DOW · Q4 2025 Earnings

Q2 lower-for-longer admission → Q3 active capacity destruction and Alberta deferred → Q4 structural re-engineering and 4,500-job cut

The cost-out narrative has been retired and replaced with structural transformation. For three quarters Dow's self-help story was sized in the hundreds of millions: $300M → $400M cost target, $1B run-rate by 2026, three European asset shutdowns. This quarter management explicitly said "Transform to Outperform builds upon the self-help actions that we have implemented over the past few years. But importantly, it goes a lot further, representing a structural re-engineering of our operating model and cost base." The numbers attached are an order of magnitude larger: $2B near-term EBITDA uplift, $1.1-1.5B of one-time charges, $600-800M of severance covering ~4,500 roles. When a company moves from optimization vocabulary to "re-engineering" vocabulary and the severance line alone exceeds the prior cost program, the implicit message is that the prior actions weren't enough.

Alberta has gone from "January decision" to "two-year deferral at lower returns." Last quarter the Alberta path-to-zero decision was pushed to January at the earliest; this quarter management delivered the answer: "completing the project with a two-year delay is the most value-creating option. This moves phase one startup to late 2029... we now expect returns of at least 8 to 10 percent." The returns reset is the more important data point — the original project was underwritten at higher returns, and the explanation that "most of the change… is the capitalized interest" admits that the delay itself is destroying NPV. The reframing as "Path to Zero will put another cracker in the first quartile for us while we exit positions that are in the fourth quartile. I think that's something that we have to do in every cycle" repositions Alberta from growth project to portfolio housekeeping.

The cycle language has shifted from "lower-for-longer" to "fundamentally shifting markets." Two quarters ago the framing was a prolonged trough; this quarter it is "many markets are fundamentally shifting, geopolitical dynamics, rapid advances in AI and automation and economic volatility require new breakthrough approaches." This is no longer a cyclical posture. The accompanying admission that the Q4-to-Q1 data point reflects "bottom of the cycle integrated margins" and shouldn't be extrapolated is an explicit warning that the next two quarters will not show recovery either.

Inventory and pricing language picked up modestly — the one constructive note. Management said "we're getting some pricing power and moving things up... The drawdown in inventories at the end of the year in North America has really helped as we lean into first quarter." This is the first tactical positive in three quarters, but it sits inside a -8% YoY local-price print that hasn't budged, and management immediately hedged that they aren't extrapolating from it. The signal is "stabilization at the trough," not "recovery."

Capital allocation language has hardened defensively. "We anticipate that Dow's CapEx spending will remain at or below DNA until we see mid-cycle earnings" — CapEx now capped at depreciation until earnings recover. The Q3 framing left open the possibility of resumption; the Q4 framing makes the cap conditional on a cycle inflection management has just told investors not to expect.