In recent weeks, I have engaged in discussions with DTN Lead Analyst Todd Hultman over the relation of the final average farm price in corn to the final stocks-to-use ratio as furnished by the USDA.
In a prior post, I noted that the current forecasted stocks-to-use ratio of 22.4% would imply a far lower average farm price for the 2020/21 season than the projected $3.20 per bushel based on price vs. stock/use scatter plot studies we have done.
Todd has done work taking each WASDE report, starting with the first new-crop estimate in May and comparing 20-year correlations of each month to the final season average farm price concluding with the October WASDE, specifically, the October WASDE report which is one year after harvest and includes the result of the Sep. 30 grain stocks report.
As is to be expected, the R2 correlation improves as you get further into the marketing year.
This spawned further discussion about what factors caused the biggest changes in the stocks to use ratio and in turn the average farm price projection.
It was thought that supply issues were most influential for the first half of the year's WASDE reports from the first in May to the January report that includes the final production number.
Then demand considerations would have more impact on the stocks-to-use ratio and average farm price projections the second half of the year from February to the October report.
This study then shows the percent change in the WASDE U.S. corn total supply and total usage estimates from May to January and then February to October reports on the left hand axis vs. the respective change in the average farm price estimate in $/bu on the right hand axis for those two time periods also.
As an example in the 2012-13 season, USDA's average farm price projection increased by $2.80 per bushel from the May 2012 to January 2013 WASDE reports no doubt due to the severe 2012 drought resulting in total corn supplies off 24.2% and total usage projections down 18.2% between these two periods.
Price action and changing supply and usage swings were much more moderate the second half of the year, that is between the February 2013 and October 2013 WASDA reports where the 2012/13 average farm price projection went down by 31 cents and Feb-Oct total supplies and demand were up 0.5% and down 1.1% respectively.
This actually happens most years where price action, supply changes and demand revisions swing much more the first half of the year (May-Jan) than in the second half of the year (Feb-Oct), suggesting changing production estimates from the May to the final figure in January have a greater influence on the final stocks-to-use ratio and average farm price calculations than demand-side considerations.
This can be seen for the coefficient of variance figures, a measure of volatility that has 12.0 for changes in May-Jan supply, 21.0 for changes in May-Jan usage vs. 6.4 and 3.9 respectively for the Feb-Oct changes.
We see the correlation between the percent change in May-Jan supply changes vs. the change in the May-Jan average farm price is a high -75.3% with the percent change in May-Jan usage changes vs. the change in the May-Jan average farm price at -47.3%.
These correlation rates drop however to 37.0% for the percent change in Feb-Oct supply changes vs. the change in the Feb-Oct in the average farm price and 44.4% for the percent change in Feb-Oct usage changes vs. the change in the Feb-Oct in the average farm price.
© (c) Copyright 2020 DTN, LLC. All rights reserved.