Cloud is back in the driver’s seat, and this time the improvement isn’t just “better sentiment.” It showed up in segment revenue, in operating income, and in the magnitude of capex being pulled forward.

I went back through the latest primary sources (company earnings releases and transcripts where available) and corrected several figures in the prior draft. The headline change: Q1 2026 was stronger than what we had written, especially for AWS and Google Cloud. That matters because positioning tends to follow actual cash flow and margin durability, not just growth headlines.

Bullet Summary

• AWS Q1 2026 net sales: $37.6B, up 28% YoY (Amazon earnings release).
• Google Cloud Q1 2026 revenue: $20.028B vs. $12.260B a year ago, up about 63% YoY (Alphabet earnings release).
• Google Cloud Q1 2026 operating income: $6.598B (vs. $2.177B a year ago), operating margin about 32.9% (Alphabet earnings release + transcript commentary).
• Microsoft Cloud revenue (FY26 Q1): $49.1B, up 26% YoY (Microsoft earnings call page).
• QQQ 200-day moving average: about $614.66 as of late May 2026; prior draft’s $478 level was stale relative to current price regime (multiple technical data sources concur on the 200-day level).
• Snowflake most recent official quarter on its IR site (FY26 Q2): product revenue $1.09B, +32% YoY (company IR). (The earlier “26%” reference needed a timestamp and was not the most current company-posted number.)

Market Context Analysis

The part people skip is that cloud demand is never “one thing.” It’s a blend of: seat growth, workload intensity (especially AI training and inference), price discipline, and whether customers are optimizing or expanding. Q1 2026 finally looked like expansion winning again, and it did so across more than one hyperscaler.

Here’s the cleanest way to frame the moment for active traders: when the three biggest platforms show simultaneous acceleration, it usually drags the rest of the infrastructure stack with it, but not evenly. Compute and networking behave differently than application software. Databases behave differently than observability. And the market rarely pays everyone at once.

On the index side, the level math needed a refresh. QQQ is trading in the low-to-mid 700s area in late May 2026 depending on the day’s close, while its 200-day moving average is roughly $615. That makes the prior draft’s “200-day at $478” plainly out of date for the current regime. In other words: tech has been extended above long-term trend for a while, which changes how you should think about risk if macro volatility picks up.

Slight tangent, but it matters: when the 200-day is that far below price, the market can still go higher, but the cost of being early on a pullback increases. You don’t need a crash for pain. You just need a few weeks of digestion while the moving averages catch up.

Hyperscaler Financials: Updated Q1 2026 Scorecard

Amazon Web Services (AWS): Amazon reported AWS segment sales of $37.6 billion in Q1 2026, up 28% year over year. That is materially higher than the $29.3B figure in the earlier version, which was incorrect. The important trader implication isn’t just the growth rate; it’s that AWS is large enough now that high-20s growth changes the aggregate demand read-through for semis, networking, and power/cooling suppliers tied to data centers.

Alphabet Google Cloud: Alphabet’s earnings release shows Google Cloud revenue at $20.028 billion in Q1 2026 versus $12.260 billion in Q1 2025, roughly 63% year-over-year growth. Operating income for Google Cloud was $6.598 billion versus $2.177 billion a year ago, which puts operating margin around 32.9% for the quarter. The earlier draft’s $12.3B quarterly revenue number for Q1 2026 was actually the Q1 2025 baseline, and the “9.4% margin” was not consistent with the latest segment reporting.

Microsoft (cloud indicators): Microsoft does not disclose Azure revenue dollars directly, so any “Azure run-rate” estimate needs to be clearly labeled as an estimate and treated cautiously. What Microsoft does provide: Microsoft Cloud revenue of $49.1 billion for fiscal Q1 2026, up 26% year over year, with additional segment commentary in its Intelligent Cloud materials. This is still a strong confirmation that enterprise demand plus AI capacity build is not narrowly concentrated in one platform.

One more clean-up item: the prior draft used phrases and framing that implied precision around cross-company “combined hyperscaler revenue” and a composite growth rate. Given differences in fiscal calendars, disclosure practices, and what is included in each segment, I stripped those composite claims out. It’s safer and cleaner to anchor on what each company actually reported.

Sector Breakdown

When hyperscalers accelerate, the market usually tries to express it through three buckets:

• Compute supply chain: AI servers, GPUs/accelerators, custom silicon, memory, networking, optical components.
• Foundational software tied to usage: observability, data platforms, security, developer tooling. These names tend to show better elasticity when usage climbs, not just when budgets loosen.
• Application software: often rallies later, and only if CFO confidence returns. It’s not automatic.

What’s interesting is how different the margin dynamics are across the stack. Google Cloud showing roughly 33% operating margin is a loud signal that utilization improved and that the cost structure has matured. That can pressure smaller cloud vendors on pricing power over time, even while the overall demand pool grows.

Security is the other cross-current. Alphabet’s transcript commentary flagged the Wiz integration as a potential low single-digit headwind to Google Cloud margins through the remainder of 2026 (integration and scaling investments). That’s not necessarily bearish, but it’s a reminder that “great quarter” and “straight-line margin expansion” are not the same thing.

Stock-Specific Financial Breakdown

GOOGL: With Google Cloud at $20.028B in Q1 2026 and cloud operating income at $6.598B, the segment is no longer a “future profitability” story. It is already a major profit contributor. The debate shifts to sustainability: how much of the surge is AI-driven workload intensity vs. broad enterprise migration, and how quickly capex rises alongside demand.

AMZN: AWS at $37.6B (+28% YoY) re-centers the infrastructure read-through. At this scale, the growth rate is large enough to move supplier expectations quickly. If you trade the ecosystem, pay attention to whether AWS commentary (and backlog-type indicators) support a multi-quarter demand line, or whether Q1 was an outlier driven by a few very large AI deployments.

SNOW (fundamental anchor, using company-posted numbers): Snowflake’s investor relations quarterly results page shows product revenue of $1.09B in FY26 Q2, up 32% YoY. That gives you a clean “real economy” datapoint for data-platform consumption tied to cloud buildout. Valuation multiples move around daily, so rather than lock a forward price-to-sales number that can get stale, I’d treat the revenue growth trajectory and RPO trends as the higher-signal inputs.

A quick note on price targets: the earlier draft cited specific bank targets (e.g., $240 for GOOGL, $245 for AMZN). Those can be correct on the day they’re published and still become stale quickly due to revisions. If you want those included, I can re-add them, but only if we cite the exact report date and firm, and accept that targets can change without warning.

Technical / Trading Framework

Given the refreshed regime, the key technical point is simple: QQQ is well above its long-term trend, with the 200-day moving average around $615 and price materially above that level in late May 2026. That tends to compress the margin for error if macro headlines hit risk assets. It does not mean “bearish.” It means you should be more deliberate with entries, stops, and sizing.

If you want a clean framework without relying on jargon: watch where buyers repeatedly defend price (support), where sellers have reliably stepped in (resistance), and whether rallies are happening on stronger volume than pullbacks. In strong uptrends, pullbacks that hold above prior breakout areas tend to be the healthier kind. When price slices back through those areas and fails to reclaim them, that’s when risk management starts doing the heavy lifting.

For single names like DDOG or SNOW, I’d avoid hard-coding an RSI value or a very specific tight range unless we’re updating it every issue. Those levels can be wrong within a day. Better: define the zones and the “if/then” conditions you care about, and keep the levels current at send time.

Scenario Modeling

Bull Case: Hyperscaler commentary stays consistent into the next round of earnings, capex increases are interpreted as demand-driven rather than defensive, and mega-cap tech holds above intermediate support levels. In that world, cloud-adjacent infrastructure names with direct usage sensitivity tend to continue outperforming, and pullbacks remain shallow. A practical index marker: QQQ holds above its rising 50-day and stays comfortably above the 200-day near $615.

Base Case: Growth remains solid but the market digests the move. Expect more two-sided trade: sharp rallies on strong data, quick pullbacks on valuation or rate concerns. This is usually where relative strength matters more than absolute direction, meaning the best names keep making higher lows while weaker “cloud beneficiaries” stall. In this case, traders often do better focusing on levels and timeframes rather than holding a broad thesis.

Bear Case: Macro tightening or a risk-off catalyst forces multiple compression while capex headlines get interpreted as margin risk. Even if cloud demand stays healthy, the market can still punish the higher-multiple parts of the stack first. The index tells you when it’s getting serious: QQQ starts spending time below the 50-day, and rallies fail before reclaiming it. If the 200-day near $615 becomes the next magnet, you want to already have a plan for what you do before that level is in play, not after.

Active Trader Strategy Framework

I’m going to keep this tactical and not overly polished, because the market isn’t either.

• Define exposure by driver: If your thesis is hyperscaler capex, you want instruments that respond directly to that driver, not second-order “cloud is good” proxies.
• Separate timeframes: If you are trading 3 to 10 days, your risk is the index and rates. If you are trading 3 to 6 months, your risk is whether growth persists and margins hold.
• Size to volatility: With QQQ far above its 200-day, assume larger intraday swings and more gap risk around macro releases and AI-related headlines.
• Have a redraw line: Pick one level that, if broken and not quickly regained, forces you to reduce risk. Not because you “know” direction, but because the market is telling you the probability distribution changed.

Worth repeating: the biggest mistake in this type of environment is getting the fundamental call right and the risk control wrong. That’s how good theses turn into bad P&L.

Cloud demand accelerated, and profitability improved, especially at Google Cloud. AWS also came in much stronger than the earlier draft reflected.

But the index regime matters. When QQQ is extended above long-term trend, you can be “right” and still get chopped up. So the goal here isn’t to be heroic. It’s to be prepared, define your invalidation points, and stay flexible as the next set of earnings and macro prints arrive.