Starting with the 2025 program year, AgriStability participants choose how their reference margin gets calculated. There are two options, the choice is locked in for four program years, and for a lot of farm files the two options produce very different benefits in a bad year. If you prepare AGRI forms for clients, this election now sits with you, whether it was ever formally handed to you or not.
I spent years in a farm practice maintaining the workbook we used for CAIS and then AgriStability, so I have some sympathy for how this lands. One more moving part, one more thing to document, and the guide is not exactly light reading. This is my attempt at the plain version.
AgriStability pays when a farm's production margin for the program year falls more than 30% below its reference margin. The reference margin comes from the five prior years: drop the best and the worst, average the middle three. That part hasn't changed. What changed is the question of which version of those prior-year margins goes into the average.
The Olympic Reference Margin (ORM) builds the average from the reference years' margins on the traditional basis. The Accrual-Adjusted Reference Margin (AARM) builds it from margins that carry the full accrual adjustments in each reference year, the same inventory, receivable, payable, purchased input and deferred income adjustments you already make for the program year itself.
For a farm whose cash income tracks its actual production every year, the two bases land in about the same place and the election barely matters. The farms where it matters are the ones where cash and accrual tell different stories year after year. In my experience that describes most of the grain belt.
Take a grain operation that habitually defers tickets into January. Every reference year, cash income understates what the farm actually grew, and the deferred income adjustment adds it back. Under AARM those adjustments flow into the reference margin. Under ORM they don't. Same farm, same five years, two different benchmarks, and the gap compounds because the pattern repeats every year.
Here is a worked example from our sample farm, a barley operation that defers $40,000 of tickets each year and hit a drought in 2025:
| AARM | ORM | |
|---|---|---|
| Reference margin (Olympic average) | $491,667 | $451,667 |
| 2025 production margin | $242,000 | $242,000 |
| Estimated benefit (2025 rules, 90%) | $91,950 | $66,750 |
That's a $25,200 difference on one file, from an election that costs nothing to get right and is locked for four program years if you get it wrong. Note the direction isn't automatic either. A farm that had been drawing down inventory through its reference years can come out ahead on ORM. The deferral pattern giveth and the inventory pattern taketh away, which is exactly why this needs to be looked at per file rather than decided as office policy.
You need each client's last five years on an accrual-adjusted basis anyway, because the program year demands it. The incremental work is running the reference years both ways and comparing. Doing that by hand for a two-hundred-file farm book is the kind of thing that quietly doesn't happen in March, which is how elections end up made by default.
My bias here is obvious since I built a tool for this, but the principle stands whether you use AgriBench, a spreadsheet, or the back of a feed store receipt: compute both bases on real numbers, before the deadline, for every file where the accrual pattern isn't flat. And write it down. If a client ever asks why their neighbour's cheque was bigger, the memo is the answer.
See the election modelled on a sample farm (free, no sign-up)
AgriBench supports accurate reporting and informed elections. Benefit figures above are estimates computed under published program parameters; actual benefits are determined solely by your program administrator. Program rules summarized here are current to the 2025 program year. Refer to the AgriStability Program Handbook and Guide RC4060 for authoritative guidance.