Todd's Take

If I say, "two, four, six ... blank," there is a good chance you have already filled in the blank with a thought of "eight." You may have also noticed that the answer didn't take much brain power. It was intuitively natural and is one example of how our brains often jump to conclusions.

We can see this sort of tendency in all sorts of things in life, including the markets. In fact, noncommercial traders are notorious for anticipating short-term trends. It doesn't rain for a while, and they buy in anticipation of drought. Prices break a new low and they sell, expecting prices to go lower. While noncommercial traders typically follow and enhance trends, commercial firms tend to focus on economic value, which typically puts them on the opposite side of noncommercial trades.

In the Dec. 12 Todd's Take, I mentioned 1985 research by economists Werner F.M. De Bondt and Richard Thaler, "Does The Stock Market Overreact?" from the July, 1985 issue of the Journal of Finance (http://bit.ly/…). The paper examined long-term stock returns and found that stocks that underperformed the averages in one three-year period tended to outperform the averages the next three years. I have long suspected commodities would show a similar tendency for one-year returns and decided to fire up DTN's ProphetX software to test it out.

Ranking the annual percentage returns of 15 commodities from 1996 to 2016, I then recorded the following years' performances of the top winners and losers of each year. Similar to the findings of De Bondt and Thaler, the top performing commodities posted an average loss of 8.1% the following year, while the worst performing commodities posted an average gain of 13.8%.

Two of the most extreme examples included crude oil going from the bottom of the list in 2008 to a 78% gain in 2009. Cotton fell from the top of the list in 2003 to a 40% loss in 2004. More recently, crude oil fell from the top of the list in 2016 to fourth place in 2017 with a 12% gain, while feeder cattle rose from the bottom of the 2016 list and also posted a 12% gain.

If you are wondering if these kind of contrary results only apply to the top and bottom spots, the answer is no. Over the same 21 years, the second worst performing commodity posted an average return of 13.0% the following year. The third-worst performer had an average return of 8.6% and the fourth-worst increased 7.8%.

Addressing this contrary expression of non-random behavior, psychologist and winner of the 2002 Nobel Prize in economic sciences, Daniel Kahneman, wrote in his book, "Thinking, Fast and Slow," that "Regression effects are ubiquitous ..." (p. 178) and goes on to explain that we often make up stories to explain events that are largely the result of regression to the mean.

Regression to the mean, says Kahneman, explains all sorts of things, such as why tall people tend to have children closer to average height, why low golf scores on one day tend to be higher the next day, and why being on the cover of Sports Illustrated was once considered a jinx for athletes (Chapter 17).

For commodity prices, regression to the mean helps explain why the hot returns of one year tend to turn cold the next year and vice versa. It seems that our natural trend-following inclinations tend to take prices too far in one direction, setting prices up for a rebound. The odd thing about regression is that most market participants are blind to it. As Kahneman explains, "Whether undetected or wrongly explained, the phenomenon of regression is strange to the human mind," (p. 179).

Those of us who answered "eight" in the first paragraph are much like noncommercial traders who anticipate short-term trends. Eight feels like the right answer, and sometimes it is. However, over time, data shows that the more probable answer is closer to four, the average of two, four and six.

While I doubt commercial grain companies have regression to the mean in mind when they make their decisions to buy or sell grain, it is interesting that by focusing on economic value, they naturally act in a way that is contrary to most traders and comply with this unnatural influence.

Looking at today's grain markets, Chicago wheat was the second-worst performer in 2016 and posted a small gain of 5% in 2017. Except for feeder cattle, Chicago and Kansas City wheat contracts currently have the worst three-year returns of 17 commodities. On the same list, corn is in the lower half of three-year returns, having not seen a positive gain since 2012. True to form, commercial firms are net long in both corn and winter wheat, even though fundamental outlooks are bearish for both.

As I wrote last week, risk should not be confused with probabilities, meaning that even the worst-performing commodities are still at risk of trading lower. In terms of probability, however, spot corn and winter wheat prices are likely to trade higher in 2018 despite their current bearish fundamental outlooks.

What gives me the right to make such an outrageous claim? The history of commodity prices, which has shown consistently over long periods of time that regression to the mean outperforms our natural human tendency to project the past into the future. It won't make sense to many, but that's how the market works.

Todd Hultman can be reached at Todd.Hultman@dtn.com

Follow Todd Hultman on Twitter @ToddHultman1

(BAS/SJ)

Copyright 2018 DTN/The Progressive Farmer. All rights reserved.