When USDA updates its World Agricultural Supply and Demand Estimates today, chances are traders will focus on the bottom line of the U.S. reports for corn and soybeans. But after the initial data dust settles, the agency’s forecast for the world at large may be a more important metric to watch.
The key can be found in charts known as scattergrams for corn and soybeans.
Now, scattergram charts aren’t for everyone.
Many years ago I used one of these charts in a presentation to farmers in Michigan. Another speaker, who was actually an economist (I am not), confided he would never use this type of graphic, fearing farmers wouldn’t understand it.
But in today’s market, a scattergram chart helps explain what’s happening to corn and soybean prices – and what may happen in the year ahead.
So, what is a scattergram?
Most charts used in grain market analysis show a single variable over time. Prices, yields, exports and the like. Sometimes these are shown as bars, sometimes as lines.
A scattergram is different. It combines two variables, one on each axis, horizontal and vertical, into points that appear on the chart. A handy feature built into Excel automatically calculates the statistical correlation between the two variables and plots it as a trend line on the chart.
The classic scattergram used in the grain market compares supply and demand to prices. The ratio of U.S. marketing-year-ending stocks to total usage for the year is charted against the average cash price for the crop reported by USDA.
This analysis worked well once upon a time, but began to break down in recent years after the run to record prices in the wake of the commodity boom, the financial crisis, the 2012 drought, and the trade war. Predictions of average cash prices made from the stocks-to-usage ratio became less reliable. Sometimes other factors became important – the value of the U.S. dollar or positions of commodity funds, to name a couple.
Now, another trend has emerged that brings some much needed clarity: The global ratio of stocks to usage may be more important than what’s projected for the U.S.
Consider the scattergrams for corn. The ratio of U.S. ending stocks to total usage still matters. The relationship between the two, or r-squared as its’ known in statistics, from 2012 to 2018, is .75. This means stock-to-use explains 75% of the variance in corn prices. That’s enough to be statistically significant and not the result of mere chance.
But when world coarse grain stocks to usage are plugged into the formula, the resulting r-squared is .90. Coarse grains, which include sorghum, rye, barley, goats, millet and mixed grains as well as corn, are included because some countries don’t grow much corn. That includes Australia, traditionally one of the major feed exporters.
The importance of world stocks
At first, the importance of world stocks may be puzzling. Exports make up a declining percentage of U.S. corn demand – domestic usage to feed livestock and make ethanol are expected to be six times greater than sales to foreign buyers for the 2019 marketing year.
But the U.S. doesn’t operate in a vacuum. Its share of world corn production in 2019 is expected to be around 31%, down more than 20% over the past decade.
The numbers for soybeans are even more stark. Emerging corn competitors in South America and the Black Sea regions are only starting to ramp up production. Brazil, by contrast, is now the leading producer of soybeans globally.
The correlation between soybean stocks-to-usage and U.S. prices shows a weaker link than corn, perhaps due to fluctuating Chinese demand and differences in the U.S. and Brazilian production calendars. But world supply and demand far outstrips the U.S. as a price influence. The r-squared between U.S. stocks-to-usage and average U.S. cash prices from 2012 to 2018 is just .32. The world reading is .70, which is statistically significant.
What does this mean for prices?
World coarse grain stocks to usage looks like it will be down at the end of the 2019 marketing year for the third year in a row. And, when Chinese data is removed from the equation, supplies could be tighter in 2020 as well according to my initial forecast. That could help support U.S. prices as long as U.S. farmers don’t rush to increase corn acreage dramatically, suggesting an average U.S. cash price of $3.80.
The outlook for soybeans remains uncertain. While a smaller U.S. crop tightened stocks globally a little in 2019, the stocks-to-use ratio remains at a very high level historically. And it could be even higher in 2020 according to my forecast. But world stocks suggest an average cash price of $9.50 – assuming global demand recovers from the impact of trade wars and the coronavirus. That’s higher than the forecast using just U.S. data from the last few troubled years alone.
This points out the limitation of forecasting. Trends are always changing, along with the numbers used to create them. And trends are only visible in the rear-view mirror, when the market may have already moved on to a different dynamic.