FEATURED READ

When Beating Estimates Is Not Enough

Hey there, bargain hunter.

Let’s start with the part that does not make sense on the surface. Meta reported its fastest revenue growth since 2021 last night — 33% year-over-year, $56.31 billion in revenue, net income up 61%, EPS that blew past the street estimate by nearly a dollar. A clean, dominant quarter by almost any measure you want to use.

The stock fell 6% after hours and shed roughly $103 billion in market cap.

That is the whole story of this earnings season compressed into a single data point. The numbers are not the question anymore. The question is what the spending required to produce those numbers looks like three years from now — and whether the cash flowing out the door today will ever come back at the scale being promised. The market has been asking that question politely for about two years. Last night, it started asking loudly.

The Numbers, Quickly

All four hyperscalers beat on revenue and EPS. All four are spending more than the street expected on AI infrastructure. And the market treated each one differently enough that you would barely guess they were all in the same sector.

• META: Revenue $56.31B, +33% YoY vs. $55.45B estimate. Adj. EPS $7.31 vs. $6.79 estimate. Net income $26.77B, +61%. After-hours: -5.7%.
• GOOGL: Revenue +20% YoY — highest growth rate since 2022. EPS $5.11 vs. roughly $2.63 estimate. Cloud revenue $20.02B, +63% YoY vs. $18.05B estimate. After-hours: higher.
• AMZN: Revenue $181.52B, +17% YoY vs. $177.30B estimate. EPS $2.78 vs. $1.64 estimate — a 69% beat. AWS +28% YoY to $37.59B. After-hours: +4%.
• MSFT: Revenue +18% YoY vs. roughly $81.4B estimate. Non-GAAP EPS $4.13 vs. $4.04–$4.06 estimate. Azure +40% YoY vs. roughly 31% consensus. After-hours: mixed, modestly higher.

Sources: CNBC, LSEG, StreetAccount, StockTitan, Bloomberg — April 29, 2026

The broader Q1 season context matters here. The S&P 500 is tracking its 11th consecutive quarter of year-over-year earnings growth. Eighty-four percent of companies have beaten EPS estimates and 81% have beaten on revenue — both above every meaningful historical average. The blended earnings growth rate for the index sits at 15.1%. The Mag 7 group was expected to post 20.3% earnings growth and 22% revenue growth. Most of that happened. The reaction to it is the more instructive data point.

This Was Not a Normal Earnings Week

The U.S. began combat operations in Iran in late February. Brent crude is above $107 per barrel. Component pricing for AI infrastructure — the GPUs, the networking hardware, the cooling systems — is rising in a supply environment that was already strained before any geopolitical disruption entered the picture. Meta did not bury this. They put it directly in the earnings release as the reason for raising capex guidance. That is notable. These companies usually describe macro risks in the most abstract terms possible. Citing a specific military conflict as a cost driver is a different kind of disclosure.

The VIX closed Friday at 18.71. Options markets going into last night were pricing 5–7% implied moves across all four names — wider than historical averages. Semiconductor call premiums were running 25% larger than put premiums, which tells you the directional lean in the options market was bullish on the AI supply chain heading into the prints. That positioning had consequences, which we will get to.

Meta: What a Beat Without a Reward Looks Like

The ad business is genuinely working. Ad impressions grew 19% year-over-year. Average price per ad rose 12%. Daily active people across the Family of Apps hit 3.56 billion — up 4% from a year ago. Instagram Reels engagement was up 10% in Q1, attributed directly to AI-driven ranking improvements. The $8.03 billion income tax benefit Meta recognized inflated the net income headline, but even stripped out, the underlying business is producing at a high level. Q2 guidance of $58–$61 billion in revenue is solid.

None of that is why the stock fell.

Meta raised its full-year 2026 capex guide to $125–$145 billion, up from $115–$135 billion. The street was modeling roughly $122.6 billion. The new midpoint is $135 billion. The stated reason: higher component pricing driven by supply chain disruptions from the Iran conflict, plus additional data center build-out. Full-year expenses were guided to $162–$169 billion. That is a company spending aggressively into a cost environment that is moving against it, while simultaneously cutting roughly 8,000 employees — about 10% of the workforce — and closing 6,000 open roles.

What Zuckerberg is doing is not complicated to understand. He is betting that AI replaces the productivity of the humans he is cutting, that the infrastructure investment compounds into dominant ad targeting and new product categories, and that investors will eventually price that correctly. He may be right. The market last night was expressing some doubt about the timeline.

Muse Spark — Meta’s first foundation model out of its Superintelligence Labs — launched this quarter. That is a longer-term story by any honest measure. But it is worth tracking as a marker of whether Meta’s internal AI development eventually reduces its dependence on external model providers. That dependence is a cost and a risk that does not show up cleanly in any single quarter’s financials.

Alphabet: The Quarter the AI Bull Case Needed

If you had to pick one report from last night that the broader AI investment thesis depended on, it was Alphabet’s. And it delivered.

Net income of $62.57 billion, up 81% year-over-year. Revenue up 20% — the highest growth rate since 2022. Google Cloud came in at $20.02 billion, up 63% year-over-year, against an $18.05 billion estimate. Google advertising revenue hit $77.25 billion, up 15.5%. Search grew 19% with queries at an all-time high. Sundar Pichai said on the call that enterprise AI solutions became the primary growth driver for cloud for the first time in Q1 2026. Not a contributing factor. The primary driver.

That is the number that matters most across the entire group. It means the hyperscaler capex is generating enterprise revenue — not just future optionality. The AI spending is showing up on the revenue side of the ledger in real time. Alphabet’s cloud backlog ended 2025 at $240 billion, up 55% sequentially. The company guided 2027 capex to be significantly above its already elevated 2026 range of $175–$185 billion, which is itself roughly double its 2025 actual spend of $91.4 billion.

One soft spot worth flagging: YouTube advertising came in at $9.88 billion versus a $9.99 billion estimate. A small miss, but YouTube subscriptions are now growing faster than YouTube ads. That composition shift has margin implications that will matter more over time than it does in any single quarter.

Amazon: Clean Number, Complicated Cash

Amazon’s AWS grew 28% year-over-year to $37.59 billion. The street was modeling 18% consensus growth. Bulls were hoping for 20%. Getting to 28% is not incremental outperformance — it is a different kind of acceleration that suggests enterprise AI pull-through into cloud infrastructure is stronger than most models were capturing.

Total revenue of $181.52 billion beat the $177.30 billion estimate. EPS of $2.78 crushed the $1.64 consensus — a 69% beat. Advertising revenue of $17.24 billion also topped expectations. Q2 guidance landed at $194–$199 billion in revenue, implying 16–19% growth. The stock was up more than 4% in extended trading and the read is not complicated.

Here is the part that does not make the headline. Amazon’s free cash flow for the trailing twelve months fell to $1.2 billion. That is a 95% decline year-over-year. The capex machine — guiding to roughly $200 billion in 2026 spending, up from $131.8 billion in 2025 — is consuming cash at a rate that would alarm most investors if the growth rate were not also accelerating. The market is extending trust on the growth story. That trust is not unconditional.

Brief tangent that is actually worth noting: Amazon’s Leo satellite service — the Starlink competitor — is targeting commercial launch in Q3 2026 with 270 satellites currently in orbit against a planned constellation of 7,700. Andy Jassy named Delta Airlines as a prospective customer. This does not move the quarterly model. But it is the kind of business-within-the-business that Amazon keeps building quietly while everyone watches AWS. Worth keeping a mental tab on.

Microsoft: Good Quarter, Wrong Moment

Azure at 40% growth is a strong number. It came in well above the roughly 31% the street was expecting. Non-GAAP EPS of $4.13 beat the $4.04–$4.06 estimate. Revenue grew 18%. The company returned $10.7 billion to shareholders through dividends and buybacks. On pure fundamentals, this is a solid quarter.

The problem is not the quarter. The problem is the story around the quarter.

Microsoft ended its exclusive right to sell OpenAI models this quarter, opening that distribution to Amazon and others. That exclusivity was a real moat — it meant Azure was the only major cloud platform through which enterprises could access OpenAI’s frontier models at scale. That advantage is now gone. GAAP EPS of $3.72 was dragged down by losses on OpenAI investments, a line item that will keep creating noise in the financials for as long as the relationship continues in its current structure. Cloud gross margin percentage also declined as AI infrastructure costs scale.

Microsoft is down 12% year-to-date — the worst-performing hyperscaler in the group. The after-hours reaction was muted rather than positive. The market wants Copilot data. It wants evidence that the AI product layer is generating enterprise revenue, not just Azure workload migration. At roughly 22x forward earnings, there is not much room for execution risk to compound before the multiple becomes a problem.

$630 Billion Going Out the Door in 2026

This is the number options traders need to sit with beyond the individual stock moves. The combined AI infrastructure capex from these four companies in 2026 is tracking toward $630–$650 billion. For context on scale:

• Alphabet: 2026 capex guide $175–$185B vs. $91.4B actual in 2025 — roughly a 100% increase.
• Amazon: Approximately $200B guided for 2026 vs. $131.8B in 2025 — roughly a 52% increase.
• Meta: Raised guide to $125–$145B (midpoint $135B) vs. $72.2B in 2025 — a 75–100% increase.
• Microsoft: Q3 capital expenditures alone were up 89% year-over-year. Full-year 2026 tracking toward approximately $146B.

Nvidia does not report this week. It does not need to. The capex guidance from these four reports is more instructive about Nvidia’s forward revenue than any analyst model. Alphabet and Amazon accelerating cloud growth confirms AI compute demand is being absorbed at the enterprise level. Meta raising capex while citing cost inflation signals that demand is strong enough to absorb rising input prices. Microsoft holding its spending cadence despite margin pressure signals commitment regardless of short-term financial pain.

Tuesday of this week, Nvidia, AVGO, and AMD sold off 2–5% on a WSJ report suggesting OpenAI missed internal revenue targets. Tonight’s prints from four of OpenAI’s largest infrastructure customers directly contradict that read on AI demand. That Tuesday selloff looks like an overreaction in light of what Alphabet and Amazon just reported.

What the Options Market Was Doing Before Any of This

Pre-earnings implied volatility across all four names was pricing 5–7% moves. That is elevated — historically, mega-cap earnings events in stable macro environments price 3–4% implied moves for names this size. The wider pricing reflects genuine uncertainty about capex guidance, not just earnings beats and misses.

Here is where it gets costly if you were positioned wrong. Implied volatility compresses sharply the moment earnings are released — regardless of which direction the stock moves. That compression is called the IV crush, and it is the single most important mechanic for options traders to understand heading into any major earnings event.

Meta’s pre-earnings one-week at-the-money implied volatility was running near 65–75%. Post-earnings, that number historically collapses to the 23–28% range. A $700 call expiring May 1, trading around $15 pre-earnings, needed Meta to close above $715 just to break even — before factoring in the IV crush eating the remaining premium. Meta closed near $671 going into the report and fell to roughly $630 after. That $700 call went from $15 to approximately $3.50 or less as IV compressed from 75% toward 30%. A trader who was directionally correct that Meta would beat every estimate still lost money on that position. The fundamental call was right. The structure was wrong.

That is the core lesson. In mega-cap earnings, being right about the business is not sufficient. The structure of the trade has to survive the volatility collapse that follows the release — in every direction.

How to Think About Each Name Now

IV has reset lower across all four names at Thursday’s open. The premium environment is compressed. The structures that make sense now are different from anything that made sense 48 hours ago. Here is a name-by-name framework.

META — Capex Overhang, Fundamental Business Intact

• If you are bullish: The ad business is demonstrably strong and $630 is a level that has served as meaningful support in prior pullbacks. A bull put spread — selling the May/June $600 put and buying the $575 put — collects premium in the compressed IV environment while defining downside risk. The risk-reward works if you believe the capex story is a short-term headline event and not a structural problem.
• If you are bearish: The capex raise from $115–$135B to $125–$145B is not a one-quarter event. It reflects input cost inflation that is structural as long as the Iran conflict persists. A bear call spread — selling the $640 call, buying the $665 call — gives defined-risk downside expression without the unlimited loss exposure of a naked short position.
• If you have no view: An iron condor centered around the $590–$650 range with wings at $570 and $670 captures the post-earnings low-IV environment through theta decay. The compressed volatility means you are collecting less premium than you would have pre-earnings, but the probability of staying in the range is also higher with the event risk now cleared.
• Watch closely: Reality Labs losses, any additional capex color, and the youth safety litigation that the company itself flagged as material in disclosure language.

GOOGL — Cleanest Fundamental Story in the Group

• If you are bullish: The 63% cloud growth rate and enterprise AI commentary give fundamental bulls the cleanest story in the Mag 7 group tonight. A bull call spread — buying the at-the-money call and selling a call 5–7% higher — limits premium outlay in the compressed IV environment. Something like buying the $175 call and selling the $185 call in a near-term expiry captures the next move without the full premium exposure of an outright long call.
• If you are bearish: YouTube advertising missing estimates ($9.88B vs. $9.99B) combined with 2027 capex being telegraphed as significantly higher creates a forward margin story worth fading. A bear put spread — buying a put at current levels, selling a put 5% lower — on a 30–45 day expiry gives defined-risk exposure without fighting the positive after-hours momentum head-on.
• If you want premium income: A short put or cash-secured put 5–8% below current price takes advantage of compressed IV while establishing a price level where you would be comfortable owning the stock. With the clean beat behind it, GOOGL is one of the easier post-earnings put-selling candidates in this group.
• Watch closely: Waymo commercial expansion data and Other Bets trajectory. If the YouTube ad mix continues shifting toward subscriptions, the revenue quality argument changes even if total YouTube revenue holds.

AMZN — Best Number of the Night, Worst Cash Flow

• If you are bullish: AWS at 28% growth is the most important data point from the entire evening. A bull call spread in a 45–60 day expiry — rather than a near-term structure — gives the trade room to develop as the market digests both the AWS acceleration and the Q2 guidance of $194–$199 billion in revenue. The FCF collapse is real but documented. The street understands it.
• If you are bearish: The FCF story is worth taking seriously. Free cash flow down 95% year-over-year on a trailing twelve-month basis, with capex guiding to $200 billion for the year, means Amazon is essentially pre-funding several years of growth with current cash. If AWS growth slows even modestly in Q2 or Q3, the multiple gets questioned fast. A bear put spread targeting a 5–8% pullback from current levels gives defined-risk downside exposure without going naked short into positive momentum.
• If you want neutral income: A short strangle in a 30-day expiry with strikes 6–8% above and below current price captures the post-earnings IV compression. The implied move was priced at 5–7% pre-earnings. With that event cleared, the probability of staying within those bounds increases significantly.
• Watch closely: H2 2026 AWS guidance tone and any additional comments on Leo satellite commercial timeline. Cost overruns on the satellite build add to a capex story that is already stretched to the limit.

MSFT — Technically Strong, Structurally Unclear

• If you are bullish: The YTD underperformance — down 12% into the print — means the stock is less crowded than its peers and potentially pricing in more risk than the Azure numbers justify. A call debit spread — buying the at-the-money call, selling a call 5% higher — gives asymmetric upside with limited premium at risk. This is the least consensus bullish trade in the group, which is sometimes the most interesting one.
• If you are bearish: The OpenAI exclusivity loss is not a quarterly noise event. It is a structural change in the competitive landscape of enterprise AI distribution. For traders who believe that moat was priced into the multiple and is now gone, a bear put spread on a 30-day expiry captures downside without unlimited risk.
• If you want to wait: That is also a reasonable position. The mixed after-hours reaction and ongoing OpenAI structure questions suggest MSFT consolidates near current levels until there is a cleaner catalyst. An iron condor in a 30-day expiry around current price benefits from both the low-IV environment and the likely range-bound behavior while the market waits for Q4 fiscal year data.
• Watch closely: Copilot enterprise penetration data. If the next quarter does not show material Copilot revenue contribution, the spending argument becomes significantly harder to defend at 22x forward earnings.

Real Risks That Are Not Theoretical

• Component cost inflation: Meta named the Iran conflict as a direct driver of its capex raise. This is not a boilerplate macro risk disclosure. It is a specific cost mechanism affecting all four companies. If Brent crude stays above $100 and supply chain disruptions persist, capex cost inflation becomes a structural margin headwind, not a one-quarter event.
• Free cash flow compression across the group: Amazon’s 95% FCF decline is the most extreme example, but the dynamic affects the entire group as capex doubles while earnings growth runs at 20–30%. If revenue growth decelerates even slightly while spending stays elevated, the multiple math breaks down quickly.
• AI monetization patience is finite: Tonight’s prints bought time — Alphabet’s especially. But the market’s willingness to fund AI infrastructure spending based on future revenue promises is not unconditional. When that patience expires is unknowable. That it is finite is not.
• Legal and regulatory exposure: Meta’s youth safety litigation and European Commission DSA findings on Instagram and Facebook are material risks that keep appearing in disclosure language with more specificity each quarter. Hard to model, increasingly hard to ignore.
• Apple reports April 30: Tim Cook’s September departure is already announced. John Ternus takes over. Tariff exposure on iPhone supply chains and AI feature adoption rates are the key numbers. The Mag 7 earnings week is not finished.

Thursday Morning: Eight Things to Do Before You Put on a Position

• Check IV levels on all four names before building any structure. Post-earnings crush will have already compressed premiums significantly. Do not pay pre-earnings prices for post-earnings risk.
• Assess each stock’s after-hours move against the 5–7% implied move range that was priced in. META exceeded that range to the downside. Residual positioning pressure may create continued selling at Thursday’s open as call holders who were positioned above $700 exit.
• For META: the $630 level is the new near-term reference. Watch whether the stock stabilizes there on volume or continues lower on light volume. That distinction matters for whether you are looking at a tradable support or a falling knife.
• For GOOGL and AMZN: both reacted positively to clean beats. The post-earnings IV crush creates a window to sell put spreads against support levels for premium income in a now-low-volatility environment.
• For MSFT: the mixed reaction and OpenAI structure overhang suggest underperformance versus peers near term. Hold off on directional long premium until Copilot data provides a cleaner fundamental anchor.
• Check Nvidia, AVGO, and AMD at the open. The hyperscaler capex confirmation is broadly constructive for the AI supply chain. Tuesday’s selloff on the OpenAI revenue miss report looks increasingly like an overreaction given what Alphabet and Amazon just printed.
• Apple reports tonight. The Mag 7 week is not over. Do not add risk to a portfolio that is already carrying Mag 7 event exposure without sizing down elsewhere first.
• Defined-risk structures only. VIX near 18–19, active geopolitical conflict, and an AI capital cycle that is accelerating faster than most models captured — this is not an environment for unlimited-risk positions regardless of how strong your fundamental conviction is.

Here is where I land on all of this. The AI capital cycle is not slowing. What changed last night is that the market started paying closer attention to the cost side of the ledger — not just the revenue line. Alphabet got rewarded because its cloud growth is now being driven by enterprise AI revenue, which is the bridge the bull case has been waiting for. Amazon got rewarded because AWS acceleration was undeniable. Meta got punished because its capex trajectory raised questions the revenue line did not fully answer. Microsoft got a muted reaction because the structural picture is genuinely unclear.

Four companies. Four different positions in the same race. The ones where AI spending is visibly converting into cloud revenue are getting the benefit of the doubt. The ones where the spending is ahead of the visible payoff are getting scrutiny they have not faced before.

Apple tomorrow. Nvidia in May. The season is not done, and neither is the repricing of the AI spending thesis.

This editorial is for informational and educational purposes only and does not constitute financial advice. Options trading involves substantial risk and is not suitable for all investors. All trade structures described are hypothetical frameworks for analysis. Past performance does not guarantee future results. Always conduct your own due diligence.