By Elaine Kub
DTN Contributing Analyst
Statistics can be a strange, fickle, and misleading guide. Back in April, I did a large study of individual counties' corn yields to find which ones tended to be most closely related to overall U.S. corn yields. It turned out Lyon County in northwest Iowa, with its corn yields from 1955 to 2014 showing a +0.963 correlation to U.S. corn yields, and its neighbor Minnehaha County in southeast South Dakota (+0.956 correlation) were the winners, and I wrote in a Kub's Den column:
"Their good years tend to coincide (1994, 2004) with good years for U.S. corn yields, and their yield trends also dip in matching years (1993, 2012). Lyon County itself isn't a big enough corn producer (28 million bushels in 2014, or 0.2 percent of the nation's production) for us to conclude that its conditions are in some way 'causing' the U.S. average corn yield to change from year to year, but it certainly seems like a representative poster child. By calculating its coefficient of determination (r squared) at 0.93, we can say that 93% of the variance in its corn yield can be accounted for by whatever simultaneously causes U.S. corn yield to vary."
I emphasized the point because, at the time -- in April -- that region of the northwestern Corn Belt was quite dry and I was bullishly concerned that 2015 overall corn yields might be affected by a lack of moisture.
Well, about that.
Multiple alert readers have recently contacted me and asked me to follow up on how Lyon County, Iowa, is actually faring in 2015, now that the crop has matured enough to give us a confident sense of what yields might be.
It turns out they have a beautiful crop right now. As Dick Groen at Cooperative Farmers Elevator in George, Iowa, put it: "We're going to have high yields if not record yields. We never caught any of the excessive rainfall, but always caught just enough rain. You couldn't irrigate a crop any better than what we got for rain. We've been very fortunate."
Compare that to the prospects for U.S. corn overall. In its latest supply and demand estimates, which were widely considered to be bearish, the USDA projected nationwide corn production at 13.7 billion bushels. That's almost 4 percent below last year's record-breaking crop, due to multiple regions' struggles with excessive rain at planting, and then acres lost to flooding throughout the summer. It would seem that the Lyon County numbers and the U.S. numbers are not going to be a nice match this year.
This is a great opportunity for me to acknowledge that something I implied about the future of the grain markets obviously did not come true over time. And I'm happy to do that because I'd like to enlist your help in a larger, farther-reaching project.
Grain market analysts, advisors, traders, consultants, Twitter experts, and your neighbor farmers standing around the local elevator's coffee pot -- collectively, let's call them the Grain Market Commentariat -- all have opinions about the production and demand and prices of various commodities.
These opinions can be tentative ("Grain damage from a potentially wet harvest might add some risk premium to the market.") or they can be bold. ("Live cattle won't dip below $135 for the next three years.") They can be well-researched ("The oil market will remain oversupplied into 2017, and prices could fall below $30 per barrel.") or they can be silly. ("I think corn is going to be back above $7 within a year because the sun spot cycles told me so.") They can be helpful ("Corn is at a profitable price; lay off risk now by hedging some future sales.") or they can be self-serving. ("You should buy butterfly spread options through my brokerage and roll them forward every month.")
But they are very rarely backtested. That is to say, no one seems to go through back issues of farm magazines, or the PowerPoint slides from last year's market outlook meetings, and give out scores for how good the analysis turned out to be. Mutual fund traders eventually publish their track record of gains and losses, but as a grain market pundit, I can go on the radio and say any ridiculous thing I want, and as long as I present it as an opinion and not a fact, I'm generally not held accountable for that opinion's ultimate accuracy.
So for the next several months, I am going to keep track of every grain market opinion I read or hear, and eventually, I will go back through and see which predictions came true and which pieces of advice were ultimately wise. What is the overall success rate of publicly shared grain marketing advice?
If you encounter any snippets of grain market commentary that you would like to have tracked and backtested, you can help by sharing them (anonymously, if you prefer) on this web form:
I'll even promise that if I do something so rash as to make an actual prediction or give actual advice during this weekend's Market to Market taping, I'll include myself in the study. I'll subject my own prognostications to the same scrutiny as the rest of the commentariat's.
I'll continue collecting responses through harvest and probably through the end of the year. So save that form link somewhere. The next time you hear a nugget like, "Soybeans are in danger of sliding below $8.00 down to $7.50," or, "Stop hunters are active today, so crude will probably overrun the $50 level," let's keep track of how well it holds up over time.
By the way, back in my April column about counties whose corn yields were closely related to the nationwide average, I mentioned a third-best correlated county: Pierce County in Missouri. That's in the southwest quadrant of the state, where moisture stress this spring and summer limited their row crop production capabilities for 2015. Maybe the implications I drew from statistics weren't so far off after all!
Elaine Kub is the author of "Mastering the Grain Markets: How Profits Are Really Made" and can be reached at email@example.com or on Twitter @elainekub.
© Copyright 2015 DTN/The Progressive Farmer. All rights reserved.