Sell your grain at the seasonal high in the spring. That's good advice, especially in normal marketing years with normal weather, normal demand, and normal influence from normal outside markets. Unfortunately, even if such a year ever presents itself, we still won't be able to predict the exact future date on which the markets will hit their annual high.
Predicting that date, or even predicting the seasonal pattern of prices, has only grown more challenging in the last several years with strange influences from unseasonal demand-driven, investor-driven, or drought-driven rallies.
In the past four years, the corn chart has experienced peaks in early August, in early July, in May, and again in July. Even a farmer who genuinely believes the market will always revert back to its time-tested seasonal patterns might feel a little spooky these days about trusting too strongly in normal spring highs.
If the weirdness of the past several years could be sifted out, I wondered, would the data still show something meaningful about grain markets' seasonal tendencies? Ideally, I wanted to pinpoint which day of the year has the strongest probability of establishing the corn market's annual high. But it turns out that if you want the corn market to reveal its seasonal movements, the data requires quite a lot of coaxing.
The results turned into gibberish if I used the continuous front-month corn chart, which switches which years' crop fundamentals it tracks throughout any calendar year or even throughout any September-to-September marketing year. They were also gibberish if I averaged or even indexed all the prices of the past 15 years for each day of the year. Extreme highs or extreme dates pulled the average far off from where the actual yearly highs seemed to cluster. The problem was they didn't cluster around any one, simple answer.
Instead, I simplified the question to just this: Within a given year, when should a farmer hedge his upcoming crop of corn? The fundamental outlook for the upcoming crop is always reliably reflected in the new-crop December futures contract. So I looked at each full trading year -- December to December -- of the past 15 December corn futures contracts to find the best timeframe for pre-harvest hedging.
Fifteen lines on a chart showing the performances of the past 15 December corn contracts look like tangled, multicolored spaghetti. Fortunately, the spaghetti could be unraveled by marking the highs of each year and identifying clusters of dates (local modes in a tri-modal statistical distribution) when those highs tended to hit. Amazingly, those clusters happened in groups of years with recognizable similarities.
I pulled out three categories of years.
* Short Crop Years (2001, 2002, 2006, 2010, 2011, 2012)
Corn production in these years was less than 105% of the five-year average. The trading pattern seems to reflect ever-increasing concern about supply throughout the summer and fall. In five of the six years, the new-crop contracts hit their highs on August 21, August 30, September 11, November 4, or November 30. Discarding the one year that didn't fit the cluster (2001), the average date for the new-crop corn market high during Short Crop Years was therefore October 1.
* Normal Abundance Years (2000, 2003, 2005, 2008, 2009, 2015)
Corn production in these years was between 105% and 114% of the five-year average (yes, the definition of these categories seems arbitrary, but this is how the highs clustered). The trading pattern reflects normal anxiety and risk premium in the spring and summer when weather is most uncertain for the growing crop. In five of the six years, the new-crop contracts hit their highs on May 3, June 2, June 26, July 16, or July 18. Discarding the one year that didn't fit the cluster (2003), the average date for the new-crop corn market high during Normal Abundance Years was therefore June 18.
* Supply Crush Years (2004, 2007, 2013, 2014)
Corn production in these years was above 114% of the five-year average and typically followed a shortage, therefore straining the industry's storage capacity. The trading pattern reflects increasing bearishness as the crop develops favorably through the summer and the huge harvest looms closer and closer. In three of the four years, the new-crop contracts hit their highs on February 22, April 8, or April 18. Discarding the one year that didn't fit the cluster (2013), the average date for the new-crop corn market high during Supply Crush Years was therefore March 26.
Of course, some annual highs were higher than others, and if I opened up the timeframe to consider a multi-year selling window, it all starts turning to gibberish again. Honestly, the best selling opportunity for 2009 corn would have been to tuck it away for three years and sell it in August of 2012. Similarly, a person could have hedged the December 2016 futures contract at $5.57 1/4 back on December 19, 2012. But it's not advisable for a farmer to depend on predicting three or four years at a time. DTN Grains Analyst Todd Hultman wrote earlier this week about the chances of hitting new highs within a year-long window: "Spot corn, soybean, and wheat prices all reached or exceeded their one-year highs 61% to 68% of the time. While that may not sound bad, each grain had pockets of three or four consecutive years that went without reaching a one-year high -- runs of bad luck that could put a producer out of business."
In nine of the 15 years I studied, the pre-harvest highs for the December corn contract were the best opportunity to sell the crop, even better than anything offered post-harvest for grain that was stored. In the other six years, there were some higher highs the next spring, but they were rarely high enough to reimburse a farmer for his risk and carrying costs. That is not to say that storing *hedged* grain to capture carry isn't profitable, only that the best-priced hedging opportunities have tended to be prior to harvest. Pre-harvest hedging works, and there are patterns when the best hedging opportunities have tended to appear.
Make what you will of all this. Remember this analysis was a study of historical data, not a prediction of future events. No one has independently verified my calculations, and I'm not recommending that anyone should buy or sell commodity futures or commodity options (or cash commodities) based on this analysis alone.
Furthermore, just knowing the historical highs for certain categories of years doesn't solve the eternal problem of figuring out what kind of weather year it will be. Nothing will ever solve that problem. There's no way to truly know, in March, how large the corn crop will be, or whether it will feel like a shortage, normal abundance, or a crush of supply. It's comforting to know, though, that no matter how strange individual years have seemed, the corn market still shows recognizable seasonal patterns.
As for soybeans, they merit the same treatment and yield similarly interesting results, but for the sake of column space, we will have to wait until the next Kub's Den column to unravel their seasonal highs.
Elaine Kub is the author of "Mastering the Grain Markets: How Profits Are Really Made" and can be reached at firstname.lastname@example.org or on Twitter @elainekub.
© Copyright 2016 DTN/The Progressive Farmer. All rights reserved.