US milk production was up 3.2% in January 2026, herds were at their largest since the early 1990s, cream was cheap, milk was flowing freely across every region, and NFDM prices nearly doubled. From around $1.18/lb in early December 2025 to a fresh record of $2.2625/lb on May 4, 2026, a rally that pushed NFDM to its highest level in 12 years, higher than butter, higher than anyone in the market had modeled.
NFDM prices according to CME call in USD/lb
The counterintuitive part: there was no shortage of milk. There was a shortage of willingness to turn it into powder. This is the story of how plenty of milk produced almost no NFDM, and what that did to prices.
Why processors stopped making powder
The NFDM rally didn’t start with a demand shock or a supply disruption in the traditional sense. It started with processors deciding, one by one, that NFDM wasn’t worth making.
Through 2025, every alternative paid better. WPC80 hit $10.50/lb, making whey protein the most lucrative use of the milk stream. Cheese margins, supported by whey protein byproduct revenues, kept cheese plants running full. MPC production had surged 52.61% in 2024 as the higher-value protein alternative. NFDM, sitting between $1.11 and $1.40/lb while EU SMP undercut US pricing and the strong dollar made exports harder, was the lowest-margin option on the board.
So processors rotated away from it. Not all at once, but steadily. NFDM production had been declining since 2022, and the shift accelerated through 2025 as higher-value products pulled milk in every other direction. By Q4 2025, NFDM output had quietly dropped well below expectations, with November numbers coming in remarkably low. Nobody panicked because the assumption was that production would catch up when it needed to.
It didn’t.
The short squeeze that lit the fuse
When Q4 production data confirmed just how little NFDM had been made, the market broke.
Several contracts couldn’t be honored. Traders and buyers who had been counting on those volumes went back to the market looking for product that wasn’t there. Two Global Dairy Trade auctions in January, on the 6th and the 20th, concentrated Chinese, Middle Eastern, and Southeast Asian demand into events where not much was on offer, pushing NZ prices to the $2,600/mt range and pulling European and US prices along.
Those caught short scrambled to cover positions, triggering aggressive buying across both NFDM and SMP. In a market with thin liquidity and few sellers, every bid cleaned the board. NFDM gained 30% in January alone. By early February it had posted another 30% on top of that.
The fundamental question was how much of this was real tightness versus sentiment, because milk was running plentiful and skimmed milk was cheap, yet processors kept choosing to send that milk elsewhere. NFDM+SMP combined production was down 1.3% year-over-year in January 2026, and MPC production dropped 11.8%, meaning neither product was picking up the slack. The rotation away from commodity powders toward higher-value proteins, with WPC80 production jumping 12.2%, was accelerating even as prices surged.
Three tight markets at once
In most years, at least one of the three major producing regions has surplus powder. If the US is tight, Europe or New Zealand fills the gap. If NZ is in its seasonal wind-down, the US and EU pick up the slack. That’s what keeps NFDM as one of the most globally commoditized dairy products.
In early 2026, all three were simultaneously tighter than expected. NZ was in its seasonal wind-down with limited volumes available. European SMP stocks sat lower than anticipated after strong 2025 exports. And US production was falling for the structural reasons already described. Three tight markets reinforced each other instead of offering alternatives. Oceania prices climbed on strong Chinese and broader Asian demand, European tightness removed the usual safety valve, and the US kept surging.
By mid-March, NFDM hit $1.765/lb, a 59% increase from Q4 2025. By late March it reached $1.9375/lb, the highest since 2014 and, for the first time, more expensive than butter. In the first week of April it touched $1.9725/lb before finally breaking through $2/lb by mid-April, short-squeezed above the line on the CME Call as supply ran very short and buyers were still caught on the wrong side of volumes they needed.
The decoupling that broke every historical pattern
The most unusual feature of this rally wasn’t the price level but the spread between the US and the rest of the world. In 227 weeks of data, only three times had any region held a meaningful premium to the group average for more than a couple of weeks. The longest was an 8-week US premium in late 2024, peaking at $0.114/lb above the three-region average, and within a month of that peak the spread was back to zero.
By early April 2026, the US NFDM premium to the three-region average hit $0.31/lb, nearly three times larger than any prior decoupling, and it had been widening for six consecutive weeks. Through February and into early March, EU and NZ prices appeared to rally in sympathy with the US, but that stopped over the final weeks of March as both flattened while the US kept climbing. The sympathy rally was exactly that: sympathy, not fundamentals.
The reason this decoupling was different: domestic constraints were absorbing the supply that would otherwise compete globally. Processors were choosing fresh products, cheese, and whey proteins over powder. The US wasn’t just tight relative to global markets. It was tight because the production mix had structurally shifted away from NFDM, and that shift was unique to the US market.
Geopolitics and energy poured fuel on the fire
On March 1, a joint US and Israeli military operation against Iran sent commodity markets into turmoil. Oil surged with crude briefly touching $118/barrel before pulling back and then climbing back above $110/barrel, while European Forties Blend reached $140/barrel. Roughly 20% of global oil supply originates from the affected region.
For dairy, the flight to safety meant buyers secured supply first and asked questions later. NFDM, already the tightest dairy market in the US, absorbed the additional uncertainty without flinching. The US Weekly noted that putting SMP and crude oil in the same price comparison chart “only amplifies the fact that commodity markets can react to similar things, no matter the type of commodity.”
The Fed revised its 2026 inflation forecast to 2.7% and the ECB projected 2.6%, with energy costs trickling through every supply chain. Higher oil prices feed directly into fertilizer, logistics, and processing costs, and the general flight to safety in commodities added a risk premium on top of fundamentals that were already stretched.
Currency played a role too. The US dollar had weakened significantly in late 2025 and early 2026 with EUR/USD touching 1.20 in January before falling back. A weaker dollar normally should have pressured NFDM prices lower by improving the export picture, but instead it contributed to higher domestic prices while European and NZ exporters benefited from favorable exchange rates.
The peak: $2.26/lb and the global spillover
By the close of April 17, the CME spot call printed $2.1825/lb, a 79.19% increase year-over-year. By April 23, it had run to $2.26/lb, a fresh record. By May 4, it was at $2.2625/lb, with the YoY gain at 82.34%. The forward curve still didn’t reach $2/lb for May contracts, meaning the futures market was pricing the squeeze as temporary even as spot kept making it look permanent.
What changed in those final two weeks of April was the geography of the rally. Through February and March, EU and NZ prices had moved up in sympathy with the US, but they flattened in late March as the US kept climbing. By late April, the gap between US NFDM and global SMP benchmarks had widened to $600–$1,000/mt on futures. That spread became wide enough to trade, and traders started going short US powder against long EU and NZ SMP.
The arbitrage flow began pulling European and Oceanian prices higher even though the supply fundamentals didn’t justify it. New Zealand had just posted a 9.4% increase in milk solids for March. Europe was sitting on a milk surplus running 7% above the prior year. The GDT result that week moved higher anyway. Fundamentals didn’t matter in the short run; the arbitrage did. The decoupling that started in late February kept widening past the $0.31/lb mark, and for the first time in 227 weeks of data, the US powder market looked like it was setting global prices instead of taking them.
The first signs of a turn
By the first week of May, several things had shifted at once.
First, the pause itself. NFDM moved sideways for most of the week and ended slightly lower. Prices that had been making new highs every week since January simply stopped. Nothing dramatic — just the absence of new highs.
Second, cold storage. March stocks for both butter and cheese came in above expectations, and the month-on-month build in butter was notably sharper than seasonal norms would suggest. That is not directly an NFDM signal, but it is a signal that the broader US dairy fundamentals are looser than the powder market has been pricing. When buyers see butter and cheese stocks building faster than seasonal patterns, the assumption that everything is structurally tight starts to soften.
Third, California. The Western region had already turned positive on YoY milk production for the first time in years, with March numbers up 1.08%. The state is now into peak flush, and processors who have been receiving record milk intake for several weeks are sending more of it to the drier. NFDM production is going to come in higher than it has all year. Whether that is enough to break the squeeze depends on how much of the rally was real tightness versus contract panic, but for the first time, the supply side has a credible counter-story.
Fourth, sentiment. The May 1 US Weekly captured the shift directly: “more bears popping up at $5000/mt NFDM than at $2250/mt a few months ago.” That observation is the cleanest signal of the cycle’s turn. When the consensus moves from “how high can it go” to “how does this end,” the buying pressure that drove the squeeze starts to fade before the spot price registers anything.
Fifth, the macro. Oil briefly peaked above $120/barrel in late April as a direct result of renewed Middle East unrest, separate from the March 1 event but the same theme. That sounds like the same story that fueled the rally. But this time the move came alongside disappointing Q1 economic growth across multiple regions and historically low consumer confidence. Higher input costs without buyer purchasing power create margin compression, not price support, exactly the dynamic that ended the 2022 commodity cycle.
None of these signals is decisive on its own. Together, they are how the top of a rally usually looks in dairy: not a single dramatic event, but the quiet erosion of the conditions that created the squeeze. Vesper’s forecast tells the same story:
Vesper’s NFDM US (CME Call) forecast through May 2027.
What this means for procurement teams
The NFDM rally of early 2026 challenges standard commodity thinking in several ways, and the forecast’s shape adds a set of forward-looking implications on top of that.
Production data alone doesn’t predict price. Milk production was at record highs and total dairy output was growing, but the product mix had shifted so far away from NFDM that aggregate supply numbers were irrelevant to the powder market. Tracking total milk production without tracking where that milk goes creates blind spots.
Byproduct economics drive allocation decisions. When whey protein prices make cheese production profitable regardless of cheese prices, processors choose cheese. When MPC margins beat NFDM margins, processors choose MPC. The lowest-margin commodity gets whatever milk is left over, and in a market where alternatives pay better, “left over” can be very little.
Short squeezes don’t need shortages. The absolute supply of NFDM wasn’t catastrophically low, but it was lower than buyers had contracted for. In a thin market, the gap between “slightly short” and “squeeze” is small, and several unfulfilled contracts were enough to trigger aggressive buying that moved prices 30% in a month.
Global decoupling is rare but real. NFDM is the most commoditized dairy product globally, and decoupling between regions almost never lasts. The early 2026 US premium was three times larger than any prior episode, and it has held for weeks. For procurement teams who priced NFDM based on global SMP benchmarks, the disconnect was painful, and treating the spread as guaranteed to revert in any specific timeframe is a planning risk, not a planning input.
The buying window is longer than it looks. If the central estimate is $2.40 in summer and $1.85 next February, there is no urgency to lock the entire forward book at the peak. Phased buying, splitting forward coverage between near-term and 6-to-12-month horizons, captures most of the unwind without taking the timing risk of trying to call the exact top.
The floor matters more than the peak for budgeting. Building a 2027 ingredient budget around the 2025 trough of $1.11/lb is the wrong baseline. The forecast points to a structurally higher resting price. Anyone whose 2027 plan assumes a return to 2025 levels is likely under-budgeting, and the gap is large enough to be a meaningful surprise.
Sentiment turns matter as much as data. The shift from “how high can it go” to “how does this end” rarely shows up in production numbers first. It shows up in trader chatter, in the futures curve flattening, in cold storage data running heavier than seasonal patterns suggest. By the time the spot price actually rolls over, the move is usually half done.
What’s next for US NFDM?
The US NFDM market moves fast. Every week, new production data, GDT results, and global SMP pricing reshape the picture, and the forecast updates with each new data point.
Our US Weekly covers all of it: NFDM, cheese, butter, whey, milk production, and the global forces pulling US dairy prices in every direction. Written by Vesper’s dairy team, delivered to your inbox every Friday. Read them on the Vesper platform app.vespertool.com.
