One Lane Closed. The Whole Rate Picture Changed

A closed Strait of Hormuz sent WTI crude to $105.42, pushed the 10-year Treasury yield to 4.602% (highest in about a year), and raised the market-implied chance of no rate cuts through 2026 from 0% a week ago to 40% by Friday. Chips sold off sharply, while Figma delivered a strong Q1 beat with revenue of $333.4 million, up 46% year over year, and raised full-year guidance to $1.42–1.43 billion. Berkshire’s Q1 filing showed a major portfolio shift under Greg Abel, including a tripled Alphabet stake, a new ~$2.65 billion Delta position, and the exit of several Todd Combs-era holdings.

Analysis

The market is starting to reprice from an inflation-scarcity regime to an inflation-persistence regime, and that changes the hierarchy of winners. The immediate beneficiaries are upstream energy and anything with embedded pricing power or short-cycle revenue reset; the real losers are duration-sensitive growth names whose multiples were built on a clean disinflation path. The key second-order effect is that higher fuel and freight costs will leak into consumer margins with a lag, so the next move is likely not a broad demand collapse, but incremental margin compression in retail, restaurants, and transport-heavy businesses over the next 1-2 quarters.

Semis are the cleanest expression of this tension. The selloff looks tactical, but the risk is that a higher-yield regime compresses the entire AI capex complex just as expectations into NVDA are most elevated. That creates a classic setup where equipment names with real order visibility can outperform the headline GPU trade if the market starts to distinguish between demand beneficiaries and valuation proxies; AMAT is the tell here. If NVDA merely meets, the stock can still disappoint because the hurdle has shifted from "beat-and-raise" to "justify a euphoric multiple in a worse macro tape."

Figma’s print matters less as a single-name event than as evidence that software is adapting faster than the bear case assumed. The more important read-through is that AI usage is becoming a monetizable layer rather than a binary replacement threat, which should support high-retention workflow software with embedded collaboration or seat expansion. That said, the market may be underestimating the margin trade-off: if AI tokens become a meaningful input cost, the winners are the vendors with pricing power and usage-based monetization, while the laggards are feature-rich platforms that cannot reprice fast enough.

Key Decisions for Investors

• Go long AMAT vs short NVDA into this week's earnings window; AMAT has idiosyncratic AI capex exposure with better visibility, while NVDA carries the highest disappointment risk after the run-up. Use a 2-4 week horizon and take profits if NVDA confirms and semis squeeze higher.
• Add a tactical long in XLE or CVX for the next 1-3 months, but hedge with an oil shock reversal trigger if diplomacy reopens the Strait and crude rolls over below the low-$90s. The asymmetry favors staying long while inflation expectations remain sticky.
• Initiate a long FIG / short software basket trade versus lower-quality AI-exposed SaaS names over 1-2 quarters. The risk/reward favors companies that can monetize AI through seat expansion and workflow stickiness rather than those facing direct feature substitution.
• Avoid chasing high-multiple software and semis on up days until after NVDA prints; the setup is fragile and the first-order move is likely already crowded. If NVDA misses even slightly, use the gap down as a better entry point only in names with real cash conversion.
• For consumer exposure, prefer short-duration hedges or underweights in retail and restaurants into the next 1-2 earnings cycles. Rising fuel and borrowing costs should pressure discretionary baskets before they show up in top-line declines.