The sustainability of the AI-fueled technology stock rally may be determined tonight. On April 29th, Alphabet, Microsoft, Amazon, and Meta are scheduled to release their first-quarter earnings reports, with Apple following with its disclosure on Thursday. These four companies collectively account for 18% of the S&P 500's market capitalization, and their capital expenditure guidance is viewed as a crucial indicator for gauging the intensity of AI computing investment.
Ahead of the earnings releases, market optimism regarding AI capital expenditures has already faced challenges. According to The Wall Street Journal, OpenAI missed its targets for both revenue and user growth, raising doubts about the feasibility of its pledged $1.5 trillion investment in computing power. This report applied brakes to the enthusiasm for AI capital spending, pressuring U.S. tech stocks overnight and causing the Nasdaq Tech Index to fall over 1.3%.
Against this backdrop, Wall Street's focus has shifted beyond short-term profit statements. The core questions are: How long can the expansion of capital expenditures continue? To what extent can revenue growth translate into profits?
**After $600 Billion in Capex: The AI Narrative at a Crossroads** Market consensus anticipates that the combined capital expenditure of Amazon, Microsoft, Meta, and Alphabet will reach a massive $600 billion by 2026, climbing further to over $800 billion by 2027. According to Goldman Sachs estimates, by that time, nearly all of the free cash flow from the "Tech Megacaps" would be consumed by capital expenditures, with companies like Meta potentially no longer generating positive cash flow.
Rich Privorotsky, a Goldman Sachs trading desk executive, offered an ironic observation: in an industry widely perceived as having insatiable demand for computing power, the real distortion is actually on the demand side. He termed this phenomenon "token maximization"—where engineering teams compete to consume as much computing power as possible, as insufficient spending is seen as a career risk. This distorted incentive structure is driving significant, potentially inefficient, capital expenditure.
Privorotsky also highlighted a structural contradiction: Semiconductor stocks are direct beneficiaries of rising capital expenditures and therefore welcome sustained spending increases. However, for shareholders of the hyperscale cloud providers (Amazon, Microsoft, Google, Meta), the increased capital spending over recent quarters has not been rewarded by the market, as the return on investment for AI remains difficult to quantify. "Last week we heard from the suppliers, and the narrative for the hyperscalers, who are the spenders, is much murkier."
In his view, the critical question is not whether demand is strong, but whether capital expenditure can continue to rise. "If capex stops growing, that's a problem for the narrative. In a world of rising input costs, flat capex is, by definition, a slowdown."
**What to Watch in the Earnings Amid Crowded Positioning** From a technical perspective, mega-cap tech stocks have undergone a rapid repricing in recent weeks. The RSI for Goldman's mega-cap tech basket fell below the oversold threshold of 30 on March 30th, only to reach an overbought level of 74 by April 17th, an extreme swing completed in just three weeks. Following the rally, positioning has become significantly more crowded. Goldman's proprietary data shows the long/short ratio is at the 88th percentile of its three-year range, indicating historically high levels.
In this context of highly crowded positioning, Wall Street is closely watching for signals from the tech giants' earnings. JPMorgan trader Brian Heavey ranks the crowdedness of positions from highest to lowest as: Amazon, Meta, Alphabet, Apple, Microsoft. Goldman's ranking differs slightly, viewing Alphabet as the most crowded long position.
* **Amazon (Position #1)**: JPMorgan rating "Overweight," target price $280. The market is focused on whether capital expenditure guidance can support AI computing demand and if AWS revenue growth continues to benefit from AI workload migration. AWS growth was 24% last quarter, with AI services reaching an annualized revenue run rate exceeding $15 billion; maintaining this momentum is key tonight. A growth rate falling below 20% could be a turning point for market sentiment this earnings season. * **Meta (Position #2)**: JPMorgan rating "Overweight," target price $825. Goldman calculations suggest that, based on the current capex trajectory, Meta may cease to generate positive free cash flow. The market is watching the pressure of capital expenditures on free cash flow and the effectiveness of AI recommendation algorithms in boosting ad revenue. The $135 billion annual capital spending plan requires CEO Mark Zuckerberg to repeatedly justify it each quarter. Any phrasing suggesting "ongoing evaluation" could be interpreted by the market as a potential pullback; past hesitations have led to negative after-hours reactions. * **Alphabet (Position #3, Goldman's Most Crowded)**: JPMorgan rating "Overweight," target price $395. The focus is on whether search advertising is being impacted by AI chat tools and the extent to which the cloud business benefits from AI compute demand. Google Cloud is expected to grow 49.6%, but Gemini's commercialization has yet to provide clear financial figures. If Cloud revenue beats expectations but monetization signals remain vague, the market reaction could be cooler than anticipated. * **Apple (Position #4)**: Target price $325. The market is attentive to its response to rising memory costs and the impact of leadership changes. iPhone revenue grew 27% year-over-year to $59.5 billion in Q2 FY2026, with a gross margin of 48.5%; Q3 guidance suggests iPhone sales of $50.1 billion and a gross margin of 47.6%. Investors are particularly focused on the resilience of gross margins. * **Microsoft (Position #5, Least Crowded)**: JPMorgan rating "Overweight," target price $550. Privorotsky notes that Microsoft is likely to provide commentary on its FY2027 capital expenditure, offering a key reference point for judging the hyperscalers' medium-to-long-term investment outlook. Additionally, the market is watching Azure cloud growth and the revenue contribution from AI services. Azure growth consensus is around 38%, but the real interest lies in whether Copilot can deliver quantifiable enterprise revenue. Guidance below 36% for the next quarter would be negative, while a figure exceeding 40% would be a positive surprise.
Another suspense surrounding the tech giants reporting on the same day is whether positive surprises will be shared or negative shocks will prove contagious. With the AI narrative at a crossroads, direction awaits tonight's results.
Comments