AI and Bonds Barbell
Michael Mignano@mignano
−4.2%
since Nov 14
The idea
I believe stocks and bonds can both work from here as valuations improve, earnings hold up, and inflation eases. So this barbells two ends: AI chips, platforms, and power and cooling gear on one side, long-dated Treasuries and bond-like dividend payers on the other, sized heaviest toward AI. I keep a small cash buffer and avoid going all in on any single name. If inflation or oil prices force rates higher instead of lower, that is the signal this view is wrong.
What it owns
- The semiconductor makers supplying the processors that power every major artificial intelligence buildout, where earnings growth has outpaced the broader market. · NVIDIASupplies the vast majority of processors running artificial intelligence systems, making it the core exposure to semiconductor demand from major cloud providers building out AI infrastructure., BroadcomDesigns custom artificial intelligence chips for major technology platforms and maintains a growing dividend, bridging the growth and income sides of the take., Taiwan Semiconductor ManufacturingManufactures the processors designed by nearly every artificial intelligence chipmaker, meaning revenue from AI chips across the entire industry flows through its factories., Advanced Micro DevicesSupplies the second major source of artificial intelligence processors, offering buyers an alternative to single-source dependence while expanding into complete rack-scale systems.30%
- The technology giants investing heavily in artificial intelligence infrastructure while monetizing it through cloud services and software, with the cash generation to sustain spending through market volatility. · AlphabetMonetizes artificial intelligence through Google Cloud, search advertising, and significant investments in AI companies, making it a direct play on AI platform adoption and profitability., MicrosoftDelivers artificial intelligence to enterprises through cloud services and productivity software while developing proprietary chips to reduce infrastructure costs., AmazonOperates the largest cloud infrastructure business and holds stakes in artificial intelligence companies, gaining as enterprises scale computing resources for AI workloads., Meta PlatformsUses artificial intelligence to improve ad targeting with results visible directly in earnings, making it the platform where AI's financial impact is most transparent.24%
- Yields have reached levels not seen in many years, suggesting potential price appreciation if monetary conditions ease. · iShares 20+ Year Treasury Bond ETFHolds government bonds maturing in twenty years or longer, so it captures the full upside if yields decline from their highest levels in two decades., iShares 7-10 Year Treasury Bond ETFHolds intermediate government bonds maturing between seven and ten years, providing more stable price movements than longer-duration bonds while still benefiting from falling yields.21%
- Steady dividend-paying companies that typically move inversely to interest rates, providing ballast on the defensive side of the take alongside bonds. · NextEra EnergyA major utility with stable dividend income and growing sales to data centers, providing bond-like price behavior while capturing energy demand from artificial intelligence infrastructure., Philip Morris InternationalGenerates revenue from consumer staple products with steady demand across economic conditions and pays a high dividend, anchoring portfolio stability., Realty IncomeOwns retail properties under long-term leases and pays monthly dividends, moving like a bond while providing steady income independent of technology cycles., AbbVieA pharmaceutical company with a long track record of dividend growth independent of economic cycles, providing earnings-backed income on the defensive side of the take.13%
- The equipment makers providing electrical infrastructure and thermal management systems that every large-scale artificial intelligence data center requires to operate. · Vertiv HoldingsManufactures the electrical and thermal management equipment that artificial intelligence data centers require, benefiting from buildout regardless of which chip architectures prevail., GE VernovaSupplies turbines and electrical grid equipment necessary to power new artificial intelligence data centers, tying revenue directly to infrastructure build-out.7%
- Cash5%