Earnings season takeaways: resilient, but for how long? By Investing.com

Goldman Sachs says Q1 S&P 500 earnings grew 26% year over year, with real revenue growth ex-energy up 6.3%, the strongest since 2021. The firm sees consumer spending remaining resilient for now but expects household demand to slow in the second half as low savings, weak real disposable income growth, and persistent inflation pressure spending. AI investment remains a key offset, with nearly two-thirds of S&P 500 management teams citing the theme and hyperscale capex expected to exceed $750 billion in 2026.

Analysis

The market is implicitly telling us that AI infrastructure is still the cleanest way to express a late-cycle U.S. growth regime: earnings are fine, consumers are fading, and capex is doing the heavy lifting. That favors the picks-and-shovels more than the end-platforms because hyperscaler spend is sticky even if sentiment rolls over; the second-order beneficiary is the network of power, cooling, memory, and assembly suppliers, while the losers are sectors exposed to weaker household demand and rising input costs.

For TSLA, the upgrade is less about near-term auto fundamentals than optionality on a robotics narrative that can re-rate the multiple before the cash flow shows up. The important nuance is that this works only if investors are willing to underwrite a 2-3 year software/robotics call; if macro softens into H2, the core vehicle business becomes a larger anchor and the stock’s sensitivity to financing conditions rises. In other words, TSLA can outperform on story, but its downside remains tied to consumer credit and auto affordability.

SMCI and APP remain more tactical expressions of the same AI spend wave, but they are vulnerable to any pause in capex guidance or margin compression from component inflation. The better trade is not chasing absolute momentum, but using those names as beneficiaries of continuing spend while hedging with a short against a more consumer-sensitive or margin-repair story. Goldman’s own macro setup argues for a bifurcated tape: secular AI winners can still work even as broad cyclicals and discretionary exposure lose altitude.

The consensus may be underestimating how quickly this becomes a quality filter: if growth slows, investors will pay up for revenue tied to non-discretionary enterprise capex and punish anything reliant on household demand. The bigger risk to the bullish AI trade is not demand disappearing, but the market realizing it already extrapolated too much of 2026 into current prices; that makes near-term upside more selective and lowers the odds of broad multiple expansion.

Key Decisions for Investors

• Long SMCI vs short a consumer-discretionary proxy for 1-3 months: own the direct AI capex beneficiary while shorting weaker household-demand beta; target 15-20% relative outperformance if hyperscaler spend revisions continue.
• Buy TSLA call spreads 6-12 months out to express the robotics rerating without taking full downside from auto-demand deceleration; best entry is on post-rally consolidation, with a favorable skew if the market starts underwriting a 2026 robotics catalyst.
• Use APP as a tactical long only on pullbacks; pair with a short in a margin-fragile software name if you want to isolate ad-tech/AI optimization exposure versus broader enterprise spend risk.
• Avoid chasing GS here; the macro read is constructive but not accelerating, so financials are more of a confirmation trade than a catalyst trade over the next 1-2 quarters.
• If AI capex guidance starts to flatten, rotate from pure beta names into infrastructure-adjacent suppliers with recurring demand and better downside protection; that transition should be made quickly because the repricing will likely happen over days, not months.