In November, USDA cut its forecast of U.S. all cotton (i.e., upland and pima combined) production by 900,000 bales. This tightened up the balance sheet and, in my judgement, moved us from a strongly bearish to modestly bearish outcome for old crop prices.
Speaking of prices, ICE futures did not appear to care about the tighter supply as it showed little market reaction to this report. One possible reason is an expectation that U.S. exports will not be as large as the 16.5 million bales currently forecasted by USDA.
The first thing to note is that the official U.S. export picture is pretty decent looking. As of November 14, 2019, there were 2,862,450 running bales worth of accumulated exports of U.S. all cotton, i.e., pima and upland sold and actually shipped. It also included another 7,539,191 running bales of pima and upland sold but not yet shipped (“outstanding sales”). The total of these is 10,401,641 running bales of “total commitments."
After converting total commitments into statistical bales, the level represents 65% of USDA’s 16.5 million bale target for 2019/20 U.S. exports. For this time of the year, this is the third highest percentage in the last eight years. So assuming the outstanding sales aren’t whittled down by cancellations, this is at least one indicator of decent demand, relatively speaking.
But there is some concern about cancellations. Some of the outstanding sales of U.S. cotton to China were made with fixed price contracts when prices were over 70 cents. Lower prices now create a disincentive for those buyers to take delivery and pay the higher prices. So maybe the total commitments are not as high as they seem.
Some of the outstanding sales were made to Chinese traders who were likely expecting a quicker trade resolution. What if the trade impasse lasts until the 2020 election?
Lastly, the required shipment numbers may be hard to achieve. To reach USDA’s target of 16.5 million bales of U.S. all cotton exports, an average of 371,000 statistical bales need to be shipped every week from Thanksgiving through July (i.e., about 36 weeks). But over the last 15 years, U.S. weekly exports for December to July average only 235,600 statistical bales per week.
We could easily enter a lull period of U.S. exports. A lot of cotton is going into the CCC loan, and the no-man’s land of current prices creates little incentives for farmers to sell or merchants to redeem cotton from the loan. As a result, cotton shipments may be bunched up during the last quarter of the marketing year (May to July). Once we get into Q4, the historical data suggest only between 15 and 20 weeks of such shipment levels (Figure 1).
In summary, we need above-average weekly shipments to export all the cotton sales on the books. If we don’t get it, then there will be more leftover cotton (i.e., ending stocks) on the balance sheet. Of course, a U.S.-China trade agreement would help ... but I am not expecting much on that front until at least the Nov. 2020 elections.
