Wednesday, September 4, 2013

The Professional Agronomist | What's Your Number?

What’s Your Number?
               This time of year I get it all the time.  At the gas station.  Grocery store.   Hardware store.  At the gym.  The local tavern.  It literally happens every day.  “Hi.  What’s your number?” “Hey big guy, think I could get your digits?”  “I know you got a number in mind, wanna write it on my napkin…” “Think
I could get your number for Friday?”  It’s suffocating.  There’s no escape.
               Now as flattering as it would be to get all the attention, I’m a happy, soon-to-be married man and I’m not talking about my phone number here.  Let’s not get weird.  I’m talking about my yield number.  It’s seems like everyone has one around this time of year.   Some heard it from someone that seemed to have a grasp on the current status of the crop.  Some just repeat what they read from reports from different agencies.  Some are just taking their own stab.  Some have a number, but they’re not really sure how they got it or where it came from, they just have it.  More than likely, if your number is low, you’re more apt to share.  Misery loves company right?  On the other hand, if your number is high, you’re more apt to keep it to yourself.  Or you’ll likely share with someone that might have a number similar to yours.
               How many of you stick with your original number?  Does your number change depending on who you’re with or around?  Does your number change the closer we get to harvest?  Is your number influenced by the grain markets?  I’m sure in all of those instances it does. 
               What’s interesting to me is that most don’t come about their number by themselves.  A few of you probably do yield checks just to get an idea, but do you do enough roll that up into field averages? Farm averages?  I can understand some arguments. “  But Jake, counting kernels is a waste of time.  It’s never accurate.”  “There’s too many unknown factors, ear count, kernel weight, etc that are just hard to grasp.” “I can’t count kernels and drink beer at the same time.  At least not accurately.”  I have heard many different opinions on counting kernels and I believe anyone can get it close.  Why do I believe that? Because it gets done every year.  I have done it myself.  We’ll touch on that later.  First my perspective on some folks that have the courage to get out, do yield estimates, and then publish those numbers for everyone to see and criticize.  I’m talking more specifically about the USDA and the Pro Farmer Crop tour. 
About a week ago, the farm I work for hosted a group of traders, brokers, investors, etc.  I’m not completely sure those are the actual titles of the visitors so you’ll have to bear with me.  I’m mainly a plant and dirt guy.  I think they were fairly important people though.  But anyway, they were on a “tour” and obviously, given their professions, they had a significant interest in this year’s crop.  I used the term “tour” lightly because they were really just a group of folks from the East Coast riding around the Corn Belt, talking to farmers and farm groups, staying at nice hotels, eating nice meals, blowing up the expense accounts and drinking a lot.  The day they visited us, they stopped and made a total of three yield estimates.  Big day.
They visited us around lunch time and invited us down to Des Moines to continue the conversation with drinks and dinner at a nice hotel.  Shocker.  They set up to have Steven Johnson, Farm & Ag Business Mgt. Specialist, from ISU extension talk about the USDA’s Objective Yield Survey, how they conduct it and discuss the accuracy of their numbers.  If you’ve ever been at a meeting with Dr. Johnson before, you know that he is a very good presenter.  Very pro-ISU, very pro-Big 12.  He talked about how the USDA conducts their objective yield surveys, a very mundane process that the USDA has been doing since at least 1958. 55 years.  With the permission of the farmer, they go out to marked spots in fields, count ears and harvest them for weight.  Then they go back after harvest and check harvest loss.  It’s not a very exciting process but they been doing it the same way for over 5 decades.
               Don’t agree with the USDA numbers? They don’t really care.  How do I know?  In every August crop production report, they give us their yield estimates.  To supplement those numbers, they give you a table called the “Reliability of August 1 Crop Production Forecasts”.  Looks like this.               
This August 2013 table basically makes the statement.  “Our numbers are what they are.  This is our track record.  Love it or leave it.”  Let’s look at a couple numbers: the “Root mean square error” and the “90% confidence interval”.  Per the USDA report, the RSME basically means the most the report will be off by in 2 out 3 years is 6.0%. The “90% confidence interval” means that in 9 out of 10 years, the most they will off by is 10.4%.  Interesting fact: In 1978, the RSME was 6.4% and the 90% CI was 11.0%. Needless to say, the USDA has been pretty consistent that last 35 years.  
With this chart, they fully disclose that they are not perfect.  But they’re pretty darn good.  The USDA for over 5 decades has been guessing final corn and soybean yields in August and then comparing their August estimates to the final yields in January.   And by the time they make their final changes for the October report, they pretty much have it locked down.  As of right now, they are our best resource for knowing how good or bad the crop will be before it starts going in the bin.
               Another group that “takes a stab” at estimating crop yields in August is the Pro Farmer Crop Tour.  As I write, they are just wrapping up the final leg of the tour right here in Iowa and southern MN.  In the past, I have heard bits and pieces from different people about the tour.  But in today’s world of internet, Twitter and various media outlets, it’s really hard to miss. It appears to me to be a great group of people traveling the Corn Belt, getting out of their vehicles and out into the corn and soybean fields.  I love reading about their observations along the way as well.  All in all, it’s another great way for us to learn about the current crop and gauge for ourselves the final outcome.
               So here are some quick comparisons between the USDA August Objective Yield Surveys and the Pro Farmer Crop Tour.  Like I stated earlier, the USDA started doing objective yield surveys at least by 1957.  The Pro Farmer Crop tour started in 1993.  So the USDA has about 35+ years on the Pro Farmer Crop Tour.  The USDA publishes the reliability of their reports.  Pro Farmer does not.  If they do, I couldn’t find them anywhere.   I did look back at the past 2 years results and make some small comparisons. 
               All together, they are both pretty good.  I think all can agree that the USDA is typically better but they have been at it for a substantially longer period of time.  On a national level, you have many options as to who you want to believe.  The USDA and Pro Farmer Crop tour are who I deem as the most reliable based purely off of their methods and amount of time they have been doing it.  In the future, we may have even more options to work with.  And I predict that some organizations will start trying to make yield projections as early as June or July using satellite and aerial imagery as the base of their analysis.  In order for them to be successful, I feel they are going to need several years, maybe 5-10 years, of predicting yields and then correlating them to final numbers.  And even then so, they will have to prove their reliability to be a player in that market.       
I can think of several good reasons why a producer should be predicting his own yields well before harvest.  
1.      It physically requires you to be out in your fields, collecting samples for counting.  While you’re at it, you might as well be making observations as well.  Plant health, insects, standability, pollination and crop maturity are all things you can be looking at.  The less surprises at harvest, the better right?
2.      You don’t have to wait till harvest to find out if your assumptions on the year are correct.  Did you make a major change to your operation and you’re anxious to know the results? Get out and learn ahead of time.   You might be able to make adjustments to fall operations ahead of time based off of the knowledge you gained. 
3.      Adjustments to your marketing plan can happen prior to harvest and not while you’re in the combine seat.  You can have confidence in the sales you make, knowing the grain will be there.
4.       Ultimately, it makes you better than your competition.  If I were to estimate the percentage of grower’s who take the time to accurately predict yields, it would easily be under 10%.  Maybe under 5%.
5.      As long as you’re being somewhat conservative, the cost of being wrong is very low.  Even the time spent in the field should not be considered a waste.
So what’s the best method for predicting corn yields?  Before I walk you through how I predict yields and give you some guidelines to go by, I better expose my track record so you can differentiate whether or not I know what I’m talking about.  For the sake of confidentiality, I’ll just give you my percent errors.  And the regions are just different areas I predict yields in. 
               I’ve had some good years and some bad years.  One region I do pretty well in, one region I do pretty poor in.  I’ve stuck with the same method pretty much all the way through.   But I have made some adjustments.  However, I’ve waited to make them until I had more than 1 year of results.  You have to treat predicting corn yields like you’re sighting in a rifle.  1 shot will give you somewhat of an idea of where adjustments need to be made but 2-3 shots will give you a definitive group from which you can adjust accordingly.
               Here is the basic formula that I use and I’m sure most everyone uses.
(Kernels Long)*(Rows Around)*(Ears/Acres)
Kernels/Bushel
               General Tips & Guidelines
·        Count kernels on a lot of ears.  When I’m doing yield checks, I walk into the field 80-100 yards. I close my eyes and walk 5-10 paces and grab a stalk.  I take the next 10 ears right in a row.  Depending on how many fields you have, you might want to take more or less.   The more ears you count, the more accurate your estimate will be.   
·        Learn how to speed count or measure the length.   I count kernels long instead of measuring, mainly because that’s how I started doing it.  And I got fast at it by learning to count in multiples of 3.  3,6,9,12,15….. Don’t get hung up on where to start or stop on the ear, just find an average row.  I’m sure measuring works good, I just don’t want to change my methods.  When counting kernels around, split the ear.  Hold it up with your thumb on bottom and your index finger on top.  Count only the right or left side and double it.
·        Know your ear count. I take stand counts in every field in the spring.  I use those numbers from the stand counts as my Ears/Acre count.  If you don’t take stand counts in the spring, do ear counts as your taking ear samples.  2 stakes, small chain or cable between them.  You can use rope but know that it has the ability to stretch.  1/1000th of an acre long.  17’5” if you’re in 30” rows.   Don’t want to do either?  Use 8% as a loss factor from your planted population.   Planted 35K, lost 8%, 32.2K final stand/ear count.
·        Adjust kernels/bushel.  I have been using 95,000 kernels / bushel as a number that works for me.  I adjusted that to 105,000 in 2012 based off of moisture conditions and it turned out pretty good.  For 2013, I’ll revert back to 95,000, except for 1 region.
·        Document your guess.   Even go to the extent of showing or telling someone.  It makes you accountable to your methods.  And once everyone is done ridiculing you for how poor of estimate you made, you then have motivation to do better next year.  And the year after that.
·        Have fun.  Don’t like doing it? Think of ways to make it fun.  Make it a contest.  Hold an awards ceremony.  Organize a crop tour like the News from Coffee Shop crew. 
If you would like to continue the discussion, share your experiences, or complain about the “10 minutes of your life you won’t get back”, email me: jthoms11@gmail.com or follow me on Twitter: @cropguyjake


Jake Thomsen is an agronomist for farming operation in central Iowa.  Jake is marrying his fiancée Mollie in December and at that point will be relocating and making a career change to the Iowa City/Cedar Rapids area.  His hobbies include waterfowl, pheasant and mushroom hunting, fishing, golf, softball, and Cyclone athletics.         

No comments:

Post a Comment