Stocks Are Trading at Records—Strong Earnings Could Send Them Even Higher

First-quarter S&P 500 earnings are running well ahead of expectations, with nearly 90% of reporting companies beating estimates and aggregate earnings beating consensus by nearly 11% versus a 10-year average of 7%. Tech is leading the upside, with reported earnings up 45% and analysts lifting 2026 capex forecasts for the five hyperscalers by 25% since the start of the year to $680 billion, up 63% from last year. Despite strong results, market reactions have been muted as geopolitics and guidance dominate, while the S&P 500 and Nasdaq are near record levels.

Analysis

The key signal is not that earnings are beating, but that revisions are still rising after the quarter has started. That typically matters more for index direction than the initial print, because upward estimate drift can keep multiples elevated even if price reactions to individual beats are muted. In this tape, the market is effectively demanding forward guidance that confirms the next leg of capex and demand rather than rewarding backward-looking outperformance.

The clearest second-order winner is the AI infrastructure complex, but the better expression is not broad beta to software or semis indiscriminately. The hyperscaler capex step-up should pull through into test/inspection, networking, power management, and lithography-adjacent names before it benefits the mega-caps themselves; meanwhile, software vendors with credible AI monetization stories can re-rate only if management proves AI is accretive to revenue, not just cost discipline. That creates a relative-value setup where AI-enabling hardware and picks-and-shovels may outperform the platform names that already have full multiples.

The underappreciated risk is that “good enough” earnings may stop mattering if geopolitics or rates dominate the tape. After a sharp recovery to highs, the market is vulnerable to any guidance cut, margin compression from energy, or a cooling in AI capex assumptions; those would hit the highest-duration names first. Over the next 1-2 weeks, the trade is less about absolute earnings and more about which management teams validate the capex chain and which ones expose demand elasticity.

Consensus may be over-crediting the idea that low multiples alone make software attractive. If investors still fear AI displacement, cheaper names can stay cheap until managements show tangible usage, pricing, and retention data; that argues for selective longs, not a blanket sector call. The more interesting contrarian view is that the market may be underpricing the durability of semiconductor equipment demand if hyperscaler spend holds above current street assumptions into 2H, which would force another round of estimate revisions and keep index-level earnings growth ahead of price pressure.

Key Decisions for Investors

• Go long TSM vs. short a software basket into the next 2-4 weeks; the setup favors names with direct AI demand visibility over names where AI is still a narrative risk.
• Add on pullbacks to ASML and LRCX on any post-earnings volatility; use a 1-2 month horizon with a stop if hyperscaler capex commentary rolls over.
• Buy a call spread on NOW into earnings: limited downside, but the stock can re-rate sharply if management quantifies AI as revenue-positive rather than merely non-disruptive.
• Pair long GOOG/MSFT with short a lower-quality software index only if capex guidance stays firm; if guidance softens, unwind quickly because the trade is purely dependent on infrastructure spend staying hot.
• Maintain tactical underweight in TSLA ahead of its print unless management can offset multiple compression with a clear margin/volume bridge; the risk-reward is poor if the market keeps prioritizing guidance over beats.