price premiums ranging from 17.9 to 27.5 cents per bushel, only slightly lower than those of the futures contracts strategies. While most of the differences in average prices for the pre-harvest strategies relative to the baseline strategy are found to be significant, many of these differences are only significant at the 10% level. The results indicate that the standard deviation and probability that the price is below the cost of production are similar to or greater than the baseline for many of the pre-harvest strategies. This implies that risk may not be reduced through the use of these strategies. However, the standard deviation for the pre-harvest strategies is somewhat


University of Guelph, Ridgetown Campus iii | P a g e

inflated due to the price volatility that occurred in 2008. If this year is omitted from the analysis, the standard deviation for these strategies would be lower than for the baseline strategy. The results also indicate that making multiple cash sales throughout the year does not generate significantly higher prices relative to the baseline strategy. Any price increases that occur in the months after harvest may be offset by storage and inventory costs. In addition, spreading cash sales out over the year does not reduce risk, as the standard deviation of returns is greater for these strategies as compared to the baseline, while there is little difference in the probability of generating a price below the cost of production. Table 1: Results of the simulation model by strategy, 1992-2008

a Figures are rounded off to the nearest cent. b Asterisks (*,**) indicate statistical significance at the 10% and 5% levels, respectively. Further analysis was conducted to determine whether differences in the relative effectiveness of strategies are evident across different types of market conditions. The model results were split into two groups – results for years in which pre-harvest prices are greater than costs of production and results for years in which prices are below costs of production throughout the primary pre-harvest periods (i.e., spring and early summer). The results for the high-price years are similar in nature to those of the initial analysis, where strategies that involve the use of futures contracts and options generated the highest average prices, but with much greater differences between pre-harvest strategies and the baseline strategy. Conversely, the results for the low-price

Average Standard ProbabilityStrategy Description Pricea Deviationa Below COP

($/bu) ($/bu) (%)

1 Baseline 3.21 - 0.61 64.72 Jan. cash sale 3.32 10.4 0.61 55.63 3 cash sales 3.29 7.8 0.67 61.04 5 cash sales 3.31 9.4 0.86 68.85 Forward contract 1 3.32 11.2 * 0.55 68.06 Forward contract 2 3.38 17.0 ** 0.51 68.57 Forward contract 3 3.39 17.5 * 0.63 76.58 Futures contract 1 3.53 31.8 * 0.93 58.19 Futures contract 2 3.52 31.1 1.34 68.510 Futures contract 3 3.54 32.8 * 1.13 65.611 Put option 1 3.39 17.9 * 0.77 64.712 Put option 2 3.49 27.5 1.18 65.113 Put option 3 3.44 23.1 * 0.96 65.514 Call option 3.26 5.1 0.73 62.615 Basis contract 3.38 16.4 ** 0.67 62.916 Forward & Futures 3.51 29.8 * 1.07 70.017 Forward & Put option 3.49 27.3 * 0.97 71.018 Dynamic 3.48 27.2 ** 0.92 58.7

DifferenceFrom

Baselineb

(cents/bu)


University of Guelph, Ridgetown Campus iv | P a g e

years indicate that only minimal differences occur in average prices across all strategies. None of the differences between the pre-harvest strategies and the baseline is statistically significant. Overall, the results indicate that pre-harvest marketing strategies for corn that use futures contracts and options are most effective relative to the baseline strategy. These strategies were found to return a higher average price than strategies where the entire crop is sold through cash sales, particularly in years where the pre-harvest prices are greater than the costs of production. As a result, in such years there may be significant benefits to using these pre-harvest strategies, rather than waiting for prices to get even higher later in the year. In years when prices are lower, and are below costs of production, pre-harvest strategies may not offer a significant improvement over the baseline strategy. The differences in the relative effectiveness of specific strategies between the high-price and low-price years imply that using only one specific strategy in every year may not be the best approach to marketing, even if the selected strategy has been found to perform better on average across all years compared to other strategies (in this case, Strategy #10). Rather, it may be prudent to use different strategies in response to the current set of market conditions.


University of Guelph, Ridgetown Campus v | P a g e

Table of Contents 1.0 Introduction ............................................................................................................. 1 1.1 Background .................................................................................................... 1 1.2 Literature Review ........................................................................................... 5 2.0 Methods .................................................................................................................. 8 2.1 Simulation Model Development ...................................................................... 8 2.2 Marketing Strategies ...................................................................................... 9 2.3 Additional Model Considerations .................................................................. 14 2.4 Methods of Analysis ..................................................................................... 15 3.0 Results .................................................................................................................. 16 3.1 Model Results across All Years .................................................................... 16 3.2 Model Results under Varying Market Conditions .......................................... 19 3.3 Strategy Rankings ........................................................................................ 21 3.4 Relative Performance of each Strategy ........................................................ 22 3.5 Summary of Results ..................................................................................... 23 4.0 Conclusions .......................................................................................................... 25 4.1 Recommendations ....................................................................................... 25 4.2 Limitations and Suggestions for Future Research ........................................ 26 4.3 Summary ...................................................................................................... 27 References .................................................................................................................... 28


University of Guelph, Ridgetown Campus vi | P a g e

List of Tables

Table 1: Summary statistics of historical daily cash prices for corn in Chatham-Kent .... 2 Table 2: Description of marketing strategies ................................................................ 11 Table 3: Results of the simulation model by strategy, 1992-2008 ................................ 17 Table 4: Results of the simulation model for the nine years in which pre-harvest prices are greater than costs of production .................................................... 20 Table 5: Results of the simulation model for the eight years in which pre-harvest prices are less than costs of production ......................................................... 21 Table 6: Strategy rankings based on average prices across all three analyses ........... 22 Table 7: Relative overall performance for each strategy .............................................. 23

List of Figures

Figure 1: Average cash prices for corn ($/bu), Chatham-Kent, 1992-2008 ................... 3


University of Guelph, Ridgetown Campus 1 | P a g e

1.0 Introduction One of the most difficult decisions that corn producers must make is how and when to market their crop to ensure that they receive prices that can enable them to be profitable, or at the very least, competitive. Marketing decisions may often be avoided or delayed due to anxiety about selling too early or too late and, in some cases, due to a lack of knowledge about appropriate strategies and tools. As a result, these decisions may not be based on sound marketing fundamentals, and may involve, for example, only making sales when necessary to meet financial obligations. Such strategies may not be effective for earning the best possible prices. Thus, without increased focus on the development of marketing strategies, it will become more difficult to be profitable, particularly in the face of increasing costs of production. Because corn prices can vary significantly throughout the year, it is important that producers implement a marketing strategy that can allow them to take advantage of periods of higher prices. However, with seemingly endless possibilities in terms of marketing tools and timing of sales, the development of a marketing strategy can be very difficult, particularly due to a lack of information on the relative effectiveness of various marketing strategies. To address this issue, this report involves the development and use of a simulation model to examine and compare various marketing strategies for grain corn, for the purpose of determining the strategies that tend to generate the highest average prices per bushel. The strategies that are examined are comprised of marketing tools that are regularly utilized by corn producers in Ontario, such as forward contracts, basis contracts, futures contracts, options, cash sales, and combinations of these tools. The marketing simulation model uses historical pricing data to determine the range of prices received under various marketing strategies in each year. The results of this model may demonstrate to corn producers which marketing strategies and marketing tools tend to generate greater returns. Based on these results, comparisons can be made among strategies that use primarily cash sales versus those that involve marketing a substantial portion of the crop during pre-harvest periods using tools such as forward contracts, futures contracts, and options. This information may benefit producers in the development of marketing strategies that may increase their returns. Prior to the description of the simulation model and its results, some background information is given in the following section on price trends as well as the potential advantages and risks associated with various marketing tools. In addition, as a means of providing context for this study, section 1.2 describes the findings from a number of related studies on marketing strategies for agricultural commodities. 1.1 Background There can be significant variation in the price of corn throughout the year. Variation in cash prices for the municipality of Chatham-Kent in southwestern Ontario between 1992 and 2008 is demonstrated in table 1, which provides the average, minimum, and maximum daily prices for each year. It is evident from the range, which is the difference between the maximum and minimum prices, that there is often considerable variation in



the price of corn. In fact, there are only five years over this 17-year period in which the range of prices is less than $1.00 per bushel, and only two years in which the range is less than $0.88 per bushel. This suggests that while the average price may be low in many years relative to the cost of production, there are often opportunities throughout the year to market corn at a much higher price. There may be specific strategies or specific marketing tools that tend to be better suited for taking advantage of these opportunities. Table 1: Summary statistics of historical daily cash prices for corn in Chatham-Kent

Source: Ontario Commodity Reports There also tends to be certain price trends that occur in the relative price level during the year. For example, prices tend to be lower during the harvest period and tend to be higher during the spring and early summer. Wisner, Blue, and Baldwin (1998) suggest that the pre-harvest price premiums may reflect uncertainty in expectations of domestic and foreign production. As harvest approaches and production expectations become more certain, prices may decline. Such trends are evident in figure 1, which displays the annual trend for cash prices in Chatham-Kent, averaged across the time period between 1992 and 2008. However, because these trends do not occur consistently each year, structuring a marketing strategy around these trends may not be successful every year. For example, in 1995 and in 2006 the price of corn increased substantially through the harvest period and continued to climb after harvest, such that the post-harvest price was much higher than prices during the previous spring and summer. But even though these unexpected scenarios may occur from time to time, strategies designed to take advantage of these trends by selling in the pre-harvest period when

Year Average Minimum Maximum Range

1992 $2.75 $2.32 $3.20 $0.881993 $2.99 $2.73 $3.66 $0.931994 $3.29 $2.63 $3.88 $1.251995 $3.62 $2.84 $4.59 $1.751996 $5.43 $3.54 $7.40 $3.861997 $3.79 $3.24 $4.35 $1.111998 $3.24 $2.71 $3.89 $1.181999 $2.79 $2.46 $3.04 $0.582000 $2.81 $2.39 $3.42 $1.032001 $3.24 $2.75 $3.66 $0.912002 $3.56 $3.11 $4.38 $1.272003 $3.48 $2.89 $3.95 $1.062004 $3.31 $2.23 $4.32 $2.092005 $2.47 $2.25 $2.91 $0.662006 $2.69 $2.09 $3.80 $1.712007 $3.72 $3.23 $4.44 $1.212008 $4.89 $3.59 $6.61 $3.02



prices are generally higher may still in the long run prove to be significantly more successful. These strategies also help to avoid the speculative position and the associated price risk that producers take on when crops are planted without a known selling price. Figure 1: Average cash prices for corn ($/bu), Chatham-Kent, 1992-2008

Source: Ontario Commodity Reports However, while there may be advantages to using marketing tools such as forward contracts, futures contracts, and options to take advantage of pre-harvest marketing opportunities, there are different risks associated with these tools. With forward contracts, there is production risk. Since the contracted amount must be delivered, producers generally do not forward contract their entire expected production, as they would not be able to fulfill their contract in the event of a yield shortfall. As a result, the price for a relatively large share of their production remains unprotected, which leaves producers in a speculative position for this portion of their production. The use of futures contracts can help to avoid this speculative position. Since these contracts can simply be offset without the requirement for delivery, there is no production risk associated with hedging through futures contracts. This enables producers to hedge their entire expected production, as a yield shortfall would not affect

$2.90

$3.00

$3.10

$3.20

$3.30

$3.40

$3.50

$3.60

$3.70



their ability to fulfill the contract. However, with futures contracts there is the risk of margin calls. If the price moves against a futures position, the producer would need to put more money into their trading account. In addition, just to enter into a futures position, a significant amount of funds must be put into the account to cover the margin requirement. The use of futures contracts, unlike forward contracts, also leaves producers open to basis risk. Since futures contracts only protect the futures component of the price, producers are not protected from unfavourable changes in the basis. For example, producers that use short hedges to protect against futures price declines would be negatively impacted by the weakening of the basis that often occurs around the harvest period. But in general, basis risk tends to be considerably less than price risk. While both forward contracts and futures contracts protect against unfavourable price changes, they do not allow producers to benefit from price increases that may occur after the contracts have been entered. Conversely, the use of options does allow producers to benefit from such increases. However, this comes at a price, as the cost of purchasing options can be quite high, particularly in volatile markets. Similar to futures contracts, options can be used to protect the futures component of the price for the entire expected production, as there is no production risk associated with the use of options. Producers in Ontario who use both futures contracts and options face exchange rate risk, as the cost of hedging with these tools as well as the gains and losses that accrue from the use of these tools vary with the exchange rate. In contrast to futures contracts, the use of options avoids margin risk, as purchasers of options are not subject to margin calls even if the price moves against their position. In these situations, rather than realizing a loss in the futures market (as with futures contracts), the options can simply be allowed to expire. As a result, the use of put options allows producers to establish minimum prices by protecting against price declines without giving up the opportunity to benefit from price increases. As such, with options producers will be better off in years when prices move against their position (i.e., prices increase) as compared to the use of futures contracts. But due to the cost of option premiums, in years when prices move with their position the returns from options would be slightly less than from futures contracts. Thus, the use of options reduces price variability relative to the use of futures – the lows are not as low, but the highs are not as high. Among these pre-harvest marketing tools, producers tend to use forward contracts substantially more than futures contracts or options. Forward contracts are more easily understood by producers, and they are easier and cheaper to enter into as compared to futures contracts and options. Forward contracts are made through local elevators, with which producers are very familiar, while futures contracts and options require the services of a broker, and the actual marketing activities may be conducted far from home. In addition, these tools have costs associated with them such as commission costs, premium costs for options, and capital costs of margin requirements for futures contracts. These factors have contributed to a much greater use of forward contracts as part of a pre-harvest marketing strategy rather than the use of futures contracts or options.



When no pre-harvest marketing tool is used to protect a selling price, the producer would have a speculative cash market position. This speculative position extends beyond the harvest period if the crop is not sold at this time. Since prices are often relatively low at harvest time, producers will store their crop and delay marketing with the hope that prices will increase after the harvest period. To reduce the price risk associated with this speculative position, there are post-harvest marketing tools that can be used for production that has not yet been sold. One alternative is to sell the remaining crop and purchase call options to establish a minimum price and be able to benefit from a future price increases. This reduces price risk associated with the futures component of the price, but the producer must pay a premium to receive this potential benefit. Another alternative is to use a basis contract, which locks in the basis component of the price but allows for benefiting from increases in the futures price. This reduces basis risk, but price risk associated with the futures component of the price still remains. Both of these alternatives reduce storage costs and the risk of crop spoilage while in storage. 1.2 Literature Review There is an existing body of literature that has examined price trends and compared marketing strategies for agricultural commodities, often using some type of simulation or modeling approach. Using either actual or simulated market price data, these studies tend to use harvest sales as a baseline strategy to which the level of returns (and often variance) from other marketing strategies are compared. The results of these studies are largely mixed in terms of the relative effectiveness of different strategies, and have generated considerable debate as to whether a market that is presumably efficient offers opportunities to consistently increase expected returns through alternative (e.g., pre-harvest) marketing strategies. A number of studies have found that alternative strategies to selling at harvest can increase returns for producers. Wisner, Blue, and Baldwin (1998) simulated returns under various marketing strategies for model farms in Iowa and Ohio, using actual price data from 1979 to 1996. They found that corn and soybean new-crop futures prices tend to be biased upward early in the year, generally peaking before mid-July. This upward bias implies that hedges made during these times can be profitable, and would most likely result in greater returns compared to waiting until harvest to market the crop. Accordingly, the results of their research indicated that several strategies that used futures and options in the pre-harvest period had significantly greater returns relative to cash sales at harvest. Curtis et al. (1987) examined a large number of marketing strategies for soybeans that used cash sales and various types of futures hedges. The results of this study indicated that incorporating hedges into marketing strategies increased returns for producers. Frank et al. (1989) expanded this research to include options in the set of marketing strategies under examination. They found that strategies utilizing options increased returns by up to 6.7% and reduced the variance of returns by up to 33%.



However, not all studies have concluded that alternative strategies to selling at harvest will be more effective risk management strategies. Using Illinois corn and soybean prices from 1985 to 1997, Zulauf and Irwin (1998) found that, while average returns were higher for pre-harvest marketing strategies that included the use of futures and options, these returns were not significantly different from returns from selling at harvest. They concluded that routinely selling before harvest would not increase income over selling at harvest. They also suggested that although new crop futures prices declined between May 1 and November 1 in 65% of the years examined, it is difficult for producers to fully benefit from this trend unless they can predict the years in which this pattern will occur. Using simulated market prices (rather than actual market prices), Peterson and Tomek (2005) found that strategies using diversifying cash sales or futures positions do not necessarily increase returns. Additionally, in the case of cash sales, any increases in price that may occur after harvest are also accompanied by an increased variance. Peterson and Tomek (2007) expanded upon this research by using simulated prices to compare returns from specific marketing strategies over the long run and in individual 40-year periods. They found that pre-harvest conditional hedging strategies, while yielding lower variance of returns, resulted in small but insignificant increases in returns. These hedging strategies generated returns that were higher than the base strategy (cash sale at harvest) in only just over half the years of a typical lifetime. Other studies have indicated a lack of consistency in situations where differences are found in returns generated by specific strategies. Tomek and Peterson (2005) note that commodity markets, even when efficient, can display diverse price behaviour over time. This can cause the relative performance of alternative marketing strategies to vary from year to year. This study found that strategies that tend to generate lower average returns would actually perform relatively well in 30-50% of the years modeled. The relative performance in the mean and variance of returns may also vary depending on location (Tennyson, 1996). The studies described above have focused primarily on strategies and pricing data for farming operations in the United States. However, cash price determination differs in Canada due to the influence of the exchange rate. This may have a significant impact on the relative performances of specific marketing strategies, and may affect the ability to apply the results from American studies to Canadian situations. One exception from the set of American-based studies is a study by Martin and Hope (1984), which developed a set of marketing strategies for corn producers in Ontario and compared the risk and returns associated with these strategies. This study found that using fixed dates for making sales throughout each year did not improve risk or returns as compared to selling the entire crop at harvest. The results also indicated that returns were higher when pricing was done later in the year, and that using target prices when considering marketing decisions increased returns and reduced risk. However, the current study takes a different approach to comparing returns from different marketing strategies by implementing a simulation model, which makes this



study more dynamic than the study by Martin and Hope. The use of more recent pricing data may generate results that better reflect the relative performance of strategies under current market conditions. In addition, this study includes options in the set of strategies for comparison. The development of the simulation model, as well as the description of the strategies that are compared by this model, are given in the following section of this report. Section 3 provides an overview of the results of the simulation model, while the final section offers some recommendations based on these results and discusses the limitations of the study.



2.0 Methods The methods for this study involved the development of a model using Microsoft Excel and Visual Basic. This model incorporates a substantial amount of decision rules and ‘if’ statements to simulate a set of marketing strategies and to generate average prices for each strategy across a 17-year time period. The specifications of this model are described in the following section. The strategies simulated by this model, which incorporate tools such as forward contracts, futures contracts, and options, are outlined in section 2.2. The final section describes some additional considerations for this model. 2.1 Simulation Model Development The simulation model was developed for the purpose of deriving and comparing average annual prices for specific marketing strategies based on historical price data. This data was drawn from historical daily Ontario Commodity Reports, which provide nearby and deferred corn futures prices as well as old-crop and new-crop cash prices for specific locations from 1992 to the present time. The location from which prices for cash sales and forward contracts in the model are derived is Chatham-Kent1. Data was also drawn from historical Chicago Board of Trade data, which provides margin requirements and daily option premiums. Specific strategies, including timing of contracts and sales, were developed through consultations with marketing specialists and producers to ensure that these strategies are representative of those currently used by corn producers. To create simulation models, a number of assumptions have to be made. In some cases, these assumptions may be limiting and may impede the ability of the model to reflect reality. However, they are necessary in order to conduct analysis from which useful results can be derived. Since a model can never fully depict human decision-making processes, a number of decision rules and other specifications must be put in place. Some of the assumptions and specifications imposed for this model are described below. The prices derived from the historical data for the marketing strategies are not based on a specific date of sale; rather, a price is randomly selected from a range of dates, based on the criteria outlined for each marketing strategy. Using a range of dates rather than just one date allows for a range of prices to be used for each strategy, which better represents the returns from marketing activities that would actually occur during that time period. This also avoids the potential issue of selecting a date on which the price is very high or very low relative to the average price during that time period, which may skew the comparison of prices generated by different strategies. Sales or contracts for each strategy are made over the same range of dates in every year. For simplicity, there are five time periods during which sales or contracts are made for this set of 1 The use of Chatham-Kent prices rather than, for example, average prices across Ontario would not affect the results of the analysis, as any differences between these sets of prices should be consistent throughout the period of observation.



strategies, with each period being three weeks in length: harvest (October 21 – November 10), early in the new year (January 5 – 25), spring (April 10 – 30), early summer (June 20 – July 10), and late summer (August 10 – 30). While marketing of corn may be done at any time throughout the year, these periods tend to be times when a greater amount of corn is marketed by producers in Ontario. For marketing strategies that involve making multiple sales at different points in time over the season, some variation is permitted in the amount of the crop that is sold or contracted at each point in time. In these cases, the share of the crop that is marketed is randomly selected from specified ranges (e.g., 40% to 60%), where the percentage selected is based on the level of the current price relative to the cost of production (COP)2. For most strategies, if the current price is less than 95% of the COP, the percentage marketed is selected from the bottom third of the specified range, while if the current price is greater than 105% of the COP, the percentage is selected from the top third of the specified range. This variation is permitted to account for the likelihood that producers would sell more of their crop when prices are higher and less of their crop when prices are lower. Generally, the share of the crop that producers sell through forward contracts and futures contracts is based on expected production, while actual production may be somewhat different due to yield variation from year to year. But for simplicity, this model assumes that actual production is equal to the expected production. 2.2 Marketing Strategies There are eighteen marketing strategies that have been developed for analysis by the simulation model. This set of strategies, as described below and outlined in table 2, involves the use of specific marketing tools at set time periods throughout the year, including cash sales (Strategies #1 – 4), forward contracts (#5 – 7), futures contracts (#8 – 10), options (#11 – 14), basis contracts (#15), and combinations of these tools (#16 – 17). The final strategy (#18) is a dynamic strategy where the tools used depend on market conditions at specific times of the year. The development of these strategies, which are assumed to be followed consistently in each year, allows for the direct comparison of returns over time from specific strategies. Cash Sale Strategies There are four strategies in which the entire crop is sold using only cash sales. The prices for cash sales are based on the old-crop prices for Chatham-Kent. Two cash sale strategies involve selling the entire crop at one point in time, while two strategies involve making multiple sales throughout the year as a means of diversification. In Strategy #1, the entire crop is sold on the cash market at harvest time in each year. This strategy is used as the baseline scenario, to which all other strategies are compared. Strategy #2 involves selling the entire crop on the cash market early in the new year. Sales are often 2 Cost of production data is derived from Ontario Ministry of Agriculture, Food & Rural Affairs’ annual Field Crop Budgets (Publication 60).



made at this time for tax-related purposes. In Strategy #3, cash sales are made at three different times (harvest, new year, and summer), with a larger share sold early in the new year. Strategy #4 also involves diversifying cash sales, with sales made at five different times throughout the year. Smaller shares of the crop are sold after harvest and early in the new year, while a greater share of the crop is stored with the expectation of higher prices later in the year. Forward Contract Strategies There are three strategies that focus primarily on the use of forward contracts. The prices available for forward contracts are based on new-crop prices for Chatham-Kent. Strategy #5 involves forward contracting between 40% and 60% of the crop in the spring and selling the remainder of the crop on the cash market at harvest. Strategy #6 is similar, but uses different time periods. The forward contract is made in early summer, and the remainder of the crop is sold on the cash market in the new year. Strategy #7 involves forward contracting 25-35% of the crop at two different times (spring and early summer), with the remainder of the crop sold on the cash market at harvest and in the new year. Futures Contract Strategies There are three strategies that incorporate short hedges using futures contracts. These short hedges are carried out by selling December corn futures contracts. The timing of the contracts is similar to the timing of the forward contracts in Strategies #5-7. Strategy #8 involves selling a futures contract in the spring for 75-100% of the crop. This short futures position is offset at harvest by buying December (nearby) corn futures contracts, at which point a cash sale is made. Strategy #9 is similar, but with the short futures position taken in early summer rather than the spring. Strategy #10 involves selling futures contracts at two different times, spring and early summer, each for 40-60% of the crop. Again, the crop is sold on the cash market at harvest when the futures contracts are offset. Options Strategies Of the four options strategies, three incorporate put options while the fourth makes use of a call option. The three put option strategies correlate directly with the futures contract strategies (#8-10), where put options are purchased rather than selling futures contracts. Put options are purchased for December corn in the spring for Strategy #11, in early summer for Strategy #12, and at both time periods for Strategy #13. In each case, the options are offset (if profitable) at harvest by selling the same December corn options at the current premium level, at which time the crop is sold on the cash market. Strategy #14 uses call options, where the crop is sold on the cash market at harvest and call options for May corn are purchased for 50-75% of the crop to take advantage of potential price increases after harvest. For this strategy, if the price is relatively low, the share of the crop for which call options are purchased will be at the upper end of the range, in order to allow for a greater share of the crop to benefit from potential future price increases. If profitable, the call options are offset the following spring.



Table 2: Description of marketing strategies3

3 Three strategies (#3, #4, and #18) had to be slightly adjusted for the 2008 marketing year, as prices were not yet available for the summer of 2009.

Strategy Tool Time Period Share of Crop

1 Cash sale Oct. 21 – Nov. 10 100%

2 Cash sale Jan. 5 – Jan. 25 100%

3 Cash sale Oct. 21 – Nov. 10 20-30%Cash sale Jan. 5 – Jan. 25 40-60%Cash sale Jun. 20 – Jul. 10 10-40%

4 Cash sale Oct. 21 – Nov. 10 10-20%Cash sale Jan. 5 – Jan. 25 10-20%Cash sale Apr. 10 – Apr. 30 20-30%Cash sale Jun. 20 – Jul. 10 20-30%Cash sale Aug. 10 – Aug. 30 10-40%

5 Forward contract Apr. 10 – Apr. 30 40-60%Cash sale Oct. 21 – Nov. 10 40-60%

6 Forward contract Jun. 20 – Jul. 10 40-60%Cash sale Jan. 5 – Jan. 25 40-60%

7 Forward contract Apr. 10 – Apr. 30 25-35%Forward contract Jun. 20 – Jul. 10 25-35%Cash sale Jan. 5 – Jan. 25 15-25%Cash sale Apr. 10 – Apr. 30 5-35%

8 Futures contract - short Apr. 10 – Apr. 30 75-100%Futures contract - long Oct. 21 – Nov. 10 75-100%Cash sale Oct. 21 – Nov. 10 100%

9 Futures contract - short Jun. 20 – Jul. 10 75-100%Futures contract - long Oct. 21 – Nov. 10 75-100%Cash sale Oct. 21 – Nov. 10 100%

10 Futures contract - short Apr. 10 – Apr. 30 40-60%Futures contract - short Jun. 20 – Jul. 10 40-60%Futures contract - long Oct. 21 – Nov. 10 80-100%Cash sale Oct. 21 – Nov. 10 100%

11 Put option Apr. 10 – Apr. 30 75-100%Offset option (if profitable) Oct. 21 – Nov. 10 75-100%Cash sale Oct. 21 – Nov. 10 100%



Table 2: Description of marketing strategies (cont’d)

Basis Contract Strategy In Strategy #15, a cash sale for 25-50% of the crop is made at harvest and a basis contract is signed early in the new year for the remainder of the crop, using the old-crop basis for Chatham-Kent. The basis contract is priced out in the spring. Combination Strategies Strategies #16 and #17 use a combination of marketing tools to sell the entire crop before harvest. In Strategy #16, 40-60% of the crop is forward contracted in the spring and a short hedge is made with futures contracts for the remainder of the crop in early summer. The futures contracts are offset at harvest and a corresponding cash sale is made at this time. Strategy #17 is similar, but with put options purchased rather than futures contracts sold.

Strategy Tool Time Period Share of Crop

12 Put option Jun. 20 – Jul. 10 75-100%Offset option (if profitable) Oct. 21 – Nov. 10 75-100%Cash sale Oct. 21 – Nov. 10 100%

13 Put option Apr. 10 – Apr. 30 40-60%Put option Jun. 20 – Jul. 10 40-60%Offset option (if profitable) Oct. 21 – Nov. 10 80-100%Cash sale Oct. 21 – Nov. 10 100%

14 Cash sale Oct. 21 – Nov. 10 100%Call option Oct. 21 – Nov. 10 50-75%Offset option (if profitable) Apr. 10 – Apr. 30 50-75%

15 Cash sale Oct. 21 – Nov. 10 25-50%Basis contract Jan. 5 – Jan. 25 50-75%Price out basis contract Apr. 10 – Apr. 30 50-75%

16 Forward contract Apr. 10 – Apr. 30 40-60%Futures contract - short Jun. 20 – Jul. 10 40-60%Futures contract - long Oct. 21 – Nov. 10 40-60%Cash sale Oct. 21 – Nov. 10 100%

17 Forward contract Apr. 10 – Apr. 30 40-60%Put option Jun. 20 – Jul. 10 40-60%Offset option (if profitable) Oct. 21 – Nov. 10 40-60%Cash sale Oct. 21 – Nov. 10 100%

18 See text for description



Dynamic Strategy Strategy #18 is a dynamic strategy for which the timing of sales and the tools utilized can vary from year to year based on relative pricing conditions at specific decision points. This strategy is included to allow for more flexibility in the type and timing of marketing activities and to allow for more responsiveness to current market conditions, in contrast to all other strategies in which the same tools are used consistently from year to year regardless of the current market situation. As such, this strategy may better reflect the marketing decisions that producers would make in response to current market conditions. While an effective comparison of strategies requires consistency in marketing activities across all years, this strategy deviates from this approach to allow for the comparison of the regimented strategies with one in which flexibility is permitted, as a means of context. This strategy includes seven decision points at which sales or contracts can be made, ranging from the spring before harvest to early summer after harvest. Decision Point #1 occurs in the spring before harvest. If the new-crop price has increased since the beginning of the year, and if the price is greater than the cost of production, then a forward contract is signed (the greater the difference between price and COP, the larger the share of the crop is contracted, up to 60%). If prices are less than the COP, no forward contract is signed. Decision Point #2 occurs in early summer. If the price has increased since the spring, and if the price is greater than the COP, then a forward contract may be signed (up to 60% of the crop). However, if a forward contract had been signed in the spring, then a forward contract would not be signed at this point, but instead a put option would be purchased for the remainder of the crop. Decision Point #3 occurs in late summer prior to harvest. If the price has increased since the spring, if the price is greater than the COP, and if a put option was not purchased at Decision Point #2, then a put option would be purchased for the remainder of the crop (if a forward contract had previously been signed) or for the entire crop (if no forward contract had been signed). Decision Point #4 occurs at harvest. If a put option was previously purchased, then it would be offset (if profitable) at this point and a corresponding cash sale would be made. If no put option had been purchased and no forward contracts had been signed, a cash sale for 20-30% of the crop would be made if the price is greater than the COP. If a forward contract had been signed, then the remainder of the crop would not be sold at this point. Decision Point #5 occurs early in the new year. If forward contracts had been signed or put options had been purchased, and if some of the crop has not yet been sold, a cash sale for 10-20% of the crop would be made if the price is greater than the COP. If no contracts or options were used, a cash sale for 40-60% of the crop would be made if the price is greater than the COP. If the price is less than the COP, a cash sale would be



made for 40-60% of the crop and a call option would be purchased for the same share of the crop. Decision Point #6 occurs in the spring following harvest. If a call option had previously been purchased, it would be offset if the futures price has increased. Any remaining crop would be sold if the price is greater than the COP. Decision Point #7 occurs in early summer. At this point, any remaining crop is sold, even if the price is less than the COP. Linked Strategies Links can be drawn between the sets of pre-harvest strategies, which are designed similarly to allow for direct comparison of returns using forward contracts, futures contracts, and put options. Strategy #5 (forward contract) is linked to #8 (futures contract) and to #11 (put option), as all three strategies involve the use of a pre-harvest marketing tool in the spring. Similarly, Strategy #6 is linked to #9 and to #12, where pre-harvest marketing activities occur in the early summer, while Strategy #7 is linked to #10 and to #13, where pre-harvest marketing tools are used in both the spring and early summer. The results of the model for these sets of linked strategies may indicate whether one of these marketing tools tends to generate greater returns as compared to the other tools. 2.3 Additional Model Considerations For strategies that involve selling a share of the crop later in the year, storage costs are taken into consideration and deducted from the selling price. Interest is also charged on inventory when the crop is not sold at harvest, using the value of the crop at harvest and an interest rate of the prime business rate plus one percent. Strategies that involve the use of futures contracts factor in the interest costs on funds required for trading accounts and margin calls. The margin requirements are derived from the Chicago Board of Trade’s historical records of corn margins. When daily fluctuations in the futures price cause the funds in the trading account to drop below the level of the maintenance margin requirement, margin calls would occur. All trading account transactions are made in U.S. currency. However, when funds are added to the trading account, they are converted to Canadian currency using the exchange rate for the day in which the transaction takes place. Interest is charged daily on the total amount of funds (in Canadian currency) in the trading account, using an interest rate of the prime business rate plus one percent. Commission fees are also accounted for and deducted from the price generated for strategies that use futures and options. Any gains or losses from futures contracts are converted to Canadian currency using the exchange rate at the time the futures position is offset. For strategies that incorporate options, a strike price close to the current futures price is selected. While there are many strike prices to choose from, ranging from well below the current futures price to well above, the selection of a strike price close to the futures



price provides a balance between the higher cost of strike prices that are in-the-money and the lower price protection level of strike prices that are out-of-the-money. The option premiums associated with these strike prices ranged for the most part from $0.10 to $0.30 per bushel. Options are offset if there is any value left at the time the crop is harvested and sold on the cash market, as long as the premium is greater than the commission cost that would be incurred from selling the option. Premiums are converted to Canadian currency using the exchange rate at the time the premium is paid or received. The returns for each strategy in this model are determined on a per-bushel basis. These returns take into account all revenues received from sales or contracts as well as all costs associated with each strategy, such as interest costs, storage costs, and commission costs. These revenues and costs are each converted to a per-bushel basis. Thus, the average price per bushel that is generated for each strategy is calculated as the selling price for corn plus any profits from the futures market less all marketing costs associated with the strategy. 2.4 Methods of Analysis Due to the semi-random selection of some elements that are inherent in the marketing strategies, such as date of sale and percentage of crop sold, the simulation model is run 1,000 times for each strategy for each year to account for the variance in these elements. The average prices generated by the simulation model for each marketing strategy are compared within each year as well as across all years. T-tests are conducted to determine whether statistically significant differences in average prices exist among the marketing strategies. In addition, comparisons are made between the baseline strategy and pre-harvest strategies that involve marketing a substantial portion of the crop through forward contracts, futures contracts, or options. In addition to generating average prices, the simulation model also provides a measure of risk associated with each strategy. There are two factors that serve as a proxy for risk. The first factor is the standard deviation, which is the square root of variance, a measure of the dispersion in the average prices for each strategy. In general, a larger standard deviation indicates greater variation in prices, which results in greater risk. Another measure of risk, following from Martin and Hope (1984), is the probability that the price that is generated will be below a certain threshold – in this case, the cost of production. Both of these measures of risk are estimated by the simulation model. The results of the model, for both the average prices and the measures of risk generated for each strategy, are provided in the next section.



3.0 Results The analysis of the results of the simulation model is conducted using several different approaches. First, the results are analyzed across the entire study period (1992-2008) to determine the strategies that generate the highest average prices when followed consistently across all years. The results are then broken up into two groups based on the level of pre-harvest prices relative to the costs of production (i.e., higher vs. lower) to determine whether there are differences in the relative effectiveness of strategies across years with varying market conditions. The strategy rankings in terms of average prices are then compared across both sets of results. Finally, the strategy rankings within each year are examined to determine the degree to which strategies that generate relatively lower average prices can still be effective in certain years. 3.1 Model Results across All Years The average prices and measures of risk for each strategy are generated through running the simulation model 1,000 times for each of the eighteen marketing strategies. T-tests are conducted to determine whether the average prices for Strategies #2 through #18 are significantly greater than those generated by Strategy #1, the baseline strategy where the entire crop is sold at harvest. Table 3 provides a summary of the relative returns and risk for each of the strategies, including the average prices across all years, the differences in average prices from the baseline strategy, the standard deviation of prices for each strategy, and the probability of generating prices below the costs of production. The strategies with the highest average prices across all years and the greatest price premiums over the baseline are Strategies #10, #8, #9, and #16, all of which incorporate short hedges with futures contracts. These strategies generated prices that are between 29.8 and 32.8 cents per bushel higher than the baseline strategy. The put option strategies (Strategies #11, #12, #13, and #17) also have much greater returns than the baseline, generating price premiums ranging from 17.9 to 27.5 cents per bushel, only slightly lower than those of the futures contracts strategies. The dynamic strategy (#18), which was included to provide a means of context for the results of the other more structured strategies, also performed well relative to the baseline, with a price premium of 27.2 cents per bushel4. While all other strategies generated average prices that are greater than the baseline strategy, the price premiums are somewhat lower. The average prices for forward contract strategies (aside from those in which forward contracts are used in combination with futures contracts or options) are 11.2 to 17.5 cents per bushel higher than selling the entire crop at harvest. The main reason that the price premiums for these strategies are lower than other pre-harvest marketing strategies is that only 40% to 70% of total 4 With many different ways to construct dynamic strategies, additional research would be necessary to determine whether the performance of the dynamic strategy in this study is representative of the range of results that could be generated by such strategies.



production is marketed through forward contracts in these strategies. As a result, the higher prices that tend to be available during the pre-harvest marketing periods are not received for the entire production, as the remaining production is sold through cash sales at or shortly after harvest. Conversely, strategies that use futures contracts or options can take advantage of the higher prices for up to 100% of total production. Strategies that use cash sales at different times of the year are found to have only marginal improvements in price relative to the baseline strategy, ranging from 7.8 to 10.4 cents per bushel. Table 3: Results of the simulation model by strategy, 1992-2008

a Figures are rounded off to the nearest cent. b Asterisks (*,**) indicate statistical significance at the 10% and 5% levels, respectively. Aside from the baseline strategy, the strategy that generated the lowest average price per bushel is Strategy #14, which involved selling the crop at harvest and purchasing a call option. Part of the reason this strategy performed relatively poorly is that there were few years in which the call option was profitable enough to offset in the spring. This may have been due in part to the parameters of the strategy, in which a May call option was purchased at harvest and potentially offset in the spring (April 10 – 30). By selecting a range of dates for the offset period that is so close to the expiration date of the option, this could have reduced the potential for the option to be worth offsetting. Generally, option premiums will decline substantially in the two months prior to expiry as the time value component of the premium erodes. Thus, when close to expiry, the option would only be worth offsetting if it were in-the-money. An alternative to this strategy that may


($/bu) ($/bu) (%)

1 Baseline 3.21 - 0.61 64.72 Jan. cash sale 3.32 10.4 0.61 55.63 3 cash sales 3.29 7.8 0.67 61.04 5 cash sales 3.31 9.4 0.86 68.85 Forward contract 1 3.32 11.2 * 0.55 68.06 Forward contract 2 3.38 17.0 ** 0.51 68.57 Forward contract 3 3.39 17.5 * 0.63 76.58 Futures contract 1 3.53 31.8 * 0.93 58.19 Futures contract 2 3.52 31.1 1.34 68.510 Futures contract 3 3.54 32.8 * 1.13 65.611 Put option 1 3.39 17.9 * 0.77 64.712 Put option 2 3.49 27.5 1.18 65.113 Put option 3 3.44 23.1 * 0.96 65.514 Call option 3.26 5.1 0.73 62.615 Basis contract 3.38 16.4 ** 0.67 62.916 Forward & Futures 3.51 29.8 * 1.07 70.017 Forward & Put option 3.49 27.3 * 0.97 71.018 Dynamic 3.48 27.2 ** 0.92 58.7

DifferenceFrom

Baselineb

(cents/bu)



have generated different results would have been to use a call option for a later month, such as July. The premium for this option would be slightly higher than for a May option when purchased, but there would have been a greater likelihood that the option would be worth offsetting, since the period in which the offset would occur is well before expiry which would leave plenty of time value in the premium. Further research may be necessary to explore different alternatives for the use of call options. The results in table 3 also indicate whether the differences in prices relative to the baseline are statistically significant, as determined by the t-tests. Many of the differences are found to be significant, though several of these differences are only significant at the 10% level. In some cases (e.g., Strategies #9 and #12), even with much greater prices than the baseline, the differences are not significant. The reason for this lack of significance is that these strategies have large variance, where the returns from year to year encompass a relatively wide range. The significance of the differences between the average prices generated by the pre-harvest strategies and those of the baseline strategy is due in large part to the price volatility in 1996 and 2008. In these years, the prices during the spring and early summer pre-harvest periods were substantially greater than the prices at harvest, as the prices declined considerably prior to the harvest period. As a result, pre-harvest strategies generated much higher prices than the baseline strategy for these years, which increased the average differences in prices across the entire study period. Without these years included in the model, the pre-harvest strategies would still have generated higher average prices than the baseline, but the differences would not have been as statistically significant in all cases. The extremely high pre-harvest prices in these years also contributed considerably to higher standard deviations for the pre-harvest strategies. In general, the strategies using futures contracts tend to have a much higher standard deviation than do many of the other strategies. This is due in part to the nature of futures positions. When the futures price moves with a position, the profits made from the futures contracts can add substantially to the returns for that year. However, when the futures price moves against a position, the associated losses from the futures contracts and interest paid on margin calls can detract from the returns for that year. As a result, the range in average prices generated by strategies using futures contracts can be very wide. Strategies using options do not experience quite as much variance as compared to those using futures contracts due in part to the nature of options. Once options are purchased, if the price moves against the option position, no losses would be incurred as the option can be allowed to expire. In addition, the premium paid for options also detracts somewhat from the profits that are made when the futures price moves with the option position. As a result, when comparing futures contracts to options, the highs tend to be higher with futures contracts but the lows also tend to be lower, which causes the variance to be greater for futures contracts. This is evident when comparing similar strategies that use futures and options. Among the three sets of linked strategies,



Strategy #8 has a greater standard deviation than Strategy #11, Strategy #9 has a greater standard deviation than Strategy #12, and Strategy #10 has a greater standard deviation than Strategy #13. There are few strategies that had substantial improvements over the baseline strategy with respect to the probability of generating prices that are less than the costs of production. In fact, there are a number of strategies for which this probability is greater, even though the average prices are much greater than under the baseline strategy. This suggests that the price premiums of these alternative strategies occur primarily in years when prices are relatively high. These results differ from those of Martin and Hope (1984), which found that strategies that had greater returns than the baseline tended to have lower probabilities of returns below a threshold based on costs of production. One of the strategies for which this probability improved relative to the baseline is the dynamic strategy, which had one of the lowest probabilities across all strategies. This suggests that allowing for flexibility in the marketing strategy in response to current market conditions may reduce the level of risk. But overall, the relatively high probabilities (between 55.6% and 76.5%) that prices are below the costs of production provide an indication of the difficulties that corn producers have experienced in recent years in terms of profitability. This further reinforces the need for the development of an effective marketing strategy. 3.2 Model Results under Varying Market Conditions While the initial analysis assumes that strategies employing pre-harvest marketing tools, such as forward contracts, futures contracts, and options, follow the same pattern of marketing in every year regardless of the current market prices, this may not be an appropriate course of action in actuality. Generally, producers are more likely to pursue pre-harvest marketing opportunities when they are able to lock in a price that will ensure that they make a profit. If prices are below their costs of production, they may be more likely to hold off with marketing their crop in the hope that prices will improve by harvest or in the months following harvest. Following from this concept, and considering that there are certain years in which specific strategies may perform well and other years in which they do not, further analysis is conducted to determine whether there are differences in the effectiveness of specific strategies across years with different market situations. The study period is broken up into two types of years – those with pre-harvest prices that are greater than the costs of production and those where the price remains below the costs of production throughout the primary pre-harvest marketing periods (i.e., spring and early summer). The years in which pre-harvest prices are greater than the costs of production include 1994-1998, 2000, 2004, 2007, and 2008. The average prices for each strategy as well as the differences from the baseline strategy are given in table 4. The differences that are significant are noted, based on the results of t-tests. The results are very similar to



those of the initial analysis in terms of the relative performance of the strategies, but the differences from the baseline for strategies that include pre-harvest marketing tools are much greater for the high-price years than across all years. The price premiums relative to the baseline are greatest for the strategies that use short hedges with futures contracts, ranging from 57.5 to 60.6 cents per bushel. The magnitude of these premiums is almost 30 cents greater than for the premiums across all years. The probabilities of generating prices that are below the costs of production are lower than in the initial analysis, but are still relatively high. Part of the reason for this is that the prices may not have been greater than the costs of production across the entire pre-harvest period. In addition, the marketing costs associated with various strategies that are deducted from the selling price may have caused the prices to drop below the costs of production. Table 4: Results of the simulation model for the nine years in which pre-harvest prices

are greater than costs of production

a Figures are rounded off to the nearest cent. b Asterisks (*,**) indicate statistical significance at the 10% and 5% levels, respectively. The results of the model for years in which the pre-harvest prices are less than the costs of production are given in table 5. The differences from the baseline in the average prices are minimal, and in some cases the average prices for alternative strategies are less than that of the baseline strategy. The only strategy for which the difference is statistically significant (at the 10% level) is Strategy #15, in which a basis contract is used. This suggests that in years where the market prices are relatively low


($/bu) ($/bu) (%)

1 Baseline 3.35 - 0.66 55.62 Jan. cash sale 3.42 6.7 0.73 45.03 3 cash sales 3.46 11.0 0.78 47.94 5 cash sales 3.58 23.0 1.03 53.85 Forward contract 1 3.59 23.1 * 0.57 52.76 Forward contract 2 3.62 26.7 ** 0.55 47.17 Forward contract 3 3.70 34.2 ** 0.71 55.68 Futures contract 1 3.93 57.5 * 1.10 45.09 Futures contract 2 3.95 59.4 1.73 63.610 Futures contract 3 3.96 60.6 * 1.41 59.611 Put option 1 3.69 34.0 * 0.90 55.612 Put option 2 3.82 46.6 1.53 55.613 Put option 3 3.76 41.0 * 1.20 55.614 Call option 3.46 10.7 0.86 46.115 Basis contract 3.55 19.5 0.83 44.416 Forward & Futures 3.94 58.6 * 1.34 58.017 Forward & Put option 3.86 50.9 * 1.20 53.618 Dynamic 3.77 41.1 * 1.09 53.5

DifferenceFrom

Baselineb

(cents/bu)



there is no clear answer as to the strategy that tends to be most effective. The standard deviation was substantially lower for these years across all strategies, and in particular for pre-harvest strategies. In fact, there was only one strategy (#18) in which the standard deviation was greater than that of the baseline strategy. However, Strategy #18 is one of only a few strategies for which the probability of receiving a price that is less than the cost of production is lower than that of the baseline strategy. In general, these probabilities are much higher than those reported in the initial analysis. Table 5: Results of the simulation model for the eight years in which pre-harvest prices

are less than costs of production

a Figures are rounded off to the nearest cent. b Asterisks (*,**) indicate statistical significance at the 10% and 5% levels, respectively. 3.3 Strategy Rankings For comparative purposes, the results of the three analyses discussed above can be outlined in terms of the overall rankings of the strategies based on average prices. These rankings are presented in table 6. The strategy rankings are very similar between the initial analysis that combined all years and the analysis of years with higher pre-harvest prices, where the strategies using futures contracts had the highest average prices followed by the strategies that used put options and the dynamic strategy. The strategies that only involved cash sales are all at or near the bottom of the rankings for both analyses. The rankings for the analysis of years with lower pre-harvest prices are


($/bu) ($/bu) (%)

1 Baseline 3.05 - 0.53 75.02 Jan. cash sale 3.20 14.5 0.46 67.43 3 cash sales 3.09 4.1 0.50 75.84 5 cash sales 2.99 -6.0 0.52 85.65 Forward contract 1 3.03 -2.2 0.37 85.26 Forward contract 2 3.12 6.2 0.30 92.67 Forward contract 3 3.04 -1.3 0.26 100.08 Futures contract 1 3.08 2.9 0.40 72.89 Futures contract 2 3.04 -0.9 0.41 74.010 Futures contract 3 3.07 1.4 0.40 72.411 Put option 1 3.05 -0.3 0.41 75.012 Put option 2 3.11 5.9 0.48 75.813 Put option 3 3.08 3.1 0.42 76.614 Call option 3.04 -1.1 0.52 81.315 Basis contract 3.18 12.9 * 0.40 83.816 Forward & Futures 3.03 -2.6 0.29 83.417 Forward & Put option 3.06 0.8 0.33 90.518 Dynamic 3.17 11.4 0.60 64.5

FromBaselineb

(cents/bu)

Difference



substantially different from the first two analyses. The pre-harvest strategies that used futures contracts and forward contracts dropped considerably in the rankings, with the exception of Strategy #6. The cash sale strategies were no longer grouped at the bottom of the rankings, with average prices for Strategy #2 ranking first while the baseline strategy ranked eleventh. Table 6: Strategy rankings based on average prices across all three analyses

3.4 Relative Performance of each Strategy While a number of strategies have relatively low average prices across the study period, such as the cash sale strategies, there may still be years in which these strategies perform relatively well. This was found to be the case in a study by Tomek and Peterson (2005), where strategies that generated lower average prices still performed well in 30-50% of the years modeled. An examination of the relative performance of each of the eighteen strategies in this study involved determining the number of years (out of 17) in which each strategy generated prices in the top third of all strategies. These results are given in table 7, along with the average price and the overall ranking for each strategy. Strategies #8 and #10, the futures contract strategies that have the highest average prices, also have the greatest number of years with prices were in the top third of prices for all strategies. These are the only two strategies which had prices in the top third in over 50% of the years of this study period. The baseline strategy, which ranked last overall in terms of average prices, had just three years in which the price was in the top

Rank All Years Above COP Below COP

1 10 10 22 8 9 153 9 16 184 16 8 65 12 17 126 17 12 37 18 18 138 13 13 89 11 7 1010 7 11 1711 6 6 112 15 5 1113 5 4 914 2 15 1415 4 3 716 3 14 517 14 2 1618 1 1 4

Strategy Rankings



third. Only Strategy #5, a forward contract strategy, had fewer years with prices in the top third. One interesting result was that the three other cash sale strategies (#2-4), which ranked very close to the bottom for average prices, had similar numbers of years with prices in the top third as compared to the strategies with put options (#11-13, #17), which generated average prices that ranked near the top. This may be due to the moderating effect of option strategies, which suppress the highs and the lows. But this also indicates that these cash sale strategies that have relatively low average prices are still effective strategies to use in certain years. Table 7: Relative overall performance for each strategy

3.5 Summary of Results Overall, the results of this study indicate that pre-harvest marketing strategies can generate higher prices for corn producers in Ontario as compared to selling the crop at harvest. These results are consistent with the findings of Wisner, Blue, and Baldwin (1998), Curtis et al. (1987), and Frank et al. (1989), but the finding that many of these differences are statistically significant is contrary to the findings of Martin and Hope (1984), Zulauf and Irwin (1998), and Peterson and Tomek (2005, 2007). The disparity in the results of this study between the years with higher pre-harvest prices and the years with lower pre-harvest prices is consistent to some degree with the results of Tomek and Peterson (2005), who found that the relative performance of alternative strategies varied from year to year. The finding that alternative cash sale strategies performed

Average Number of

($/bu) Rank Top Third

1 3.21 18 32 3.32 14 63 3.29 16 54 3.31 15 55 3.32 13 26 3.38 11 57 3.39 10 48 3.53 2 109 3.52 3 610 3.54 1 1011 3.39 9 512 3.49 5 513 3.44 8 414 3.26 17 515 3.38 12 716 3.51 4 717 3.49 6 718 3.48 7 6

Strategy OverallPrice Years in



relatively well in about 30% of the years of the study period is also consistent with the results of Tomek and Peterson. The results with respect to the relative variance of specific strategies differed somewhat from those of Frank et al. and of Peterson and Tomek (2007), which found that pre-harvest strategies such as options reduced the variance of returns relative to the baseline strategy. However, if returns from 2008, which added considerably to the variance of pre-harvest strategies, are omitted, the standard deviation estimated by this model for the pre-harvest strategies is less than that of the baseline strategy. Some of the differences in the results of this study as compared to previous studies could be due in part to the finding of Tennyson (1996), who suggested that the relative performance of strategies could vary by location. Differences in the results across previous studies may also occur due to variation in the relative performance of strategies across different time periods.



4.0 Conclusions Based on the results that were discussed in the previous section, there are a number of recommendations that can be made with respect to marketing decisions and the development of marketing strategies. These recommendations are outlined in the following section. However, the results of this study may be affected to some degree by some limitations that are inherent with the design of the simulation model. These limitations are discussed in section 4.2, along with suggestions for future research. The final section provides a summary of the findings of this study. 4.1 Recommendations While making multiple cash sales throughout the year may reduce risk, this strategy does not result in average returns that are significantly greater than the baseline strategy. Part of the reason for this is that any increase in price that occurs after harvest may be offset to some degree by storage costs and inventory costs. However, the availability of on-farm storage may reduce these costs to some degree, which could improve the performance of these strategies. But it is important that producers know these costs and take them into consideration when evaluating the use of such strategies. While forward contracts are used substantially more by producers than futures contracts and options, the results indicate that the returns from strategies that include forward contracts tend to be considerably lower than those of strategies that involve hedging with futures and options. This suggests that it may be worthwhile for producers to consider becoming more familiar with how futures and options work, and to incorporate these tools into their marketing strategies. A better understanding of how these tools work may help to alleviate the concerns regarding the risks associated with them. Pre-harvest marketing strategies, particularly futures contracts and options, appear to be most effective relative to the baseline strategy in years when prices are higher. In years with higher pre-harvest prices, the trend of declining prices into the harvest period is more pronounced. As a result, there may be significant benefits to using pre-harvest strategies in these years, rather than waiting for prices to get even higher later in the year. In years when prices are lower, and are below costs of production, pre-harvest strategies do not offer any significant improvement over the baseline strategy. In these years, there tends to be less of a difference between the pre-harvest prices and the prices during the harvest period. The differences in the relative effectiveness of specific strategies between the high-price and low-price years implies that using only one specific strategy in every year may not be the best approach to marketing, even if the selected strategy has been found to perform better on average across all years compared to other strategies (in this case, Strategy #10). Rather, it may be prudent to use different strategies in response to the current set of market conditions. This is supported by the results of the dynamic strategy



across the different analyses. This strategy, for which the timing of marketing activities and the tools used were allowed to vary in response to current market conditions, performed relatively well compared to other strategies in both high-price and low-price years. There are many factors to consider in the selection of marketing tools to use from year to year, not all of which have been accounted for in the model. For example, if the basis level is relatively high in the pre-harvest period, it may be better to use forward contracts rather than futures contracts or options, as forward contracts lock in both the futures and basis components of the price. Conversely, if the basis level is relatively low, then the use of futures contracts or options may be better, as they would allow for benefiting from a future strengthening of the basis. 4.2 Limitations and Suggestions for Future Research There are some limitations to this study as well as some considerations that were not accounted for in this model. For example, this model does not take into account the risk of fluctuations in production of corn over time due to factors such as drought. However, accounting for yield fluctuations would not allow for an appropriate comparison of returns under different strategies, as some factors must be held constant in order to conduct an effective comparison. Additionally, since the model measures returns on a per-bushel basis rather than an aggregate or whole-farm basis, the effects of omitting production risk from the model would be minimal. For example, a strategy that involves selling half of the total crop on the cash market at two different times would return the same price per bushel whether total production was 50,000 bushels or 75,000 bushels. The only component of the model that incorporates yields is the cost of production, for which the cost per bushel is determined by dividing the cost per acre by the average corn yield in Ontario over the previous five years. This moving average yield accounts for yield fluctuations at a macro level, but not at a micro level (i.e., yield fluctuations for individual producers). While the cost of production does not affect prices generated by specific strategies in the model, it does affect the share of the crop that would be marketed at specific points in time. There are a number of limitations related to the specification of the marketing strategies. Certain assumptions had to be made in the specification of marketing strategies, such as the timing and amounts of sales. Changes to these specifications may affect the results of the model to some degree. For this reason, caution must be used in extending these results to make generalizations about specific types of marketing strategies. However, there did appear to be relatively consistent results among strategies that used the same marketing tool, particularly for strategies that incorporated pre-harvest tools, which allows for some confidence in using these results to draw conclusions regarding the relative effectiveness of specific tools. But additional research would be necessary to determine the degree to which changes in the parameters of each marketing strategy would affect the relative returns.



The strategies that included futures contracts and options could have used these tools in combination with hedges on the Canadian dollar to minimize the exchange rate risk. However, this would add another element of difficulty to the development of marketing strategies, and may not be a component that is generally included in marketing strategies for smaller producers, given the size of Canadian dollar futures contracts and their associated margin costs. The specification of the options strategies required the selection of specific strike prices for specific delivery months. Given the wide range of strike prices that could be selected, the results may have differed under alternative specifications of these strategies. To fully evaluate the relative effectiveness of options strategies, additional research is necessary to test for the sensitivity of returns from options strategies to different strike prices (i.e., both below and above current futures prices) and to different delivery months. The specification of the marketing strategies requires that the same tools are used during the same time periods in every year for each strategy (with the exception of Strategy #18). This may not be representative of how producers do their marketing, as they may change their approach and the tools utilized based on current market conditions. However, the objective of this report is not to model producer decision-making that occurs in response to relative market conditions, which would present a significant challenge as producers may respond in many different ways to the same set of market trends and conditions. Rather, the objective is to compare returns from different marketing strategies, which requires the assumption that the same strategy is used every year. While the set of marketing strategies is limited in this report to the eighteen strategies described above, the model can be adapted to represent any strategy with any number of tools used and range of dates over which marketing activities may occur, provided there are specific decision rules that can be followed. 4.3 Summary This report provides evidence indicating the timing of sales or contracts and the types of marketing tools that tend to result in higher average prices received for corn. This evidence may provide Ontario corn producers with information that will be beneficial for the development of effective marketing strategies that may increase their returns. The results of this study indicate that there may be advantages to using pre-harvest marketing strategies that include futures contracts, options, and, to a lesser degree, forward contracts, as these marketing strategies return higher average prices than strategies where the entire crop is sold through cash sales. Particularly in years where the pre-harvest prices are greater than the costs of production, these pre-harvest marketing strategies generate much greater average prices for corn. However, these strategies do not consistently outperform cash sales in all years, as cash sales are found to perform relatively well in about 30% of the years in the study period.



References Curtis, C. E., Jr., Pfeiffer, G. H., Lutgen, L., & Frank, S. D. (1987). A target MOTAD

approach to marketing strategy selection for soybeans. North Central Journal of Agricultural Economics, 9, 195-206.

Frank, S. D., Irwin, S. H., Pfeiffer, G. H., & Curtis, C. E. (1989). Further evidence on

soybean marketing strategies: The role of options. North Central Journal of Agricultural Economics, 11, 213-219.

Martin, L. J. & Hope, D. (1984). Risk and returns from alternative marketing strategies

for corn producers. Journal of Futures Markets, 4, 513-530. Peterson, H. H. & Tomek, W. G. (2005). How much of commodity price behavior can a

rational expectations storage model explain? Agricultural Economics, 33, 289-303. Peterson, H. H. & Tomek, W. G. (2007). Grain marketing strategies within and across

lifetimes. Journal of Agricultural and Resource Economics, 32, 181-200. Tennyson, L. (1996). Tailoring a grain marketing plan for highest revenue. South Dakota

Farm and Home Research, 47, 6-7. Tomek, W. G. & Peterson, H. H. (2005). Implications of commodity price behavior for

marketing strategies. American Journal of Agricultural Economics, 87, 1258-1264. Wisner, R. N., Blue, E. N., & Baldwin, E. D. (1998). Preharvest marketing strategies

increase net returns for corn and soybean growers. Review of Agricultural Economics, 20, 288-307.

Zulauf, C. R. & Irwin, S. H. (1998). Market efficiency and marketing to enhance income

of crop producers. Review of Agricultural Economics, 20, 308-331.

comparing returns for grain corn production under various ... · comparing returns for grain corn...

Documents