Investors should prioritize companies with high returns on invested capital as AI-driven borrowing costs rise, favoring infrastructure plays like ASML and GEV over firms with weakening cash flow profiles.
ASML maintains a unique competitive advantage as the sole provider of EUV lithography systems, which are essential for advanced chip manufacturing. The company has demonstrated significant pricing power and strong demand, evidenced by its raised full-year guidance and a capacity sprint that is already largely spoken for by customers.
The company's moat is built on complex calibration processes that competitors have failed to reverse-engineer, even after taking machines apart.
Oracle faces significant headwinds due to declining returns on assets and a heavy debt burden, leading to an S&P downgrade to BBB-. Analysts argue that the company's massive capital expenditure is not yielding the same value as hyperscalers, as its infrastructure buildout is less efficient than the services layer provided by competitors.
The concern is not the spending itself, but the efficiency of the returns; Oracle's capacity buildout is viewed as less productive than the services layer monetized by hyperscalers.
Rising Treasury yields near 4.79% reflect heavy corporate borrowing for AI infrastructure rather than economic distress. While companies like ASML, GE Vernova, and Comfort Systems leverage strong backlogs and pricing power to thrive, Oracle and Rocket Companies face deteriorating returns on assets that threaten their long-term valuation.
Rising yields are primarily driven by massive corporate demand for capital to fund AI infrastructure, with estimated debt issuance reaching $1.5 trillion this year.
The key metric is the spread between a company's borrowing cost and its return on invested capital; companies funding projects with high returns are compounding value, while others are destroying it.