I want to talk about a chemical that makes most investors' eyes glaze over - chlorine trifluoride - because the real story isn't the chemical itself. It's the two industrial gas giants that dominate its supply, and whether their stocks still deserve the premium the market pays them.
The headline around chlorine trifluoride is that China, South Korea, and Taiwan are driving demand through 2036, with Air Liquide, Linde, and Merck KGaA strengthening electronic gas supply chains. It sounds like a growth breakout. In reality, the global chlorine trifluoride market was valued at roughly $72 million in 2025 and is projected to reach somewhere between $83 million and $114 million by 2033. A CAGR of 4 to 6 percent on a $70 million base is not a catalyst event. It's a rounding error for any of the three named suppliers.
But here's what actually matters. Chlorine trifluoride is just one specialty gas among dozens used in semiconductor fabrication - etching, cleaning, doping - and the broader electronic gases market is a different beast entirely. It was valued at $6.8 billion in 2025 and is projected to reach $12.4 billion by 2034, growing at nearly 7 percent annually. That is the real demand base, and it's the one that turns Linde and Air Liquide into something that looks remarkably like infrastructure.
Let me start with the cash flows. Linde generated $10.4 billion in operating cash flow for all of 2025, after investing $5.3 billion in capital expenditures. That leaves roughly $5 billion in free cash flow - money available for dividends, buybacks, or M&A. Air Liquide reported gas and services revenue of €25.8 billion in 2024, with its Electronics business segment growing 8.2 percent. For 2025, the company declared a €3.70 dividend per share against a payout ratio in the low-60s range. Both companies run businesses that are, for practical purposes, fee-based: they lock semiconductor foundries into long-term take-or-pay contracts, deliver gas onsite from plants built next door to the fab, and collect on a schedule that doesn't bend with commodity prices or macro headlines.
From a cash-flow predictability perspective, this is the closest thing to a midstream toll-road model you'll find outside of energy. Packaged and specialty gases deliver 55 to 60 percent gross margins, and the contract duration stretches into double digits. TSMC doesn't cancel its gas supply agreement because consumer spending softens in Q2. That predictability premium is exactly what justifies the multiples.
Now let's talk about valuation, because this is where the story gets uncomfortable. Linde trades around 13 to 19 times EV/EBITDA depending on which measure you use, and Air Liquide sits in a similar band - roughly 13 to 15 times on some measures, higher on others. The Price-to-Free-Cash-Flow ratio for Air Liquide has run near 39 times on a trailing basis. These are not value entry points. They are quality premiums, and the market is charging them.
For comparison, Linde's historical net leverage average has run between 2.5 and 3.0 times. As of end-2025, net leverage sat near 1.1 times - well below its own history. The balance sheet is conservatively managed, though net debt did climb from $16.8 billion at the end of 2024 to $21.9 billion at the end of 2025. Air Liquide has approximately 88 percent of its debt fixed-rate, which insulates cash flow from rising borrowing costs. Both balance sheets are solid, but they don't offer the kind of distressed discount that creates margin of safety. There is no bargain bin here.
While it's true that the semiconductor tailwind is real - global semiconductor sales hit $791.7 billion in 2025 according to the SIA - the market already knows it. These stocks are priced for execution, not for discovery. The gap between price and intrinsic value that makes a deep-value case is narrow or nonexistent. Re-rating potential from peer parity isn't the driver because there is no significant peer discount to close.
Merck KGaA is an interesting contrast. Its Electronics segment posted net sales of €886 million in the first half of 2025, down 7.4 percent on a reported basis and 5.6 percent organically. For a company with €21.1 billion in total net sales, the electronics gas business is a sliver - not the backbone. That makes Merck a poor analogue for the Linde-Air Liquide model, where specialty gases are a meaningful margin accretor atop a massive contracted revenue base.
All things considered, the chlorine trifluoride headline is a distraction. The electronic gases business that carries it is genuinely durable, and the cash flow it produces is infrastructure-grade. But Linde and Air Liquide are not the hidden-value compounders they might have been five years ago at lower multiples. They are well-managed, well-priced quality names trading at quality multiples. Even if semiconductor demand accelerates further, the upside from here is execution-dependent, not mispricing-dependent.
I would rate these a Hold at current levels. The cash flows are excellent, the balance sheets are clean, and the contract model is recession-resistant. But value investing requires a margin of safety, and at 13 to 15 times EV/EBITDA with P/FCF near 39 times, the market has already done its homework. There are better risk-adjusted opportunities elsewhere - names where the cash-flow durability is comparable but the multiple hasn't caught up to the quality. That's where the asymmetric returns live.