Sterling Terrell

The question that I am still asked most frequently is, “So what do you think cotton prices are going to do?” Naturally, my answer usually lies somewhere between “I wouldn’t tell you if I knew” and “What drawer did I put that crystal ball in?”

In reality, while I am entirely unsure of exactly how much a significant event (or more usually, the net effect of thousands of events happening simultaneously) will affect cotton prices, I am much more certain of identifying events that may or may not be significant.

For example, if you buy a 100 percent cotton dress shirt this weekend, will that action affect the dollar price of cotton?  The answer is yes, because it increases the demand for cotton and, ceteris paribus (all other things being equal) will make cotton prices increase. However, nobody is analyzing or talking about your weekend purchase, because ceteris paribus is not reality. The purchase of one little shirt will not have enough of an effect on prices to matter.

But, what if China decided to outlaw cotton imports tomorrow?  Or, what if three or four countries got together and dropped their use of the Euro? You bet those events would have an effect on the dollar price of cotton!

My examples are obviously exaggerated to prove the point that, when it comes to a given event’s effect on supply and demand, some events simply do not matter. So, how much does the production of cotton from the U.S. affect the dollar price of cotton?  Let’s look at the data.

Using annual data from 1981-2011, I found that changes in the U.S. annual production of cotton can explain just over .036 percent of the variation in the average annual dollar price of cotton. Said differently – but using the same data – a one percent increase in the U.S. production of cotton should lead to approximately a .1 percent decrease in the dollar price of cotton, on average. Conversely, a one percent decrease in the U.S. production of cotton should lead to approximately a .1 percent increase in the dollar price of cotton.

So, in 2009/10, total U.S. cotton production was at 12.2 million bales of cotton. However, if U.S. production had fallen by 10 percent, from 12.2 million to 10.98 million bales, we could have reasonably estimated that prices would have risen one percent. So instead of a 2009 average crop year price of 71.6, we would have expected prices to rise to 72.32.

This brings me back to my point.

First, keep in mind that the U.S. produces somewhere between 15 and 18 million bales per year. Second, considering current world supply figures, the expected 2013/14 U.S. production of cotton (14.00 million bales) will account for about six percent of the total world supply (233.25 million bales). According to our model, this means that eliminating half the U.S. cotton crop would lead to, on average, approximately a five percent increase in prices.

Is that even a valid estimate? Yes and no.

First, yes. That is what the data says on average should happen. How is this possible?

It is possible because, as a percentage of overall world cotton supply, the U.S. cotton crop is smaller than most of us realize, and because U.S. (and world) cotton production may be less correlated with the dollar price of cotton than we thought. For example, during the 1986/87 season, U.S. cotton production was 9.73 million bales, with an average price of 55 cents. In 2005/06, U.S. growers produced 23.89 million bales, also at an average price of 55 cents.

And second, no. This model looks only at average changes in the production of U.S. cotton and dollar cotton prices (monthly data might give a better picture). It also assumes other variables are constant. And as I said before, neither of these facts reflect reality. Also, much of this data reflects points in time when China did not hold, by some estimates, 60 percent of the world’s cotton, as is the case today.

Even still, my advice would be to not look at the three counties you periodically drive through and think, “The crop looks bad this year, therefore supply will be down and prices will soon go higher.” You may end up being correct that prices are going to increase. But rest assured that if prices do increase, it will most likely not be the result of the reason you thought. Said reasoning once again confuses correlation with causation.

In a given year, there can be little to no cotton in a very large area without decreasing cotton supplies significantly affecting the dollar price of cotton. Cotton is bought and sold in an extremely large global market, and – as an American cotton grower – you play an important role as a supplier to that market.

And while some events have a marked effect on the price of cotton and are worth taking note of, minor changes in the total world cotton supply is simply not one of them.

First published at  Cotton Grower.