
DD · Basic Materials
The market is pricing DuPont as a restructuring story still in progress, but the more interesting question is whether its semiconductor process chemistries — qualified into the most demanding fab lines on earth — are quietly becoming a chokepoint asset in AI infrastructure that no one is valuing correctly because the historical returns data is contaminated by the businesses that have already been jettisoned.
$46.73
$95.00
The switching cost moat in semiconductor process chemistries is real, but five years of sub-six-percent ROIC is the business confessing that 'specialty' in the marketing brochure doesn't yet show up where it counts — in capital returns. Serial restructuring doesn't build compounding machines; it builds complexity.
The underlying cash conversion is genuinely solid — OCF consistently running ahead of reported earnings is the signature of a business with real economic substance — but the Altman Z-Score of 0.88 and years of balance sheet scar tissue from merger-era goodwill mean the fortress walls are thinner than the income statement implies.
The AI infrastructure investment cycle is a legitimate tailwind for semiconductor process chemistries, and advanced packaging demand creates new chemistry complexity that DuPont is positioned to exploit; the organic growth rate is modest but the mix is getting meaningfully better, not just smaller.
Across every DCF scenario — including the pessimistic one — the current price implies meaningful upside, and a six-percent-plus FCF yield on a business with genuine switching costs is not demanding; the market is still applying a diversified-chemicals discount to what is rapidly becoming a focused semiconductor materials and water technology company.
Chinese state-backed domestic substitution in semiconductor chemistries is the single most credible existential threat — it attacks both the revenue line and the qualification lock-in that makes the moat durable — and PFAS regulatory overhang, governance opacity, and concentration in a cyclical semiconductor capex cycle compound the risk stack materially.
The investment case rests on a mismatch between what DuPont was and what it is becoming. Historical ROIC in the mid-single digits drags the stock's quality reputation down, but those numbers were computed over a portfolio that included businesses now sold — the remaining entity, centered on semiconductor materials and water purification, likely earns structurally higher returns than the blended history suggests. At current prices, investors are not paying for that quality re-rating, which creates the gap between the FCF yield and what the DCF scenarios imply. The trajectory matters more than the level. Every generation of chip complexity — heterogeneous integration, chiplets, high-bandwidth memory stacks, sub-2nm gate-all-around transistors — increases the chemical precision required in the manufacturing process and tightens the switching cost for qualified suppliers. DuPont is not riding semiconductor volume growth; it is riding semiconductor complexity growth, which is a more durable and less cyclical driver. The 2026 guidance of accelerating organic growth and margin expansion, if delivered, begins to close the gap between the transformation promise and the returns data that long-suffering shareholders need to see. The single biggest risk is Chinese domestic substitution in semiconductor process chemistries — not gradual erosion, but state-directed acceleration. Beijing has the capital, the industrial policy, and the national security motivation to qualify local chemistry suppliers even at performance cost, and SMIC and CXMT are under explicit pressure to reduce foreign input dependence. If Chinese fabs successfully onboard domestic alternatives to DuPont's qualified chemistries, the company loses both revenue and the lock-in economics that justify the strategic premium — simultaneously. That scenario is not today's reality, but it is tomorrow's credible risk, and it deserves far more weight than the consensus assigns it.