Chip Stocks and Bank Earnings Extravaganza

Bank of America beat Q1 estimates across most lines, with earnings up 17% year over year, net interest income up 9%, and credit loss provisions about $200 million below expectations; Schwab also posted record revenue up 16% and added 1.3 million brokerage accounts, though results were slightly below expectations. In semiconductors, ASML sold 79 lithography machines for just over $10 billion in revenue, while TSMC delivered 35% local-currency growth and record gross, operating, and net margins as AI demand stayed exceptionally strong. The discussion also highlighted buybacks at Schwab ($2.4 billion) and continued investor focus on Lyft, PayPal, and Toast heading into earnings.

Analysis

The key second-order read-through is that AI infrastructure is still in an “orders first, narrative second” phase: the bottleneck is not demand but capacity, and that favors the toll-takers upstream of model adoption. ASML and TSMC together imply the spend cycle is broadening from leading-edge GPUs into memory refresh, process upgrades, and installed-base monetization, which should keep equipment and foundry utilization high even if headline AI spend growth moderates. That makes the ecosystem more resilient than a pure Nvidia-centric trade; the better risk-adjusted exposure is in the enabling stack where revenue is locked to capacity constraints rather than to end-user enthusiasm.

Banking signals are more nuanced. BAC’s strength suggests consumer credit deterioration is still more “late-cycle normalization” than outright stress, which reduces near-term systemic risk and supports a stable funding backdrop for financials. SCHW is the more interesting tell: elevated trading and asset inflows are powerful in volatile tape, but that also means earnings quality is unusually sensitive to market calm; if volatility fades into summer, the earnings setup can deteriorate even as the stock looks optically cheap on peak activity. The market may be underpricing that lumpy mix of brokerage beta plus bank-like rate sensitivity.

The most attractive contrarian setup is in businesses where the street is extrapolating disruption too mechanically. Lyft’s multiple is compressing as investors price autonomous-vehicle displacement, but the near- to medium-term path likely benefits from the opposite: platform scale, buybacks, and incremental autonomy partnerships that extend the demand curve rather than end it. Toast has a similar dynamic—consensus is worried about SaaS commoditization, but the embedded workflow depth creates switching costs that are hard to unwind quickly, so the burden of proof is on bears to show actual churn, not just AI rhetoric.

Key Decisions for Investors

• Long TSM / ASML into the next 1-2 quarters: favor the foundry-capacity and tool-enablement trade over NVDA if you want AI exposure with less multiple risk; risk/reward improves if capex commentary stays elevated and memory demand continues to accelerate.
• Add BAC on weakness over the next 4-8 weeks: this is the cleaner financials expression versus SCHW because earnings are less dependent on trading volumes; upside comes from operating leverage and credit normalization, while downside is mostly a macro recession shock.
• Trade SCHW tactically, not structurally: buy only on post-earnings or volatility-driven dips, and hedge with a short financials vol position if available; the stock can underperform quickly if market calm reduces trading revenue over the next 1-2 quarters.
• Initiate a LYFT long on valuation dislocation, ideally paired vs UBER or against a basket of autonomous-disruption beneficiaries; the thesis needs 2-3 quarters of execution plus buybacks to re-rate, with downside limited by current low free-cash-flow multiple.
• Build a TOST starter long ahead of earnings with a 6-12 month horizon: the key catalyst is proof that the installed base can re-accelerate without the prior big account, and the bear case only wins if churn or net adds visibly slow.