Tool Projects Crop Production
Lanworth combines frequent satellite imagery with intensive ground verification, knowledge of historic crop rotation and yield patterns, and computer-based simulations to project planted and harvested acreage and yield for corn, beans, and wheat in the U.S. and South America.
The Chicago-based firm provides clients — grain companies, ethanol companies, financial funds, large farmers — with biweekly reports, and it releases summaries to Bloomberg subscribers ahead of USDA’s monthly crop production report.
Now in its third year of reporting, Lanworth begins developing its planted acreage projections in the spring.
“We have a fundamentally different process from anyone else, including USDA,” says Nick Kouchoukos, a geographer and Lanworth’s vice president of resource intelligence.
Rather than relying on farmer surveys or expectations based on market signals, Lanworth uses space-based satellite imagery, mapping the planting field by field, refined by the “agronomy side” and the “geography side,” Kouchoukos says.
The best predictor of what farmers plant is what they planted last year, Kouchoukos told DTN. “Farmers practice rotation for a very important reason — to manage limited resources. When USDA is surveying farmer intentions, we are studying rotation patterns (last year’s crops and rotation frequencies), and complex geographical patterns we struggle to understand.”
Lanworth’s third step in projecting acreage is ground verification. Field teams cover preset routes from mid-May to mid-June through major crop areas in major states. “They check to see if what we said was corn is corn, and we gather rotation information for the database,” Kouchoukos said.
As specific events occur — flooding or hail, for example — Lanworth teams resurvey affected areas to provide updates on those areas to clients.
Once the crop is in the ground, Lanworth begins its yield assessments. Lanworth’s methods in determining yields also depart from USDA and other private firms, says Corey Cherr, Lanworth’s lead agronomist.
Rather than starting with trendline assumptions and trying to account for the impact of current conditions, Cherr said Lanworth bases its forecasts on the way the crops work and the environment they are in.
“A good example is the record cool temperatures in July this year. It’s different than assuming a trend line yield and seeing where another year’s yields varied. We look for the biological basis for the trendline. In corn and soybeans, that comes primarily from duration of grain fill and resource capture. When you account for flowering and potential maturity date, you’ve accounted for trendline and potential deviation,” Cherr told DTN.
Some other analysts are using 2004 as a reference year, since there were cool temperatures then, too. But flowering that year was earlier, so the corn had more “resource availability,” and yields were not reduced the way Lanworth expects is happening this year.
“This year, the biological meaning of the cool temperatures is very different from other years,” Cherr says. In places like Iowa and Illinois, aside from the delayed development, light levels since June are down an average of 10 percent, he said.
Cherr said other forecasters make use of USDA weekly crop progress reports, but “from our approach, the crop may look good, but if it can’t put that into the seeds, yields won’t be any higher. … We use a fundamental understanding of how biological processes function” to estimate yields.
Lanworth’s track record for 2007 and 2008 is pretty impressive. In 2008, for both corn and soybean production, for example, USDA estimates fluctuated from month to month, while Lanworth said its estimates were generally stable. By the end of the season, USDA’s totals for both crops were within one percent of Lanworth’s.
This year, Lanworth is projecting about the same split between corn and soybeans as USDA reported June 30, but Lanworth projected planted acreage for the two crops about a million acres less than USDA. Its production estimates in August were about 600 million bushels below USDA on corn and 200 million bushels below USDA on soybeans.
“We are not saying yields won’t be good,” Kouchoukos said. “We may be heading for the third highest yield on record. But satellites can’t see and condition ratings can’t say what the seeds are doing.” For that, Lanworth relies on a fundamental understanding of how biological processes function, Cherr said.
It’s important for risk managers using Lanworth’s data to realize the projections are for final totals, not the month-to-month changes USDA may make, says David Levine, Lanworth vice president of resource solutions.
But USDA’s monthly reports do provide hedging opportunities. “By understanding where the USDA will be in the future, relative to where they are presently, our clients find important opportunities to predict and monetize commodity price fluctuations when executing hedging, risk management, procurement and investment strategies,” Lanworth said in a promotional presentation Levine provided to DTN. That was the case this spring when Lanworth’s acreage projections proved to be much closer to USDA’s than the average pre-report estimates from other firms, Lanworth said.