For three years, the Magnificent Seven moved like one thing. You bought the basket, you got the ride. Up together, down together, priced like a monolith.

That trade is over.

In Q1 2026, the group collectively fell 10.5% while the S&P 500 dropped just 4.6%. The S&P 493 — everything else in the index — was actually up 4%. For the first time since 2022, the Mag Seven is dragging the index down instead of carrying it.

But the real story isn’t the group average. It’s the spread inside the group. One name is up on the year. One has shed nearly a quarter of its market cap. The distance between them is where the opportunity lives — if you’re willing to do the work of telling them apart.

Scoreboard: Where Each Name Stands as of April 10

Here’s the 2026 YTD leaderboard, best to worst:

• Alphabet (GOOGL): up approximately +1.5% YTD — the only Mag Seven name in positive territory
• Apple (AAPL): roughly flat to slightly positive — holding up better than anyone expected given the AI criticism
• Amazon (AMZN): down approximately -8% YTD — weighed down by a $200 billion capex announcement and tariff exposure in its retail business
• Meta (META): negative YTD after paring earlier gains — operationally strong, but the market is nervous about the infrastructure bill coming due
• Nvidia (NVDA): negative YTD — the AI arms-race supplier is giving back gains as investors ask who’s actually getting ROI on all that spend
• Tesla (TSLA): down approximately -20% YTD — and down even more from the January peak
• Microsoft (MSFT): down approximately -23% to -24% YTD — market cap fell from $3 trillion to roughly $2.6 trillion. That’s roughly $400 billion in market value gone in one quarter.

That is not “tech is having a rough stretch.” That is stock picking disguised as passive investing. If you own a standard S&P 500 fund, 32.5% of your money is in these seven names. They have collectively cost you four percentage points of relative performance against the other 493 companies in the index — so far this year.

Why This Is Happening

The bundle trade worked because of three things that no longer all apply at once:

• Capital was essentially free — which made long-duration growth stories easy to own without demanding near-term proof
• Narrative beat numbers — AI capex was seen as a feature, not a cost
• Liquidity compressed correlations — owning the biggest, most liquid names was a risk-management decision, not a stock-picking one

Now? Rates are still above 4%. Tariffs are creating real margin uncertainty across every business with hardware in its supply chain. A US-Iran ceasefire was only announced April 8 — a fragile two-week truce — after weeks of energy-price pressure that was already raising recession probability estimates.

And the market has flipped its central question from “is this company building the future?” to “is this company paying for its future in cash flow, today?”

That is a very different scorecard. Consensus earnings growth estimates for the Mag Seven still sit at roughly 18% for 2026 — nearly double the 11% expected for the rest of the S&P 500. The earnings story hasn’t collapsed. The valuation story is the problem.

What These Businesses Actually Do (Plain English)

Before the numbers, a quick grounding on how each company actually makes money — because the divergence starts here.

• Meta (META): sells your attention to advertisers. About 96% of revenue is ads. AI makes that attention more valuable — if it shows up in conversion rates and CPMs, not just press releases.
• Alphabet (GOOGL): sells intent (Search) and attention (YouTube), with Google Cloud as the emerging profit engine. Ad revenue is tightly correlated with global economic confidence — which is exactly why the Iran ceasefire gave the stock a 4% single-session pop on April 8.
• Amazon (AMZN): runs a massive retail-and-logistics operation that cross-subsidizes two higher-margin businesses: AWS and advertising. The retail side is taking tariff fire right now.
• Microsoft (MSFT): sells enterprise software subscriptions and is trying to layer Copilot (AI) on top as an upsell. Azure is the growth engine — but it decelerated last quarter, and the market is penalizing capex that hasn’t produced the margin expansion that was underwritten.
• Apple (AAPL): sells premium hardware and takes a toll on every transaction that runs through its ecosystem (Services). When hardware stalls, it becomes a buyback-and-compounding story. Right now, paradoxically, being cautious on AI capex is working in its favor.
• Tesla (TSLA): sells cars — a cyclical, increasingly competitive business — while the market prices in software, robotaxi, and energy optionality that hasn’t arrived on schedule. The income statement is starting to matter more than the roadmap.
• Nvidia (NVDA): sells the picks and shovels for the AI gold rush. Still growing fast. But the stock has given back a chunk of its gains as investors start asking: what happens when the capex cycle peaks?

The Numbers: Company by Company

Meta: The Margin Machine Is Back — For Now

• Q4 2025 ad revenue growth: +19% year-over-year
• Operating margin: back above 40% — the clearest sign yet that the “Year of Efficiency” stuck
• Reality Labs: losses narrowing quarter over quarter, though still burning cash
• New front: Meta launched a standalone AI assistant app (Llama 4) in April, competing head-on with OpenAI and Google — the market is watching whether this becomes a revenue line or just a distribution play
• The tension: Meta is guiding toward heavy capex in 2026. The market punished that once already (Q3 2025). The question is whether 19% ad growth can justify the infrastructure bill — and whether that growth holds as comps get harder.

Alphabet: The One Name With Room to Breathe

• Google Cloud Q4 2025 growth: +28% year-over-year, crossing a $12B+ quarterly revenue run rate
• YouTube ads Q4 2025: +14% year-over-year
• 2026 capex forecast: analysts expect $115B+ — significant, but Alphabet has a better track record than most of turning capex into cloud revenue
• Custom chip edge: trained its Gemini model family on its own Ironwood TPUs — that’s a real cost advantage vs. relying entirely on Nvidia
• Next catalyst: Q1 2026 earnings due late April — the first read on whether Cloud keeps accelerating

Microsoft: The Surprise Laggard

Nobody had Microsoft as the worst performer in the group. Here’s how it happened:

• Q2 FY2026 revenue: $81.3 billion, up 17% year-over-year — fine in isolation
• Azure: growth decelerated toward 30% when the market expected it to hold at 40% — supply-constrained data centers are the bottleneck
• Market cap hit: from $3 trillion to roughly $2.6 trillion — down approximately 23-24% YTD
• Capex trajectory: analysts project nearly 57% capex growth in 2026, to over $110B. Goldman models it at $125B this year, rising to $144B in 2027.
• Copilot: still no clear disclosure on paid seats or ARPU lift — the market is in “prove it” mode and Microsoft hasn’t proven it yet
• The parallel: this is the same dynamic that crushed Meta’s stock in 2022. Spend first, earn later. Meta eventually delivered. The question is whether Microsoft does too — and on what timeline.

Amazon: A $200 Billion Commitment the Market Didn’t Love

• YTD: down approximately -8%, and roughly 17% off its 52-week high
• The announcement: $200 billion in 2026 capex — disclosed on the Q4 2025 call. Stock fell 5.5%+ the same day.
• Q1 2026 operating income guidance: $16.5B to $21.5B — came in below the Street’s $22.2B consensus midpoint
• FCF watch: trailing 12-month free cash flow has been declining quarter over quarter; more capex means more compression ahead
• The binary on April 29: AWS Q1 earnings. Above 20% growth keeps the margin thesis alive. Below 20% forces a full valuation rethink.
• The bright spot: advertising grew 24% in Q4 — an acceleration, and a higher-margin business that partially offsets tariff drag in retail

Tesla: 50,000 Cars Sitting in a Lot

The delivery number gets the headlines. The production-vs-delivery gap is the real tell.

• Q1 2026 deliveries: 358,023 vehicles — missed Wall Street’s 365,645 consensus by about 7,600 units
• Production: 408,386 vehicles built. Only 358,023 delivered. That’s a 50,363-unit inventory build, almost entirely Model 3/Y. Tesla has historically built to order. This is not normal.
• Sequential decline: deliveries fell 14.4% from Q4 2025’s 418,227 — steeper than seasonal norms
• The easy YoY comp: up 6.3% from Q1 2025’s 336,681 — but Q1 2025 was Tesla’s worst quarter in years due to a Model Y production halt. This is not a rebound. It’s a recovery from a self-inflicted low.
• Energy storage — the narrative that was supposed to hold: deployed 8.8 GWh in Q1 2026 — a 38% drop from Q4 2025’s record 14.2 GWh, vs. analyst consensus of 14.4 GWh. The one segment bulls were leaning on missed badly.
• Europe: Tesla registrations down 17% in January while the European EV market grew 14%. BYD outsold Tesla there for two consecutive months.
• Full-year 2026 math: consensus calls for 1.69 million deliveries — just 3.3% above 2025. The Q1 annualized run rate is about 1.43 million. Tesla needs a significant second-half acceleration just to hit a modest target.
• Next hard data: Q1 financials due April 22. Automotive gross margin and autonomy commentary are the two variables that matter most.

Apple: Winning by Not Losing

Nobody called this one. Apple was supposed to be the AI laggard. In 2026, being cautious on capex turns out to be a feature.

• YTD: roughly flat to slightly positive — the best relative performer in the group after Alphabet
• Fiscal Q1 2026: iPhone revenue +23% year-over-year, Greater China revenue +38% — a genuinely strong quarter
• Services: up 14% year-over-year to a record high; installed base now exceeds 2.5 billion devices
• Why it’s working: in a market penalizing companies for announcing $100B+ capex programs, Apple’s measured AI approach looks like discipline, not weakness
• Forward P/E: approximately 32x — not cheap in absolute terms, but the market is paying for perceived stability
• The risk that flips the story: tariff exposure on hardware. The India manufacturing pivot is multi-year. If tariff costs start landing on iPhone pricing, the defensive narrative cracks fast.

Is It Cheap? The Valuation Reckoning

“Cheap” doesn’t mean a low P/E. It means price paid versus reality being delivered. Here’s the honest framing:

• Alphabet at ~19x forward earnings: Cloud growing 28%, stock barely up on the year. The market is pricing in Search cannibalization as a certainty. That may be too pessimistic.
• Amazon at ~26-29x forward earnings: near a two-decade valuation low relative to its own history. Wells Fargo calls it a top internet pick at 21x 2027 GAAP earnings. 92% of analysts rate it a buy. The median price target implies roughly 33% upside. The capex overhang is real — but so is the discount.
• Meta at ~17-22x forward earnings: 40%+ operating margins. Some analysts see it at roughly 17x 2026 estimates right now — arguably the cheapest it’s been vs. its growth rate in two years.
• Microsoft at ~24% off its highs: still not “statistically cheap.” You need Azure to re-accelerate and Copilot to monetize in the same window. Two variables. One bet.
• Apple at ~32x on mid-single-digit growth: a quality premium. Fine as a hedge. Expensive as a growth position.
• Tesla at a significant premium on forward earnings: while carrying a 50,000-unit inventory build. The multiple has always assumed the autonomous/energy/software story arrives on schedule. It hasn’t yet.

Here’s the structural point worth anchoring to: the Mag Seven is expected to grow earnings at roughly 18% in 2026 — nearly double the 11% forecast for the rest of the S&P 500. The earnings case hasn’t collapsed. The valuation-vs-capex tension is what’s creating the dispersion. And dispersion is where mispricings live.

Bull / Base / Bear

Bull case — divergence pays: earnings season (kicking off mid-to-late April) delivers beats from Alphabet and Amazon. AWS holds above 20% growth. Meta re-accelerates. Microsoft telegraphs Azure capacity relief in H2. Tariff fears continue to fade after the Iran ceasefire holds. Active managers keep rotating out of laggards and into cash-flow compounders.

Base case — messy but directional: earnings are uneven — some beats, some guidance cuts. Sticky inflation (CPI for 2026 revised up to 3-4%) keeps the Fed parked longer than hoped. But names with the cleanest margin profiles and upward revisions still outperform over a 12-month horizon. Stock picking beats the basket.

Bear case — the macro resets everything: some institutional models are pricing in a 50% recession probability by year-end. If that lands, multiples compress across all seven regardless of fundamentals. In that tape, position sizing and cash discipline matter more than any internal rotation trade.

Action Plan: What to Do With This

If you’re still running an equal-weight Magnificent Seven basket out of habit, that’s a portfolio construction choice — not an investment thesis. Here’s how to think about fixing it:

• Step 1 — Sort by capex bucket: Microsoft and Amazon are being punished for spend that hasn’t hit margins yet. Meta and Alphabet have better track records of converting infrastructure spend into revenue. Know which bucket your names sit in.
• Step 2 — Mark your calendar for April 22 and April 29: Tesla’s Q1 financials (margin and autonomy commentary) and Amazon’s Q1 (AWS growth rate) are the two most important near-term data points in this entire trade. Both are binary reads.
• Step 3 — Follow revisions, not vibes: rotate toward names where the next 2-3 quarters have the highest probability of upward estimate changes. Alphabet is the strongest current case. Amazon is the conditional one.
• Step 4 — Scale in like a cheapskate: 2-3 tranches, spread over several weeks. Don’t chase a name after a strong session. The 4% single-day pop Alphabet got on the Iran ceasefire news April 8 is exactly the kind of entry point that looks great in hindsight and feels terrible to buy into.

Cheap Investor Scorecard: 10 Things to Track

• Alphabet Cloud growth above 25% YoY in Q1 2026 report (due late April) — passing today
• Alphabet Search margin stability despite AI overview rollout — watch TAC trends
• Amazon AWS growth above 20% on April 29 — the binary that matters most right now
• Amazon FCF trough visibility — is free cash flow bottoming, or still falling under the $200B capex cycle?
• Meta operating margin above 40% with ad revenue in high-teens growth — passing today, watch the comps
• Microsoft Azure re-acceleration signal in Q3 FY2026 guidance — any capacity relief language is the add trigger
• Microsoft Copilot commercialization — still no clear paid-seat or ARPU disclosure as of April 10
• Tesla automotive gross margin above 15% on April 22 — below that, the multiple becomes very hard to defend
• Tesla inventory build resolved in Q2 — if the 50,000-unit overhang persists, that’s a structural demand signal, not a timing one
• S&P 493 vs. Mag Seven spread — if the rest of the market keeps outperforming on a rolling basis, the concentration unwind has further to run

Bottom Line

Here’s where things stand: one Mag Seven name is up on the year. One has shed nearly $400 billion in market cap. One built 50,000 cars it couldn’t move. One just committed to spending $200 billion. And the company everyone was criticizing for being slow on AI is quietly outperforming all of them.

If you still think of the Magnificent Seven as one trade, you’re not getting diversification. You’re getting seven businesses — at seven different points in their cycles, carrying seven different capex loads — all priced at different levels of optimism about a future that may or may not arrive on schedule.

The bundle is broken. For anyone willing to do the work of separating the compounders from the story stocks, that’s not a risk. That’s the setup.

— The Cheap Investor Editorial Engine