Air Products Bets on Discipline Over Ambition: New CEO Draws a Hard Line on Project Size, NEOM Timing, and the Louisiana Blue Decision
Bernstein 42nd Annual Strategic Decisions Conference, May 27, 2026
Air Products and Chemicals CEO Eduardo Menezes used his first major public conference appearance since restructuring the company's engineering organization to deliver a remarkably candid message: the era of multi-billion-dollar speculative megaprojects is over, the Louisiana Blue hydrogen-to-ammonia project has not yet gone to the board, and NEOM's first ammonia volumes are now expected in early 2027. For investors who have spent the past three years trying to understand what Air Products actually is, the picture coming out of Bernstein today is considerably clearer — and considerably more conservative — than the one the company previously projected.
Louisiana Blue: No Board Decision Yet, July Is the Target
The single most important disclosure of the day came on Louisiana Blue, Air Products' proposed blue hydrogen and ammonia project on the U.S. Gulf Coast. Contrary to market speculation that construction bids had already been submitted to the board and sent back for revision, Menezes flatly corrected the record. "We didn't take it to the board yet," he said. "We're going to — we're working on the project, we're working with our suppliers, we're working for our customer. We are trying to finish everything for the next board in July, and then we'll try to make a final decision."
The framing around the decision itself was notably unsentimental. Menezes acknowledged the core structural problem — that the project was conceived without a committed end customer for the product — and declined to speculate on whether the economics might force a renegotiation with the offtake partner. "If it's a good project, we'll do it. If it's not a good project, we're not going to do it," he said. At current ammonia prices, he noted almost parenthetically, "any project would be printing money today, but that's not how you need to think about that." The July board meeting is now the pivotal catalyst for this project.
NEOM Is On Track, But the Ramp Is the Unknown
On NEOM, the $8.5 billion green hydrogen-to-ammonia project in Saudi Arabia, Menezes confirmed that the renewable power infrastructure — nearly four gigawatts of wind and solar — is essentially complete and that substation energization is underway. The bottleneck is the electrolyzer commissioning process, which is sequential by design. "We have 110 electrolyzers, and you cannot commission 110 at the same time," he explained. "You need to go one by one, and you cannot start commissioning the ammonia loop before we have a certain amount of hydrogen running." The base case is first ammonia production at the start of 2027, but the pace of ramp to full capacity remains the open question.
The commercial structure of NEOM carries significant commodity price risk that Menezes was direct about. Air Products holds a 30-year obligation to purchase the entire ammonia output at essentially a fixed price — there is no meaningful raw material cost variation given the renewable power input — and its financial return depends on the spread between that contracted cost and prevailing ammonia market prices. "The more the time goes on, the better this project will look like," Menezes said, pointing to the lack of meaningful cost escalation on the buy side. But the ammonia price risk is real and not fully hedged, with the Yara preliminary marketing agreement providing only partial insulation.
Electronics Is Accelerating, Not Slowing
Investors who have been concerned that semiconductor fab CapEx has been running negative for two consecutive years may be misreading the signal for Air Products specifically. Menezes described the company's semiconductor business — roughly 17% of revenues — as sitting "right in the middle of the largest CapEx cycle the industry has ever seen." The specific data point he offered was striking: the Phase 5 expansion at Air Products' Korean semiconductor site will be three times the size of Phase 1, and the total gas volumes across all industrial gas suppliers at that single site now exceed the nitrogen consumption of the entire U.S. Gulf Coast. The Taiwan multiphase project announced previously was $900 million, and the Korea project is described as larger than that, spread over four years.
Productivity on Track, $100 Million Target for This Year
CFO Melissa Schaeffer confirmed that the productivity program initiated in 2023 is running on schedule. Headcount has been reduced by more than 10% from peak levels, and the company is targeting $100 million in cost savings for fiscal 2026, with roughly $50 million already delivered through the first two quarters. The program is expected to wind down in 2027. Menezes was clear that headcount reduction is only part of the story — power consumption optimization, natural gas efficiency, and distribution rationalization are all components of a global productivity organization that he is rebuilding along the lines of what he ran at Linde and Praxair.
The Engineering Overhaul: China Is Now the Center of Gravity
Perhaps the most structurally significant organizational change that has received little attention externally is the complete restructuring of Air Products' engineering and fabrication operations. Menezes was unusually direct in explaining what went wrong under prior leadership: the engineering organization expanded from roughly 1,000 people to 5,000, including more than 2,000 engineers in India working on U.S. and European projects without deep familiarity with industrial gas culture or Air Products' standards. "That created issues on the business," he said flatly.
The solution is consolidation around the company's manufacturing facility in Caojing near Shanghai, which Menezes described as "first class, second to none." A Chinese executive now runs all industrial operations engineering globally. "The engineering, the cold boxes, everything will come from there," he said. The logic is straightforward: volume drives innovation, and Air Products' Asian operation has consistently delivered on time and on budget in ways the expanded global engineering organization did not. Schaeffer reinforced the point, noting that the Asian team "takes it personal when there's a dollar overrun."
Guidance Looks Conservative, but Management Is Deliberately Cautious
First-half fiscal 2026 EPS growth of 15% — with Q1 up 11% and Q2 up 19% — makes the Q3 guidance of 5% to 8% growth look like a meaningful deceleration. Menezes and Schaeffer attributed this to a combination of turnaround timing (a significant planned maintenance event that was deferred from Q2 into Q3 and Q4), continued caution on European and Asian underlying demand, and potential one-time charges if the Louisiana Blue project is cancelled, which would affect capitalized interest accounting. Schaeffer also flagged ongoing caution around helium: the company has locked in longer-term supply contracts, reducing volume volatility, but at some pricing concession. The 4% helium revenue headwind for the full year is being maintained. On the full-year guidance of 8% to 10% EPS growth, management's tone suggested confidence in meeting the range but deliberate conservatism in not raising it given the number of moving parts.
Helium: Supply Secured, Market Will Be Structurally Long Again
The Qatar Energy disruption — which has effectively removed a major source of helium supply from the market — is being actively managed through Air Products' cavern storage in the U.S., which had been accumulating product over several years when the company was buying more than it was selling. Menezes confirmed that the Air Products-operated Qatari plant was not physically damaged, but that logistics — container availability, shipping insurance, airport congestion at Jeddah as operators reroute cargo to avoid the conflict zone — remain the binding constraint. He expects the plant to complete its maintenance in the near term, but when shipments actually resume is uncertain. His strategic read on helium pricing is notably unsentimental: the market was structurally long before the crisis and will return to that state within six to twelve months of resolution, which argues against locking in spot pricing gains in favor of long-term customer agreements, particularly in electronics.
Space Is Real but Still Small — and the In-Source Risk Is Genuine
Air Products has supplied liquid hydrogen, liquid oxygen, and helium to NASA for decades, and the commercial launch boom is creating material new demand. Menezes offered a vivid illustration of the scale shift: a single recent SpaceX launch consumed close to 4,000 tonnes of liquid oxygen, compared to roughly 600 tonnes for a large NASA rocket. At the cadence being discussed — one launch every other day or daily — the logistics of trucked supply become physically impossible, requiring on-site production. The company has announced a new liquid oxygen plant in Texas to serve this market. However, Menezes was candid about the principal risk: SpaceX is already building its own on-site production capability in Texas, and the fundamental question for Air Products is whether it can offer enough scale advantage and operational efficiency to win outsourcing decisions against customers who are fully capable of building for themselves. The business remains a small fraction of total revenue today.
The Cultural Reset: Monthly Reviews and Smaller Projects as a Feature, Not a Bug
Menezes described a significant increase in governance cadence since his arrival, with separate monthly full-day reviews covering business performance, project execution, and new capital proposals. His comment about decision-making under the previous regime was pointed: "If I believed that the people in the business today were involved in some of these projects, they wouldn't be in the business anymore." He attributed the prior problems to a very small group approving large commitments without involving the experienced operators who would have asked harder questions. Going forward, he was explicit that the right project size for Air Products is one where a single large program — like the Korea semiconductor expansion at approximately $1 billion over four years — represents a fraction of the annual $2 billion project target, not the majority of it. "I don't see us doing a lot of these very large projects going forward," he said.