Ferroglobe Turns the Trowel on Itself: Five Plants, a $100 Million Promise, and a Critical-Minerals Option That Actually Ships
A negative-EBITDA quarter that was also, oddly, the most forward-leaning call the company has given in years — if you believe the numbers it just put a range around.
GSM · Earnings Call · 2026-11-04
The trough, quantified — and a discrepancy nobody reconciled
Ferroglobe's third quarter of 2026 was bad in a way that was easy to name: revenue of $370 million, down 2% sequentially but up 19% year over year, and adjusted EBITDA of negative $2 million. What made it interesting was the diagnosis. “Excluding the $18.5 million impact from high energy prices, the adjusted EBITDA would have been approximately $16 million in the third quarter.” — Marco Levi, Chief Executive Officer · 2026-11-04 In other words, roughly nine-tenths of the sequential collapse came from an exogenous input — power, chiefly in Spain — that the company does not control. That maps almost exactly onto a market-level theme rather than a Ferroglobe one: Middle East conflict impacting energy and freight has been one of the most persistent global keywords across the last several quarters, and Ferroglobe is simply a high-voltage casualty of it. The company's own vocabulary mirrors that: high energy and energy cost both rank at the very top of its keyword set this quarter. One small oddity worth flagging for anyone modelling off the transcript: the CEO compared the quarter to "positive $30 million in the second quarter," while the CFO referenced "a positive $13 million in the prior quarter." The gap was never explained on the call. Treat the sequential EBITDA delta with caution.From "premature" to a hard number
Here is where the quarter actually turns. In August, an analyst asked for the dollar scope of cost reductions and got deflection: “Marketing is premature. What we have decided to do really revisit our asset footprint... we will communicate more precise numbers later in the year.” — Marco Levi, Chief Executive Officer · 2026-08-05 Three months later, the number arrived, and it is not small.Against a company that just printed negative $2 million of adjusted EBITDA and carries a market capitalization of roughly $721 million, that target is existential-scale. The mechanics are the story: close 5 plants, cut corporate and operations headcount by about 20%, and lift capacity utilization from under 60% to about 80%, with a $40–60 million implementation cost net of working capital over 18 months. The first dominoes — the South African plants at eMalahleni and Polokwane — are already down. Notably, management would not name the other three, citing consultation with employee representatives. That is defensible, but it means the $100 million is a promise with an undisclosed bill of materials. What makes it credible rather than aspirational is that the *cadence* of delivery broke. In the Q2 call, the CFO explicitly confirmed costs in silicon-based alloys would rise in Q2 and then decline through the second half: “Yes. You're right.” — Beatriz García-Cos Muntañola, Chief Financial Officer · 2026-05-06 They didn't decline — energy blew them up instead. Management's response was to stop waiting for the market and attack the fixed-cost base directly. That is a different company than the one that spent 2025 talking about hiring freezes and discretionary spending.Last month, we initiated an aggressive cost reduction program to increase our capacity utilization, enhance our operating efficiency, and strengthen our competitive position. Once fully implemented, we expect the cost reduction program to enhance our annual EBITDA by approximately $100 million, plus or minus 15%.
Critical materials: the option is shipping, not slideware
The second real change is that critical material expansion moved from proposal to purchase order. Ferromolybdenum shipped to commercial customers in Q3, well ahead of a year-end target, with ferrovanadium, low-carbon ferrochromium and ferrotitanium in trials. The company's stated rationale — that existing furnaces can make these alloys with minimal capital investment — is the cheapest strategic optionality in the entire Basic Materials universe right now. The September 16 proposal to the US Department of War covers magnesium, manganese and seven additional materials; a final proposal lands this month, with negotiations starting in early 2027. The tension: Magnesium was a marquee keyword for this company as recently as Q2 — a 20,000-tonne plant pitched at $180–200 million — and it has since faded sharply in the keyword set even as it remains inside the government proposal. Management is deliberately keeping the capex-heavy leg quiet while the cheap ferroalloy leg proves itself. Investors should watch whether magnesium re-enters the narrative when the final proposal is public, or quietly stays in the appendix. On Venezuela, the posture is unchanged in substance but firmer in tone: an OFAC licence is still pending — “We should get the license anytime soon” — Marco Levi, Chief Executive Officer · 2026-11-04 — and the binding constraint is not permits but reliable electricity supply. This has been a live discussion in Venezuela for multiple quarters now; it is the one item on the strategic list that has never moved.Trade protection works; the audience doesn't
Trade protection is the one lever that has visibly paid. Silicon metal index prices rose 10% in the US and 6% in Europe in Q3; the US ITC's final determination against Australia and Norway landed August 3; the European Commission opened an interim anti-dumping review on Chinese silicon metal on September 1 and registered imports on October 23, opening the door to retroactive duties. “And now we have changed our approach. We are going for antidumping review against China.” — Marco Levi, Chief Executive Officer · 2026-11-04 This is no longer a lone grievance — a global trade defense mechanism theme sits in the market's own top-20 keywords this quarter. Ferroglobe's realized prices improved 3% in silicon metal and 5% in Manganese based alloys, and steel demand is forecast to grow 3% in Europe and 2% in North America in 2027, with aluminum production adding 2% and 3% respectively. So what did the market do with all this? Almost nothing visible — and that is the third story. The Q&A ran four questions from a single analyst before the operator declared “There are no further questions at this time.” — Operator · 2026-11-04 Compare that to prior calls where two analysts pushed through six or seven questions apiece. And after repeatedly promising to revisit the buyback "on a weekly basis," the topic vanished entirely this quarter; the only capital return mentioned was the unchanged $0.015 quarterly dividend. The company is asking investors to underwrite a self-help turnaround at a cyclical bottom with no price signal in the tape and little analyst scrutiny.The honest read: Ferroglobe just converted a vague restructuring story into a quantified one, at the cost of an ugly quarter and a promise of a still-weaker Q4. The $100 million is the whole thesis. If it lands, EBITDA roughly triples off a near-zero base regardless of whether Chinese silicon metal gets duties. If it doesn't, the critical-materials option — real, shipping, and cheap — will be the only thing left to own.Even though our third quarter earnings results are disappointing, we are taking decisive action to improve profitability... These actions are within our control and can deliver long term benefits regardless of market conditions.