Policy Uncertainty. The last seven years have been a case study in how foreign government policies have influenced the cotton market. Several major examples are India’s domestic minimum support price program, the 2018 U.S. farm bill, the U.S.-China Phase One Agreement, the U.S. Market Facilitation Program, and the U.S. CARES act funding. A more recent example is the Chinese tariffs on Australian cotton imports.
With established USDA benchmarks for acreage (12.19 million planted, although FSA certified acres show almost a million fewer), abandonment (24%), and yield (938 pounds per acre), the new crop production question will be influenced by harvest conditions across the Cotton Belt. Lately that includes a lot of rainy regions from Central Texas to Georgia.
The impact of recent weather, field conditions, and influences should be (albeit imperfectly) reflected in the weekly crop condition rating. The latter started off the season kind of in the middle, historically speaking, slid lower through June, stabilized in July, slipped over the first half of of August, and gyrated sideways in a gradual down-trend through the week ending October 18 (see light green line below). The condition rating reflects the cotton left standing, which includes a lot of maturing cotton in West Texas (reflecting mainly the surviving irrigated). Also, for the week ending October 18, U.S. boll opening and harvest rates were a little ahead and on-par, respectively, the pace of recent history. U.S. crop conditions for the same period were rated 40% “Good-Excellent”, and another 27% “Fair”. The weekly adjustments have mainly involved slippage from the “Fair” to the lower categories. As of mid-October the market’s focus appears to be on the possibility of quality and yield loss from cumulative rain events in the eastern Cotton Belt.
Ultimately the summer’s growing conditions did what they did. Now harvest conditions everywhere will have their influence, including too cool temperatures, unwanted rain, and mid-October hard freezes. USDA makes routine adjustments to their initial August estimate of U.S. production. The process they use to gather crop data include 1) farmer operator surveys (phone/mail/internet) and 2) objective yield surveys involving field sampling. For cotton, USDA did an August farmer operator survey across Texas and limited objective yield surveying in South Texas. Starting in September, they began conducting the first of four statewide objective yield surveys in Texas (as well as Arkansas, Georgia and Mississippi). In addition, starting in November USDA will be collecting ginnings data as an alternative, and perhaps more reliable, way of forecasting production.
The graph below shows how USDA’s U.S. cotton production estimate varies, higher or lower, from it’s initial August estimate. In years where the estimate gets smaller (possibly like in 2020), the average downward adjustment over eleven months is 6%.
And lastly, because cotton is a global crop, it is important to note new crop developments in major foreign producing countries. For example, the Indian crop is likely to be strong because of strong monsoon rains in some of its important cotton growing regions.
Demand Uncertainty. For U.S. cotton, the two main demand categories are domestic mill use and exports. Domestic U.S. consumption is estimated by USDA at 2.7 million bales. Exports are generally a more important source of demand as they represent over 85% of projected total use of 2020/21 U.S. cotton. The main indicators of export demand are weekly sales and shipments of U.S. cotton. The pattern of export demand has generally followed what we would expect from economics (lower quantity demanded with higher prices, and vice versa). The pace of actual export shipments of all cotton (i.e., upland and pima combined) was sub-par for the week ending October 15, but really too early to judge. The week ending October 15 saw a moderate level of net export sales for the 2020/21 marketing year.
It is an open question how the demand destruction from shuttered businesses and sheltering consumer will affect the cotton supply chain. Retail indicators of reduced demand include a dramatic 79% decline in U.S. clothing store retail sales in April (see chart below), and a correspondingly large decline in estimated consumer spending on apparel. Even May’s recovery of U.S. clothing retail sales is still below normal levels, and the June retail clothing sales show the same thing, i.e., higher month over month, but lower year over year. July retail clothing sales continue the trend of recovery. Consumer sentiment, another indicator, tells a similar story as retail sales, i.e., near term improvement, but not normal (certainly not bullish).