Stocks To Buy Now: July 2026

THE MARKET SETUP

The market is ripping the bandaid off two fears at once.

Fear one: Iran. The truce fractured this week. Tehran hit tankers near the Strait of Hormuz, the US reinstated the blockade, and Trump slapped a 20% toll on all cargo through the strait. Oil jumped over 9% in a day at one point, its biggest spike since 2020. The Fear & Greed index is sitting in the Fear zone. Consumer sentiment just printed 54, its best reading since the war started in February, but still down 12% from a year ago.

Fear two: AI capex. Hyperscalers are spending roughly $725 billion on AI infrastructure in 2026,

headed toward $1 trillion in 2027. The market cannot decide if that is the greatest investment cycle in history or the greatest overbuild.

Here is why that combination is your opportunity. Fear compresses prices, not businesses. Microsoft is down 20% this year while Azure accelerates. ServiceNow is down 47% in a year while growing 22%. Vistra and Constellation sit near 52-week lows while signing 20-year contracts with the richest companies on Earth. Celestica and Corning both got smoked 30% in a month while raising guidance.

THE HYPERSCALERS

1. Meta Platforms Inc (META)

The attention toll booth. Facebook, Instagram, WhatsApp. Four billion people, 81% gross margins.

Numbers: $215B revenue, $70B profit, 33% margin. Forward PE ~20, below its own 10-year

average. Down 9% over the past year while earnings grew.

Why: AI capex here pays off immediately. Better AI means better ad targeting means more revenue this quarter, not in 2030. You get the discount because the market fears the $125 to 145B they are spending this year.

Catalyst: Q2 earnings July 29. Any sign capex converts to revenue growth and the multiple re-rates.

Risk: If the spend has not shown up in profits by 2027, the market punishes it hard.

Zone: $550–$600. Steal: under $560.

2. Amazon Inc (AMZN)

Three businesses in a trench coat. The store you see, plus AWS and advertising where the profit

lives, plus a chip business nobody prices in.

Numbers: AWS growing 28%, its fastest in almost four years, with a $364 billion backlog. In-house chips passed a $20B run rate. Forward PE ~25.

Why: The backlog is contracted future revenue. Their Trainium chips deliver the same AI work at 30 to 40% less cost than Nvidia, and they are exploring selling them to outsiders.

Catalyst: Q2 earnings July 30. AWS acceleration plus any capex ROI proof reprices the stock. Street targets cluster at $300 to $325.

Risk: $200B of capex crushed free cash flow to nearly zero. If AWS slows while spend stays

elevated, it gets ugly.

Zone: $200–$235. Steal: under $210.

3. Alphabet Inc (GOOG)

The only company that owns the full AI stack. Models, chips, data centers, and distribution to billions through Search, YouTube, and Android.

Numbers: $110B quarterly revenue, up 22%. Cloud growing 63%, fastest of any major, with a $462 billion backlog. Record 36% operating margins. Buffett owns $30B+ of it.

Why: Two years ago the fear was AI kills Google. Now Google is winning AI and selling TPU chips

directly against Nvidia.

Catalyst: Q2 earnings July 22, first of the giants to report. Cloud backlog growth and TPU

commentary are the numbers to watch.

Risk: Up 74% in a year. The stock ran ahead of the business. EU regulators circling. Least margin of safety of the four giants at this price.

Zone: $300–$335. Steal: under $290. Do not chase $370.

4. Microsoft Corp (MSFT)

The hated blue chip and my favorite setup on this list. Down 20% this year, its worst sentiment stretch since 2008, while the business accelerates.

Numbers: Azure growing 40% and speeding up. AI revenue $37B annualized, up 123%. $627 billion in already-contracted future revenue. Forward PE ~20, cheapest since 2023.

Why: The price says broken. The business says accelerating. That gap is the entire trade.

Consensus targets sit near $550, roughly 40% above today.

Catalyst: Earnings July 29. Management already guided Azure to accelerate in the back half of 2026.

When supply constraints ease, growth is unlocked, not created.

Risk: The AI-eats-software theory, plus $190B of capex squeezing margins.

Zone: $350–$375. Steal: under $350.

SOFTWARE

5. ServiceNow Inc (NOW)

The software that automates how big companies run: IT, HR, operations. Boring, essential, nearly

impossible to rip out.

Numbers: Stock down 47% in a year. Business growing 22%. Forward PE ~24 with a PEG near 1,

meaning you pay fair value for the growth and get the quality free. 43 analyst buys, 1 sell.

Why: The cleanest panic-versus-reality gap in software. Renewal rates near 98%. Customers stayed. Only the stock price left.

Catalyst: Q2 earnings July 22. Four straight beats. A fifth beat plus raised guidance forces the

re-rate.

Risk: If AI agents genuinely replace enterprise software, this is the stock that theory is about. No

cloud business to hide behind.

Zone: $90–$115. Steal: under $90, generational.

Salesforce Inc (CRM) is the deep value alternate: ~$150, down 33% YTD, ~14x forward, $50B

buyback, Agentforce growing 205%. But organic growth slowed to 10%. I lean NOW.

CHIPS & AI INFRASTRUCTURE

Warning: chips are wobbly after the run-up and they fall together when capex fear or Iran

headlines spike. Those dips are the entries. Size accordingly.

6. NVIDIA Corp (NVDA)

The tollbooth of AI. Biggest company on Earth, still growing like a startup.

Numbers: $82B revenue last quarter, up 85%. $49B free cash flow in one quarter. $80B buyback.

Forward PE just ~23 against a three-year average near 60.

Why: An 85% grower priced like a mature company, purely because of capex fear. That is the

definition of my example: reasonable valuation for its forward growth and positioning.

Catalyst: Vera Rubin ships this fall, 10x performance per watt, and Jensen says it will be

supply-constrained its entire life. Hyperscaler capex guides keep rising.

Risk: First and hardest hit when AI spending growth merely slows. Big customers all building their own chips. China is a regulatory wildcard.

Zone: $175–$200. Steal: under $170, appears only in full panics, lasts days (Deepseek crash for example).

7. Broadcom Inc (AVGO)

When Google, Meta, or OpenAI want their own custom AI chip instead of Nvidia's, Broadcom designs it. Plus the networking silicon connecting everything.

Numbers: AI revenue $10.8B last quarter, up 143%, guided to $16B next quarter, up over 200%.

Apple deal locked through 2031. The catch: ~40x forward against a 10-year average of 17x.

Why: Every hyperscaler wants a second source to Nvidia. Broadcom is the second source.

Catalyst: Each new custom chip customer announcement has moved the stock double digits. More are coming.

Risk: Up 180% since April. A handful of customers means one cancelled program takes 20% off in a day.

Zone: $280–$320. Steal: under $270. At $380 you are momentum trading, not accumulating.

8. ASML Holding NV (ASML)

One company on Earth makes the machines that print advanced chips. This one. Not the leader. The only one.

Numbers: Just reported July 15: beat everything, raised full-year guidance to €43 to 45B. EUV sales growing 45%, memory revenue growing 75%, roughly two years of backlog.

Why: You do not have to pick the winning chip. Nvidia, Broadcom, Google, Amazon: all of them are printed on ASML machines. Picks and shovels play.

Catalyst: High-NA, the $400M next-gen machine, is now in production at Intel. Memory upcycle just ignited.

Risk: The run to $1,800 pre-paid much of 2027. Cyclical: equipment orders can vanish for a year.

China export controls are permanent overhang.

Zone: $1,300–$1,500. Steal: under $1,300. Only ever buy the 25% drawdowns, never sell.

9. Vertiv Holdings Co (VRT)

Industrial power and liquid cooling for data centers. Without it, the chips literally cook themselves.

Numbers: Revenue up 30%, profits up 83%, Q4 orders up 252%, backlog $15B. Guiding 34%

growth. The catch: 45 to 49x forward.

Why: Orders and margins inflecting simultaneously. That is when operating leverage prints money. Co-developing next-gen cooling with Nvidia.

Catalyst: Earnings July 29. Every new gigawatt-scale data center announcement is a Vertiv order.

Risk: Priciest infrastructure name here. Earnings must grow into the valuation. Dropped 25% in June alone.

Zone: $170–$200. Steal: under $160. The June-style air pockets are the entries.

10. Arista Networks Inc (ANET)

The nervous system. Thousands of AI chips are useless unless they talk to each other at insane

speed. Arista makes those switches for Meta, Microsoft, and Oracle.

Numbers: Revenue up 35%. Full-year guide raised to $11.5B with $3.5B from AI networking, a target already doubled once. Operating margin 48%. Zero debt.

Why: As clusters scale, the network becomes the bottleneck, so Arista's product gets more critical every generation.

Catalyst: New 1.6-terabit switch ramping into Meta and Microsoft now. Management expects one or two brand-new customers worth 10%+ of revenue this year. Earnings August 4.

Risk: ~48x earnings with heavy customer concentration. Memory shortages squeezing supply. Nvidia sells competing networking.

Zone: $125–$140. Steal: under $120.

11. Celestica Inc (CLS)

The Canadian workhorse. Hyperscalers design custom AI hardware, Celestica builds it: the servers, the switches, and the racks.

Numbers: Revenue up 53% last quarter. Raised 2026 guidance to $19B revenue and $10.15 EPS, then put a floor under 2027 at $25.5B minimum. And here is your setup: down 29% in the past month, from $413 to $336, sitting at the lower band while guidance went up.

Why: This is the definition of a steal window. The business accelerated, the stock got sold with the sector. At ~33x this year's earnings and ~25x next year's floor, you are paying a great price for contracted, non-cancellable orders.

Catalyst: Earnings July 27. Eight straight EPS beats. Ten active 1.6-terabit switch programs booking awards into 2028. Goldman target: $475.

Risk: 65% of revenue comes from three customers. Thin manufacturing margins. Highest volatility on this list outside SoFi.

Zone: $275–$340. Steal: under $280.

12. Corning Inc (GLW)

Hundred-and-seventy-year-old glass company that quietly became an AI play. Every data center needs miles of optical fiber to connect the chips. Corning is the world's fiber architect.

Numbers: Optical sales up 36% last quarter. Targeting a $20B annual run rate by year-end. Signed a multibillion-dollar fiber deal with Amazon in June, and a partnership with Nvidia that came with Nvidia taking warrants on 18 million Corning shares. Down roughly 30% from its recent high in the optical selloff.

Why: Fiber demand scales with every GPU shipped, and Corning pays you a dividend while you wait. Nvidia taking equity exposure tells you who they think wins the connectivity layer.

Catalyst: Q2 earnings late July. Amazon deal revenue starts flowing, plus government broadband programs layering on top.

Risk: Even after the 30% haircut it trades near 50x this year's earnings because legacy glass and

display businesses are stagnant. The AI story carries the whole multiple.

Zone: $140–$160. Steal: under $135. This is a pullback-only buy, and the pullback is happening

now.

FINANCE

13. American Express Co (AXP)

A subscription business wearing a credit card costume. Card fees are recurring revenue attached to the highest-quality spenders in the world.

Numbers: Fee revenue up 16%, the 30th straight quarter of double-digit fee growth. Gen Z spending up 38%, so the customer base is getting younger. Dividend hiked 16%. ~20x forward.

Why: This is the dividend and value slot with hidden growth from the subscription model. Premium customers do not default in downturns. Proven pricing power: they raised Platinum fees and spending accelerated.

Catalyst: Earnings July 24 with a likely guidance raise. JPMorgan just upgraded it with a $400 target.

Risk: Consumer spending proxy. A real recession hits billings. It lags in melt-ups. This is ballast, not a rocket.

Zone: $310–$345. Steal: under $300.

14. SoFi Technologies Inc (SOFI)

The asymmetric bet. A full digital bank for the phone generation: loans, checking, investing, crypto, one app, ~13 million members.

Numbers: Revenue growing 30 to 40%. Full-year guide: $4.66B revenue, adjusted EPS $0.60, up

51%. Profitable, compounding members 30%+, JPMorgan just upgraded it, and it gets floated as a takeover target. The catch: ~30x forward, trades like a 2x beta tech stock.

Why: Highest ceiling on this list. A bank compounding customers 30% with margins inflecting is how multi-baggers happen. The market still prices it as a fintech experiment.

Catalyst: Earnings July 29. Crypto and HELOC products just launched, and each cross-sell raises revenue per member.

Risk: Everything. Charge-offs creeping up. It fell 14% on its last earnings despite beating.

Growth-stage lenders get hit hardest in recessions. Size it small.

Zone: $13–$16. Steal: under $13.

VALUE

15. Berkshire Hathaway Inc (BRK.B)

$397 billion of cash. Insurance, railroads, energy, a $300B stock portfolio, now run by Greg Abel

with the largest cash pile in market history.

Numbers: $397 billion in cash, over a third of the entire company's value. Operating profits up 18%. Net seller of stocks for 12+ straight quarters.

Why: The only stock here that gets MORE valuable when everything crashes. That cash deploys at panic prices. It is the shock absorber that lets you stay aggressive everywhere else. With Iran

headlines flying, that insurance matters.

Catalyst: Abel is finally spending: a homebuilder acquisition, $10B more Alphabet, buybacks

resumed.

Risk: Lags a ripping bull market. Post-Buffett era untested.

Zone: $440–$480. Steal: under $430. Current price: hold, do not chase.

ENERGY

This is the AI trade the market refuses to price as the AI trade. With Iran pushing oil higher, power producers get a second tailwind: higher power prices lift their unhedged output.

16. Constellation Energy Corp (CEG)

America's largest nuclear fleet. Data centers need clean power 24 hours a day. Solar sleeps. Nuclear does not.

Numbers: Revenue up 64%, earnings up 28%, beat. Full-year EPS guide $11 to 12, targeting ~20% annual growth toward 2029. Near its 52-week low at ~22x forward.

Why: The catalyst machine of this list. Already inked: Microsoft restarting Three Mile Island, Meta, CyrusOne, and now Walmart's first-ever nuclear deal. These are 20-year contracts with investment-grade giants. The earnings do not care about quarterly AI panic. The stock got dumped with tech anyway.

Catalyst: Every new hyperscaler PPA is a press release that re-rates the stock. Crane restart on

track for 2027 after the June FERC waiver. Analysts sit 40% higher.

Risk: Merchant power price exposure, aging reactor outage risk, and any nuclear incident anywhere reprices the sector.

Zone: $225–$255. Steal: under $220.

17. Vistra Corp (VST)

Same thesis as Constellation, cheaper ticket. Texas power giant: nuclear, gas, solar, batteries, selling into the two fastest-growing grids in America.

Numbers: Revenue up 43%, swung from a loss to a $1B profit, beat EPS estimates by 53%. 2026

EPS estimated near $9.40, which makes this roughly 15x earnings: the cheapest AI-adjacent stock on this entire list. Down 39% from its high. Street average target: ~$225, more than 60% above today.

Why: Nearly all 2026 and 2027 output is already hedged, so the earnings are locked while the stock panicked. The January 20-year nuclear deals with Meta, plus a pending AWS deal, are not even in guidance yet. Pure unpriced upside.

Catalyst: Closing the Cogentrix acquisition and the AWS PPA, both of which force guidance raises. Earnings August 6.

Risk: More commodity exposure than CEG once hedges roll off. Heavy Texas concentration: weather and grid politics matter. Q4 miss started this slide. That is worth noting.

Zone: $125–$150. Steal: under $120. Best value entry in the energy bucket.

DEPLOYMENT

In the zone right now: Microsoft, ServiceNow, Nvidia, Constellation, Vistra, Celestica, Amazon. Start tranching today.

Alerts set, cash ready: Broadcom, ASML, Vertiv, Arista, Corning, Alphabet, Meta, Amex, SoFi,

Berkshire. The infrastructure names hand you a 20 to 30% dip every year. Iran headlines will

manufacture at least one more. No emotion, just execution.

Structure: Hyperscalers plus Nvidia are the core. ServiceNow, Vertiv, Arista, Celestica are the

growth satellites. Amex and Berkshire are the ballast. CEG and Vistra are the energy leg. SoFi is the small asymmetric swing. Corning is the income-paying infrastructure sleeper. Nobody gets hurt by a position they did not oversize.

Timing: the earnings gauntlet starts in five days. Alphabet and ServiceNow July 22. Amex July 24. Celestica July 27. Microsoft, Meta, Vertiv, SoFi around July 29 and 30. Arista August 4, Vistra August 6. Every zone in this video gets stress-tested within three weeks. That is not a reason to freeze. That is the reason you buy in tranches, never all at once.

Stay sharp.

Disclaimer: Not a financial advisor, not financial advice. My opinions and my own accumulation plans. Numbers are approximations from public data as of mid-July 2026 and will change. Do your own research and size for your own risk tolerance.