From Mag7 to Lag7: when the market asks for receipts

Breadth is the plot twist

A normal cap-weighted index gives the biggest companies the biggest voice. An equal-weight index gives every company the same vote.

When equal-weight leads by a wide margin, it usually means two things. More stocks participate in gains, and the giants do not do all the heavy lifting. In plain English, the parade gets bigger, but the biggest floats stop stealing the show.

That matters for investors because it changes what “the market is doing” actually means. If you only look at the headline index, you might think nothing much is happening. Under the surface, you can have plenty of winners, and a few heavyweights holding the index back.

It is also a reminder that concentration cuts both ways. When a small group leads for years, it can start to feel permanent. In 2026, the market starts to challenge that habit.

Why Mag7 starts to look like Lag7

The “Magnificent Seven” are Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, and Tesla. They matter because they are large enough to steer indices, sentiment, and expectations. This year, they also become a convenient place for investors to express doubt. Not doubt about whether artificial intelligence (AI) is real. Doubt about who captures the cash, and when.

A clean proxy makes the mood visible. The Roundhill Magnificent Seven exchange-traded fund (ETF), ticker MAGS, closed at 61.79 USD on 18 February 2026. Its 52-week high is 69.14 USD. That puts it down 7.35 USD, or 10.6%, from the peak.

That does not mean the companies are suddenly “bad”. It means the market’s grading system changes. In 2023 to 2025, being associated with AI often earned a premium by default. In 2026, the premium comes with homework: show that AI turns into revenue, and do it before the bill feels too heavy.

This is where the confusion kicks in. Investors seem to penalise software for disruption risk, and penalise big tech for spending risk, at the same time. Different reasons, same sector label.

The tech double penalty: disruption fear plus spending fatigue

Software faces the substitution fear. Generative AI can compress seats, automate tasks, and make cheaper substitutes “good enough”. In practice, that can mean slower user growth, more bundling pressure from big suites, and weaker pricing power.

Big platforms face the spending fatigue. They build the data centres, buy the chips, and fund the capex (capital expenditure) that makes AI usable at scale. The market can love that ambition and still worry about the payback timetable. When investors hear “we are spending more”, their next question is “and when does that show up in cash flow?”

So the market does something that looks irrational but is actually consistent. It demands receipts.

That is why our internal work on the software shortlist leans on observable signals, not predictions. The aim is not to call winners. It is to track whether AI is monetised, or quietly absorbed as a cost.

We use two layers.

First, five lenses to screen disruption risk: seat-heavy pricing exposure, bundle-away risk, “good enough” substitution, small and mid-sized business sensitivity, and AI agent disruption risk.

Second, a simple receipt checklist you can watch each quarter: pricing language shifts (per user versus per task), paid AI add-ons, net revenue retention (NRR), bundling pressure, sales efficiency, and small business churn.

This also helps frame “disrupted” versus “disrupters”. Disrupted tends to mean seat-heavy point tools in workflows an AI agent can complete end-to-end. Disrupters more often own distribution, the system of record, or trusted data, where AI can deepen the moat. Their risk flips: AI becomes expensive if monetisation lags.

Risks to keep in mind, and what to watch

This is a framework, not a forecast.

Breadth can reverse quickly if mega-caps regain momentum and pull the index back into a narrow leadership model. Watch whether the equal-weight advantage shrinks.

Capex fear can fade if companies show clear paid adoption and improving unit economics, meaning profit per customer rises as AI scales. Watch pricing and attach rates, not product demos.

Finally, “AI disruption” can become a lazy explanation for any sell-off. The antidote is boring, but effective: watch the receipts, quarter after quarter.

Investor playbook

• Treat “tech” as many business models, not one trade. Separate software seats from platform infrastructure spend.

• Follow receipts: paid AI add-on, NRR trends, and clearer pricing language matter more than AI mentions.

• Use breadth as a mood gauge: equal-weight leadership often signals rotation, not a broken market.

• Stay humble on timing: narratives move in days, business model shifts show up over quarters.

The takeaway for long-term investors

“Lag7” is a catchy label, but it points to a real shift. In 2026, the market does not abandon AI. It simply stops paying up for AI stories without proof. The equal-weight versus cap-weight gap shows that leadership spreads beyond the giants, even while the biggest names wobble.

At the same time, investors punish software for disruption risk and punish big platforms for spending risk, which feels contradictory until you see the common thread: the market wants receipts. The practical takeaway is not to predict. It is to measure what changes, and watch whether those changes get paid for.