PLC or ARC? First, Take an Antacid. Computer Models Can Disagree. – DTN

    The new farm bill’s program signup ends March 31, with yield updates and base allocations due Feb. 27.

    Picking the highest-paying farm program option for 2014-2018 crops may seem like a game of chance as random as flipping a coin, throwing darts or tossing dice.

    Even farmers who deploy the Farm Service Agency’s two supercomputers to analyze their options may discover the experts offer conflicting advice. Such conflicts can be compounded when unofficial calculators enter the mix.

    Southeastern Iowa farmer Ken Sobaski just finished entering data from land he farms in three counties on both Texas A&M‘s more sophisticated decision aid and a much simpler Iowa State University farm program payment calculator. Instead of consensus, they came in with conflicting recommendations.

    Using identical price forecasts from the Food and Agricultural Policy Research Institute (FAPRI) at the University of Missouri for 2014-2018 season crops, Texas A&M recommended he opt for Price Loss Coverage (PLC) for his corn with expected five-year payments of $10,175 on one farm unit while signing up soybeans for Agriculture Revenue Coverage-County (ARC-CO) at $4,479 over the same five-year period.

    Doctors Plastic Surgery recommended Sobaski enroll both his corn and beans in ARC-CO with projected five-year payments at $10,225. He used the FAPRI prices on both when making the entries and all the other data was the same also. Because he’s recovering from surgery, he’s had plenty of time to kill testing multiple outcomes and simulators.

    “I am wondering which one is the most accurate as this is quite a difference,” Sobaski said, still unsure if he should sign up for PLC or ARC for his corn crop.


    Land-grant university economists also found conflicts with the Texas A&M calculator and the University of Illinois’ farm program tool, both of which had been vetted by the FSA to weed out technical bugs. Some couldn’t understand how the Texas model recommended any PLC payments for corn if a grower expected the corn price to average no lower than $4 in any given year. By law, PLC doesn’t pay until corn prices fall to $3.70. In 2014, ARC begins to pay on average yields when season-average prices run below $5.29.

    James Richardson, co-director of Texas A&M’s Agriculture and Food Policy Center, has heard both questions. He can’t vouch for Iowa State’s program, or dozens of unofficial calculators available nationwide, but doubts the Texas or Illinois farm bill tools contain major flaws. Instead, he attributes the conflicting advice to methodological differences between the two computer models.

    After their own tests, Texas A&M modelers concluded they had incorporated more price risk in their simulations while Illinois included slightly more bearish yields. The architects of both computer models told DTN there shouldn’t be huge variations between the two if you input the same data and price assumptions. However, with lower prices embedded in the Texas A&M model, farmers could expect to see higher PLC payments than under the Illinois tool.

    “It’s a close call, and the real message from different results is that maybe you need to consider more than just expected payments in making program choices,” said FAPRI Director Pat Westhoff.

    “The tools both try to represent the programs as correctly as possible, and we both recognize the reality of uncertainty in agricultural markets by looking at ranges of prices around whatever averages the user specifies. Even very subtle differences in what those ranges look like and how uncertain prices and yields correlate with one another can change the results,” Westhoff added.


    But how can Texas A&M recommend PLC if a corn grower inputs $4 per bushel average marketing year corn prices, even though PLC won’t trigger until averages hit $3.70?

    That’s because the Texas A&M calculator performs the probability of 500 corn prices surrounding a $4 average marketing year price (see chart). Richardson said that range of prices means there’s actually a 40% chance corn prices will be low enough to hit the $3.70 PLC trigger even when a grower thinks he is using a calculation based on a flat $4 corn average.

    People underestimate price risk, the Texas economists stressed. “We don’t know the prices yet for 2014-15 for sure, and we’re halfway through the marketing year,” Richardson said. “We have a rough idea based on FAPRI price forecasts what corn prices will be further out, but it’s only within plus or minus $1.50 per bushel.”

    Westhoff said that doesn’t make one tool right and one wrong. “It’s just a reminder that all of these estimates can be sensitive to assumptions about things we can’t know for sure — not just average future prices and yields, but how prices and yields might vary from one year to the next around those averages. We know that can be frustrating for users, but it’s just a reminder that there’s no guaranteed ‘right’ choice.”

    Joe Outlaw, co-director of the Texas A&M program, strongly recommends farmers use at least one of the FSA-approved online simulators to incorporate a range of probabilities in their decisions. “Without getting too far down in the weeds, we will almost always have different results from models that did not include price and yield risk as we and Illinois have done.

    “We feel strongly that it is critical to incorporate risk when providing decision aids to farmers. Obviously the answers will depend on the future prices that are plugged into any of the models. Also, we received several datasets from FSA that to my knowledge have not been made publicly available, so I am not sure what data the other decision aids are using, but if they don’t have the official data — that may be another reason for differences.”

    FAPRI’s Westhoff likes an analogy he heard a speaker give an audience last week in Manhattan, Kansas: Suppose someone asks you to bet on the outcome of the roll of a dice. You have two options: In the first, you win if either a 1 or a 2 comes up; in the second, you win if a 3, 4, 5 or 6 comes up. If you choose the second option, but then happen to roll a 1 or a 2, did you make the wrong choice?

    “That’s similar to the situation here,” Westhoff said. “We won’t know until September 2019 which program choice will have yielded greater payments, but people have to make a choice today. Even if they make the most informed choice possible, there’s still a chance you’ll inadvertently make the choice that gives the smaller payment. The tools are intended to help shift the odds a bit in your favor, and to help you better understand the risks you face. They do not and cannot promise that they will tell you the choice that will ultimately have proved to maximize payments.”

    All this talk of probabilities and price ranges might make the decision process sound overly complicated, however. Given the challenges with forecasting prices, the University of Illinois’ Jonathan Coppess emphasizes a rough rule of thumb.

    “In general, ARC-CO looks to be the more favorable program for corn and soybean base due to the high prices used in the five-year Olympic average and the lower reference prices for those crops,” Coppess said. “At some point, however, low enough prices would flip that and make PLC more favorable over the five-year expected life. …In general, if the farmer thinks corn prices are going to be $3.30 or lower for the next five years, then PLC would be expected to provide more assistance. For soybeans, we estimate the break point at $7.80 and for wheat at $5.50.”

    Texas A&M tool at…

    University of Illinois tool at…

    The Latest

    Send press releases to

    View All Events

    [ecs-list-events limit="5" key="start date" order="asc"]
    Send press releases to

    View All Events