Record growth, a raised guide and a $5 billion buyback — and a stock the market is treating as a peak-cycle power play.
HWM · Earnings Call · 2026-10-29
Howmet’s Record Quarter Meets a 30% Drawdown
The turbine-blade maker just posted its strongest growth quarter on record, raised guidance, and added $5 billion to its buyback. The stock is 30% below its August peak. That gap is the story.
A record quarter the tape has already left for dead
On 29 October Howmet reported Q3 2026 revenue of $2.7 billion, up 29% year over year and 25% organically; EBITDA of $879 million, up 43%; a 32.5% EBITDA margin; EPS of $1.46, up 54%; and quarterly free cash flow of $562 million. “Revenues at $2.7 billion were up 29% year-over-year and up 25% through organic growth. EBITDA was $879 million and was up 43%, with a margin rate of 32.5%.” — John Plant, Executive Chairman and Chief Executive Officer · 2026-10-29 The company beat the high end of its own guidance on revenue, EBITDA and EPS, lifted full-year revenue to $10.3 billion, EBITDA to $3.33 billion, EPS to $5.50 and free cash flow to $2 billion, and set an initial 2027 revenue guide of roughly $11.5 billion, up 12%. It repurchased $500 million of stock in the quarter, lifted the dividend 17%, and increased its buyback authorization by $5 billion. This is not a company in operational trouble.
The fundamentals confirm the operating inflection, even if the latest filing is a quarter behind the call. The June-quarter filing showed revenue of $2.5 billion, up 24% year over year, before Q3 accelerated to $2.7 billion.Operating margin reached 27.9% in the June quarter, up 2.5 percentage points year over year. The call added that Q3 EBITDA margin rose 310 basis points to 32.5%, with 43% incremental flow-through. The balance sheet remains investment grade in shape: net debt/EBITDA of 1.3x, an undrawn $1 billion revolver, and a $1 billion commercial paper program.
The new thing: IGT is no longer a side story
The most important change in this report is not aerospace. It is the industrial gas turbine business. Gas turbine revenue grew 45% in Q3, and the company now expects to double IGT sales from the 2025 base in three years, not the three-to-five years it previously discussed.
whereas our prior commentary had been around a doubling of IGT sales from 2025 as the base year within a three-to-five-year period, we now see this as being achieved in three years... maybe it will rise to 2.5 times inside five years, which would take those sales to about $2.5 billion by the end of the decade.
That is a quantified, company-specific upgrade. Management says major gas turbine customers have 2031 sold out and 2032 likely sold out within a quarter or two. It also walked through the data-center power bottleneck with unusual specificity: SpaceX’s behind-the-meter fleet, roughly 1 gigawatt of compute power, 33 Titan turbines and 10 ProEnergy units, and Howmet’s blades inside both Titan and LM6000 aeroderivative turbines. gigawatts of power is no longer an abstraction in this call; it is a demand signal.
This is where Howmet becomes a AI data center adjacency story. The company frames data centers as one leg of broad-based IGT demand alongside industrial, oil and gas, and power generation. But the market does not always make that distinction. The global tape shows the AI data center basket down 5.3% over 90 days, with 7 advancers and 49 decliners. Howmet’s own shares are down 25.4% over the same 90 days and sit 29.6% below the 18 August peak of $292.65. The operating numbers went up; the multiple came down.
The market is selling the AI-power adjacency, not the aerospace aftermarket
This is the central tension. Howmet’s end markets are not one trade. Commercial aerospace grew 32% in Q3, defense aerospace 10%, commercial transport 28%, gas turbines 45%, and other markets 48%. Spares revenue rose 30% to about $570 million, or roughly 21% of total revenue, driven by legacy and next-generation engine programs. Spares demand is the high-margin annuity that underpins the aerospace thesis. The Commercial Transportation recovery is real but cyclical, and the North American market recovery comes with aluminum cost pass-through that dilutes margin rate even as it lifts revenue.
The tape, however, is treating Howmet as a peak-cycle power capex name. At the June quarter the shares traded at 60 times free cash flow; the August peak made that look even richer. A 30% drawdown in a company still guiding to $2 billion of free cash flow is a valuation reset, not an earnings warning. The question is whether the reset is enough. On 2027 revenue of $11.5 billion and $5.50 of 2026 EPS, the stock is cheaper than it was, but it is not statistically cheap unless IGT and spares both hold.
CPP is the new competitive variable, and management knows it
The first analyst question was about GE’s acquisition of CPP and vertical integration in blades and vanes. John Plant’s answer was revealing. He conceded that CPP is “probably of the higher order in structural castings,” then spent several minutes quantifying why losing a narrowbody first blade would not matter: one year of IGT spares growth in 2031 would exceed the total sales value of the first blade on Howmet’s largest narrowbody engine. That is a confident defense, but it is also an admission that the blade franchise is the thing investors are worried about. acquisition of CPP and structural casting are now part of the Howmet debate in a way they were not a quarter ago.
The capex language reinforces the point. Howmet is already considering another plant and a larger capacity commitment for GE that takes two years to install. “for every year we see an elevated CapEx, then it gives us a little bit of confidence about that future growth and its solidity, going into the next decade.” — John Plant, Executive Chairman and Chief Executive Officer · 2026-10-29 That is the right long-term posture if demand is real. It is also how a high-return compounder becomes a heavier capital story. The market may be discounting both the competitive risk and the possibility that free cash flow conversion temporarily troughs while the capacity is built.
What the prior calls tell us
IGT has been the drumbeat for over a year. In August 2026, Plant said “Howmet has a market share in excess of 50% globally for turbine blades in the IGT market. And therefore, the growth of that market is dependent upon our willingness to invest” — John Plant, Executive Chairman and Chief Executive Officer · 2026-08-06. In October 2025, he said “IGT and aero are very comparable in terms of margins. So there's no dilution at all” — John C. Plant, Executive Chairman and Chief Executive Officer · 2025-07-31. So the “never thought IGT would look like it is” line is not a new discovery; it is a long-building wave that has now reached quantification. The new pieces are the three-year doubling, the $2.5 billion end-of-decade target, the SpaceX behind-the-meter detail, the CPP competitive question, and the explicit multi-year elevated capex.
Defense is the other unmodeled option. “We have not yet assumed in our forward look the replenishment of inventories, which have been depleted as a result of the Iran conflict.” — John Plant, Executive Chairman and Chief Executive Officer · 2026-10-29 That means Defense sales guidance is not yet benefiting from replenishment of inventories or missile ramp. If budgets clarify after the midterms, as management hopes, there is upside. If they do not, defense remains a solid but not explosive leg.
Bottom line
The bear case is no longer that Howmet is missing the aerospace recovery. It is that Howmet is spending into a peak. The company is doing nearly everything right operationally: record growth, margin expansion, 1.3x leverage, $2 billion of free cash flow, and a $5 billion buyback authorization. But the shares are down 25% in 90 days because the market is repricing the AI-power capex cycle and the competitive intensity in hot-section castings. If IGT demand holds and defense replenishment arrives, this drawdown will look like a valuation reset inside a structural growth story. If the data-center power build pauses or CPP changes the blade economics, elevated capex becomes the problem. The quarter was excellent. The debate is about the next five years, not the last three months.