Mid-August 2026 Hay Report

August 18, 2026

Date: August 18, 2026 Market: Southeast Colorado Premium large-square alfalfa, $/ton ex-stack (McClave / lower Arkansas Valley) Updates: the August 8, 2026 report, after the USDA releases of Aug 12–13. Full grading and data trail in research/12-grading-and-update-2026-08-18.md and research/grades-2026-08-13.md.

The Bottom Line

Ten days ago we wrote our predictions down before USDA spoke. USDA has now spoken, and the honest summary is: we badly overestimated how much of Colorado's alfalfa crop failed, and we underestimated how strong prices would be anyway. The first official 2026 production forecast came in at 2.24 million tons — down just 5% from last year, far above our 1.55–2.00M prediction range and nowhere near our 1.5M "crop disaster" estimate. Yet in the same week, the only large hay trade in Colorado printed at $275/ton delivered to a feedlot — up from a $225 benchmark in June and above our prediction range — and premium stable lots in southeast Colorado traded at $350–380/ton, up from $300–310 two weeks ago.

More hay than we thought, higher prices than we predicted. The resolution: this market is no longer a story about alfalfa scarcity — it's a story about everything else failing. Pasture is 64% poor-or-worse (five-year average: 23%). Nebraska reports native grass hay running 30–50% below normal tonnage. Oklahoma's drought exploded by 59 index points in one week. Cattle money is at all-time records. Every animal that can't graze becomes an alfalfa buyer, while the alfalfa that does exist sits mostly in northeast Colorado — 200 miles and $30/ton of freight away from the buyers who need it.

The forecast barely moves: the January 2027 peak trims from ~$265 to ~$260, and the extreme scenarios above $300 lose their fuel (Colorado won't run empty — there is hay at a price). The exit rules don't move at all — and the El Niño that ends this market got scarier: forecasters now give it a 69% chance of being the strongest event since 1950.

Report Card: The Pre-Registered Predictions, Graded

On August 8–9 we committed 13 predictions to the public record before the releases. Here is how the ones that graded this week came out — the economically important ones first, misses and all.

Score: 3 of 10 numeric prediction ranges contained the outcome. After four months of reporting "bands held," our first fully pre-registered test cycle was mediocre — which is exactly what pre-registration is for, and it confirms what our independent audit said about over-precision. Note the pattern: the narrative rows all hit; the numbers mostly missed, and every price number missed in the same direction — the market keeps coming in stronger than we predict. That's three consecutive reports where the surprises broke bullish. The remaining rows (USDA July state prices) grade August 31.

Price Prediction (updated August 18)

Honest uncertainty on any single month remains ±$25–30/ton, and the January "peak" is a scenario midpoint — the months around it are nearly as likely to be the top. Our prediction ranges just went 3-for-10; read this table accordingly.

Why the Price Is What It Is

1. The production story we have to retract — mostly

We estimated Colorado's 2026 alfalfa crop at ~1.5 million tons (−36%). USDA's farmer survey says 2.24 million (−5%). Even if USDA revises down as hard as it ever has in a drought year (−12%, in 2022), the final lands near 1.95 million — so we're revising our estimate to ~1.95 million tons (−17%), and the 1.5M call is on track to grade as a miss in January. Where we went wrong: we weighted the visible catastrophe (a dry Fort Lyon Canal, fallowed West Slope projects) as if it were the whole state, while wells kept pumping, the South Platte had a good water year (C-BT quota 80%), and first cuttings everywhere were made on spring soil moisture before the canals failed. The same USDA report shows what actually failed: Colorado winter wheat −67%, corn −24%, pasture ≥70% poor/very-poor across four states. The drought is real. It just mostly missed the state's irrigated alfalfa.

2. So why did the only big trade print at $275?

Because the price of alfalfa is being set by the failure of everything around alfalfa:

• Grass is gone. Nebraska: native grass hay 30–50% below normal, "upward pressure" in the official report. CO pasture 64% poor/very-poor. National other-hay crop −10%, with record lows in two states.
• The buyers have record money. Feeder cattle at all-time highs ($388–435/cwt for 5-weights, tops past $500); a rancher with dead pasture and $2,800 calves does not haggle over $20/ton of hay.
• The neighbors are re-drying. Oklahoma's drought index jumped 217 → 276 in one week (worst move in the country); Texas is heating up again. The demand radius that collapsed in June keeps re-expanding.
• Freight is up. Diesel hit $5.45, a three-month high — every substitute got more expensive to haul in.
• The floor moved up $10–20 in two weeks. Kansas feedlot-grade hay now lands in McClave at $222–252; the 800-ton NE Colorado trade nets $245 back to a SE CO stack. Nothing commercial should clear below ~$225 ex-stack.

3. And why the top scenarios got less likely

The same USDA number that embarrassed our estimate is genuinely bearish for the blowout case: there is hay in Colorado at a price — northeast Colorado has volume, and $275-delivered is the proof it moves. A winter buyer always has the truck-it-in option near $250–280 landed. The $300+ scenarios now require Kansas, Oklahoma, and Texas to fail simultaneously (that's how 2018 and 2022 happened) — Oklahoma is trying, but it's not there yet. Third-cutting rain helps too: August 1–17 delivered 2.97" at Lamar + Las Animas (our "cutting dies" trigger was <1"), even as the canal itself finished its last run — 439 acre-feet diverted in August versus 11,136 in June.

4. The clock still runs through the snowpack

The El Niño that should refill the Arkansas basin got stronger: >90% odds of a very strong event, and now a 69% chance it's the strongest on record back to 1950. Historic El Niños are the most favorable pattern for southern-Colorado snowpack — and the one wobble (spring persistence odds slipped to 82%) doesn't change the discipline: if December 1 snowpack is at or above normal, sell everything by Christmas; never hold past the February 1 reading. If anything, a record event argues for selling into the November–January window with both hands rather than waiting for February confirmation.

5. What to do with hay in the stack

Unchanged in structure, sharpened in level: ride the winter window, target $250–260 on the commercial channel (the NE-CO netback already proved $245), and put tested Premium lots into the stable/retail channel where $350–380 just printed. The $225 ask from April is now below what Utility-grade hay fetched in a small lot. One new caution: the scarcity premium is a winter-window phenomenon, not a permanent condition — USDA just told us the state has hay, and spring will tell us the mountains have snow.

What We're Watching (next report ~early September)

Data as of August 18, 2026. Sources this cycle: USDA NASS Crop Production (Aug 12, raw file archived and verified two independent ways), USDA AMS direct hay reports for CO/KS/NE/NM/OK/TX (Aug 7–18), CPC ENSO discussion and weekly update (Aug 13/17), ACIS station precipitation (Lamar airport + Las Animas), USACE John Martin daily report, Fort Lyon Canal Co. water report + CDSS diversion records, US Drought Monitor (Aug 11 map), EIA weekly diesel, BLS producer price index, CME dairy futures. Known soft spots: the AMS Colorado report was the thinnest of the year (1,306 tons statewide — two channels absent entirely); Oklahoma/Texas reports were stale (Jul 30 / Aug 7, both refresh Aug 21); freight uses an all-commodity flatbed rate from July 22.