Understanding Agricultural Data Means Standing in the Field
We talk a lot in our business about how technology and analytics drive critical information and accurate predictions. Wherever we can analyze information in a way that facilitates accurate predictions, our customers can run their businesses more effectively, efficiently, successfully.
But agricultural data is different. It takes more than just numbers to get the job done. One of the most critical aspects of crop analysis -- appropriately so -- is the crop tour: a boots-on-the-ground, count-the-corn crop tour combining the best of new technology with good, old-fashioned, firsthand visual evidence.
Agriculture is, by nature, a global and dispersed activity, making it necessary to integrate a vast array of information, from fundamental supply-and-demand data, prices, news, and international forums, to the analytics tools to evaluate importance and set strategy. As data points turn into big data, however, it becomes increasingly challenging for traders and investors to relate it to the fundamentals that they know determine supply and demand. Making matters more difficult, crops are often idiosyncratic -- frequently influenced by novel combinations of powerful forces. Unable to recognize the novelty, apparent insights drawn from big data alone can go terribly wrong.
Seeing the needles and the haystacks
The data picture in the agricultural markets may have never gone as terribly awry as in June 2012. On the heels of devastating droughts in the breadbaskets of South America, Central Asia, and Eastern Europe, US crops appeared to be racing to a banner harvest. Official forecasts on June 12 raised expectations for US production far above previous records. That same day, new crop contracts for Chicago corn closed at $5.22, near their one-year low.
Four days earlier, however, on June 8, Thomson Reuters analysts in the Lanworth team had noted that current and forecast drought conditions posed risk of substantial losses to an apparently healthy crop. Key signals in weather data were becoming uniquely similar to those seen only in a historic drought 25 years earlier, contradicting the more widely followed visual and satellite crop ratings and official forecasts.
Our analysts were walking US cornfields within weeks and saw that rain -- if it fell at all -- would be too late. From Illinois to Nebraska, in field after field, corn was already dying, long before almost any data had predicted or the market had anticipated. By the end of June, Lanworth’s World Crop Production reports were predicting vast reductions to expected supply. On July 3, Chicago new crop corn contracts broke through the previous contract high at $6.74. Official forecasts began making sharp reductions two weeks later, and Chicago corn prices rose until mid-September, peaking at $8.38.
Comprehensive data management and state-of-the-art statistical software allow us to identify and test fundamental relationships evidenced by big data. These relationships can vary, from identifying the level at which high rainfall will become harmful to corn yield in northeast China, to testing whether anomalies in cash-price data signal a shortage of soybeans in the US. Our analysts don’t only apply economic and weather-based models they have developed for a particular crop – individuals who “know” a crop then travel through the countryside making direct observations on plantings and crop yield to combine with entirely new data from satellite observations of vegetation health and post-harvest surveys of local operators as harvests come in.
This summer we again led the market by identifying key climate signals indicating that favorable conditions will finally lift global crop production to the levels officially forecast two years earlier. The analysts -- still walking fields this season across North and South America, Europe, and Asia -- have confirmed the view as fundamental data and government forecasts have fallen into line in support and crop prices have dropped to their lowest levels in five years. And at the same time, the rapid pace of progress continues. Detailed coverage of global weather forecasts and insights, weather, and satellite data feed capabilities and mobile information tools that will soon allow farmers to contribute and benefit from data on the go.
Together with expert guidance and confirmation, all of these technologies could potentially transform agricultural analysis, allowing farmers, fund managers, and physical traders alike to have access to the analytics that they need for decision-making with the same grounding as if they were all out, standing in the field.
Dr. Corey Cherr is Head of Agriculture and Weather Research and Forecasts at Thomson Reuters. Trained in biological and physical sciences, he has fifteen years' experience in the application of quantitative analysis, remote sensing and weather models to agricultural ... View Full Bio
Tweet This