Chip Stocks and Bank Earnings Extravaganza

Bank of America delivered a strong Q1 with earnings up 17% year over year, net interest income up 9%, and provision for credit losses about $200 million below expectations; Schwab posted record revenue up 16% and repurchased $2.4 billion of stock, though results were softer than expected. In semiconductors, ASML sold 79 lithography machines and reported strong demand, while TSMC posted 35% local-currency growth with record gross margin of 66%, operating margin of 58%, and net margin of 51%, driven by AI-related high-performance computing. The episode was broadly constructive on earnings quality and AI infrastructure demand, but the tone remained measured given cyclical and valuation concerns.

Analysis

The clearest signal is that AI infrastructure is still in the “capex validation” phase, not the “demand destruction” phase. The second-order tell is that the bottleneck is shifting from model demand to manufacturing throughput: TSMC is monetizing utilization and pricing power now, while ASML is monetizing the installed base and future capacity additions with a lag. That mix argues the AI complex remains supported for the next 2-4 quarters even if headline enthusiasm cools, because the supply chain is still under-earning relative to end-demand.

The more interesting asymmetry is inside semis: TSMC is the cleaner earnings momentum story, while ASML is the better duration trade. TSMC’s margin expansion implies customers are paying up for scarce advanced-node capacity, but that also increases the risk that any later-cycle moderation shows up abruptly once capacity catches up. ASML’s order flow is the real leading indicator; if memory and logic customers keep extending tool demand, the market can keep re-rating the whole AI spend curve without needing another Nvidia-style surprise.

On financials, BAC looks like a quality cyclical winner with a less fragile earnings base than investors feared, but SCHW is the more interesting volatility trade. Schwab’s earnings are levered to market activity, not just rates, so the stock can underperform precisely when “risk-on” returns and clients trade less. That makes the setup counterintuitive: the business may be improving, but the next leg higher likely needs either sustained transaction intensity or continued buybacks to offset multiple compression. The bank/market mix also favors capital return stories over pure credit beta right now.

The most underappreciated contrarian angle is that retail-platform winners may benefit more from AI and crypto narrative expansion than the market expects. PYPL and TOST both have embedded distribution and switching costs, but the market is pricing them as if product innovation won’t matter; if either starts showing even modest revenue acceleration, the multiple response could be disproportionate. Meanwhile LYFT is the cleanest upside/downside setup if unit economics keep improving: the market is still discounting autonomous disruption as a near-term threat, but that threat is more likely to be a partnership tailwind over the next 12-24 months than an existential hit.

Key Decisions for Investors

• Long TSM on any post-earnings consolidation; use a 3-6 month horizon. Risk/reward favors continuation if AI capex stays elevated, but trim if capex growth starts slowing faster than revenue growth.
• Pair trade long TSM / short a basket of weaker AI-exposure hardware names over 1-2 quarters. The thesis is that scarce leading-node capacity should continue to outperform where end-demand is real and visible.
• Initiate a tactical long SCHW against a short position in a lower-quality market-beta financial if volatility stays elevated; if markets calm materially, fade the trade because trading activity is the key earnings swing factor.
• Buy LYFT only on weakness near current valuation levels and hold 6-12 months. The stock offers asymmetric upside if management compounds buybacks and proves current operating momentum is not a peak.
• Speculative long PYPL or TOST into the next print via call spreads, 3-6 months out. The market is underestimating how much a single quarter of better product monetization could re-rate the shares.