The USDA puts out hotly anticipated balance sheets every month but when they – and we – are making forecasts about how much corn will be left over at the start of next September, how accurate can the estimates be at this time of year?

The production side of the balance sheet is pretty close to being known now that harvest is underway. But the demand side is trickier to assess. And that is illustrated by USDA’s corn export forecast for the 17/18 crop year.

USDA is forecasting US corn exports to fall to 1.85 billion bushels from about 2.3 billion in the 16/17 crop year. The assumption is that production will be about 1 billion bushels below last year, therefore exports will be lower as lower supplies equal higher prices. July 2018 corn is trading about 25 cents above where July 2017 corn was at this time last year, despite a virtually identical ending stocks forecast of 2.3 billion bushes this year vs. this time last year.

But what is interesting is the global corn balance sheet forecast, and that shows where there is plenty of room for the US export forecast to change as time goes on. USDA is currently forecasting global ending stocks to fall from about 225 MMT in 2016/17 to 200 MMT in 2017/18, about a 900 million bushel difference (1 MMT is about 35 million bushels). But USDA is forecasting global demand to fall slightly, from 1058 MMT to 1057.

Given that global demand has risen to current levels from just 860 MMT only five years ago, it’s a little hard to believe that global demand would be flat in 2017/18. But USDA is expecting Brazilian and Argentine corn exports to total 62.5 MMT vs. 63.5 in 2016/17. There is a lot that can happen to Brazilian and Argentine production, and export totals, between now and harvest in early 2018. Corn-vs.-bean planting mixes are still being determined to some extent, while the most important portion of Brazilian production the “safhrina” second crop doesn’t even go into the ground for another 6 months. Brazil and Argentina export a large portion of their corn crop and don’t carry over huge stockpiles from one year to the next, only about 7-12 MMT combined in general.

This means that any number of things can happen between now and early 2018 – poor South American weather, a planting shortfall driven by strong soybean demand, or changes in global demand patterns – that can change US corn export demand. Given that US stockpiles are going to be pretty robust, it seems that US export demand forecasts are more likely to go up than down over the next several months.

An increase in export demand of 200 to 300 million bushels from current levels wouldn’t make that large a difference to the US balance sheet, only sending the carryout to about 2 billion bushels. But that could be enough to add about 20 cents to July 2018 futures vs. our current fair value forecast closer to $3.50 at expiration.