Good morning. Futures are firm — the bid's real after the NVDA print. NVDA WRAPS $500B OF THIRD-PARTY COMPUTE CAPITAL with Apollo/BlackRock/Blackstone/Brookfield/GS/KKR. That kills the balance-sheet-overhang narrative cold. Memory stays the cleanest fundamental on the board: MU CAN'T FILL HALF OF DATACENTER DRAM DEMAND, HBM4 8Hi pulls forward, LTAs re-widen the HBM/DDR premium. SNDK rides the same wave. The software bogy sits with MNDY: -9% AH on a weak Q2 and guide — the VARification thesis in one print. MSFT keeps ripping, +31% in eight sessions yet still only +5.5% YTD, so the catch-up has runway. AAON +8% on the beat — datacenter cooling is the quiet compounder. Asia/macro: the AI supply chain grinds higher, but the wrinkles are visible — cracked gas turbines at Colossus, batteries dying in weeks, and the SEC just opened the securitization spigot for DC bonds. That's the new battleground: power reliability and financing structure, not GPU counts. VARA -44% in July says the leveraged crowd is still deleveraging; dips stay buyable for the unlevered. Four threads today: memory is the only clean supply/demand story left; AI infra becomes project finance — power, bonds, and IG tenants win; software gets VARified; MSFT is the large-cap re-rate with more room. We'll hit up MU and SNDK first, then get to MNDY, DDOG, and TEAM in software, before circling back to NVDA financing fallout in the power names.
GTA VI PREORDERS ARE "UNPRECEDENTED AND ASTONISHING" — THAT'S THE WHOLE BALLGAME. TTWO PRINTED A Q1 BEAT (ADJ EPS $0.35, +6% VS CONSENSUS; EX-IMPAIRMENT $0.54, +64%), HELD FULL-YEAR GUIDANCE, AND THE STREET DOESN'T CARE ABOUT THE LACK OF A RAISE BECAUSE THE CATALYST IS THE LAUNCH, NOT THE QUARTER. STOCK SITS AT $253.56, ONE BLINK FROM THE $265.94 HIGH, WITH TARGETS JUST CLUSTERED $270-313.
Net bookings $1.386B — 2.1% over consensus and above the $1.32-1.37B guide. The headline EPS number understates the beat, too: a $43.4M impairment on a cancelled third-party title buried ~$0.19 of drag. Back that out and the quarter was a blowout, not a beat.
Quality of the print matters more than the headline. NBA 2K and GTA carried. RCS +7% in 2K, +3% in GTA Online (the durability of that title this deep into the cycle keeps surprising). Mobile is the wart: -7% on a horrid Color Block Jam comparison. Management says no consumer pullback — and with what they're seeing on GTA VI preorders, we take them at their word.
Management dropped the "astonishing" preorder word and then refused to touch guidance. Textbook sandbag. UBS is reading it the same way:
"UBS expects guidance will prove conservative and anticipates more significant updates closer to the GTA VI launch."
The UBS model shows the shape of the story: FY26 bookings $6.7B → FY27 $8.5B → FY28 $9.2B. EPS $4.09 → $6.51 → $11.12. That FY28 margin step is the re-rating trigger. Buy the multi-year narrative, not the quarter.
Bull: Preorders are real demand, not survey noise. Management doesn't use "astonishing" casually. The 13.4% return over the past year says the stock hasn't fully re-rated for the launch cycle yet — room to run.
Bear: Preorders can cancel. Guidance didn't move. Mobile is shrinking. At $253 with the stock this close to the high, no cushion exists if GTA VI slips or launch demand gets pulled forward into a disappointing first-month comp.
Six firms stepped up this week with PTs parked in the $270-313 range. The collective read: the quarter validates franchise strength, and the preorder commentary validates the GTA VI narrative. Nobody chases the print — they position for the event. (BTIG the high at $313; Baird the low at $270.)
Bottom line: carry the position into the launch. Dips toward $250 are buyers. The r/r skews long until management finally raises guidance and the supercycle trade becomes consensus.
SEVEN SHOPS. SEVEN BUYS. That's the entire story this morning. BofA ($15), Goldman ($18), Barclays ($17), Evercore ($17), Stifel ($16), UBS ($14), William Blair (Outperform) — the Street just tripped over itself to open coverage on Standard Nuclear, and the tape's telling you everything: $8.44 at the BofA print at 11:41, $9.88 by the time Goldman hit the wire at 2:20. Up ~17% intraday on initiations alone.
The consensus thesis is annoyingly clean: STDN is the ONLY U.S. INDUSTRIAL-SCALE TRISO FUEL PRODUCER, and TRISO is the fuel that keeps microreactors and SMRs viable (higher burn-up, inherent safety). The model is reactor-agnostic — doesn't matter which OEM wins the reactor bake-off, they all need the fuel. Near-term demand is government-driven: UBS flags the military microreactor programs, with test reactors and demos filling the gap before commercial SMRs scale.
BofA's framework is the sharpest — four pillars: fuel availability as the gating constraint on reactor deployment, the reactor-agnostic model, near-term TRISO demand from microreactor/test/demo builds, and manufacturing expansion + customer qualification + regulatory progress. But they threw the key caveat in there too: TAM is "tied to TRISO-compatible reactor designs, representing a smaller but critical subset of advanced nuclear deployment." That's the honest frame for the bear.
Goldman's model is the punchline:
Standard Nuclear requires no further external capital to fund its growth, which makes it unique among peers tied to the growth of SMRs.
That's the real differentiator. Cash > debt, current ratio 19.25, capital-light with pricing power at the early stage of TRISO adoption. GS has revenue crossing $1B by 2030 (from sub-$20M in 2026), ~65% EBITDA margins by 2030, positive FCF in 2028. Their demand math: ~15GW cumulative SMR deployment by 2035 (from zero today) = ~100 MTU of TRISO demand.
Bull: It's a monopoly-ish position on a rarefied fuel that every advanced reactor design needs, backed by a balance sheet that never needs to dilute. Defense contracts give the next 12–18 months actual revenue visibility, not just deck math. Scarcity + optionality is a hell of a combo in a market starving for power exposure.
Bear: Trailing twelve-month revenue is $3.36 MILLION. The entire r/r rests on 2029 EBITDA multiples and an SMR curve that goes 0 to 15GW in ten years. BofA says competition is coming — fuel suppliers, vertically integrated reactor developers, incumbents — and the TRISO-compatible TAM is a subset of the broader nuclear story, not the whole thing.
For PMs playing the nuclear basket: STDN is the purest fuel-supply scarcity play on the board. The risk is paying 2029 prices for a 2026 story, but in this tape, the "only" in the US narrative can compound for a while. Next catalyst: name-brand reactor OEM contract announcements. That's what gets the next leg.
BMO isn't blinking. Reiterates Outperform and $186 PT on DuPont after a Q2 beat the tape mostly shrugged off. Stock sits at $140.86 — 32% to the target — and BMO still carries it as a top pick.
The quarter: adjusted EPS $1.88 vs. $1.76 est, revenue $1.82B vs. $1.81B est. Organic +4%, operating EBITDA +8%. Beat on both lines, and the stock sold off anyway. Investors chose the growth pace over the print. BMO says that's backwards.
"DuPont will likely see solid double-digit earnings per share growth for the foreseeable future, with risk to the upside relative to buyside and consensus expectations."
That's the whole thesis. PEG of 0.49 — market's pricing a decelerator, BMO sees an accelerator. Healthcare leads, high single digits in 2027. Water growth improving. Diversified industrials outperforming. And the 80/20 initiative is expanding margins without the revenue drag bears modeled.
Balance sheet gives BMO more room to run: current ratio 2.43, D/E 0.23. Buybacks and M&A are funded — optionality, not just a balance-sheet flex.
The bear case: the tape is telling you the easy money's been made (stock's up 64% in a year) and the outlook doesn't justify chasing. BMO's retort: growth reaccelerates from here, and you're paying a PEG under 0.5 for it. I know which side of that trade I'm on.
DA Davidson holds Buy, $100 PT after a quarter that beat by a mile and STILL got sold. CONSENSUS WANTED $1.11, MNDY PRINTED $1.48. Revenue $364.6M vs $355.59M est. Market's not stupid — the guide got a "reiterated but with near-term uncertainty" qualifier, and a 20% workforce cut says growth is getting more expensive to buy.
R/r at $88.63 vs $100 PT is fine, not exciting. The bull case is cheap + AI traction. PEG 0.26 on an 89% gross margin software business is objectively attractive, and the upmarket push is landing. DA Davidson sees the AI pivot as early but real.
"The company's strategic pivot toward AI products is showing early signs of success."
Down 62% YoY, up 26% off the lows — wounded but stabilizing. Restructuring overhang keeps the multiple in check. This is a hold-and-let-it-work name, not a chase-it-green name. Watch the next guide: "near-term uncertainty" could become a cut if AI sales don't offset the headcount reductions.
Bernstein's first on the tape with an Outperform and a $65 PT — that's ~36% upside from the $47.63 close. They're not calling a beaten-down name; PDFS is already up 151% in a year, 61% in six months. They're calling a compounder.
The thesis is structural semis complexity, not the last print. Q2 was a modest beat (EPS $0.27 vs $0.26, revenue $61.52M vs $60.96M) — and the stock actually dropped in the regular session before recovering after hours. Ignore that noise.
"Bernstein expects these factors to drive revenue and earnings per share growth exceeding 20% over the next several years."
That's the bet. The $65 PT is ~33x the average FY27/28 EPS estimate of $2.02. Mid-30s forward multiple — in line with EDA and semi-cap equipment peers, but Bernstein thinks PDFS deserves a premium given its positioning in yield improvement and process analytics. Makes sense: device complexity is the ultimate pricing power for this kind of software. No other analyst moves today — this one is the signal.
BUY MAINTAINED, PT SLASHED TO $43 FROM $55 — THIS IS A MULTIPLE PROBLEM, NOT A BUSINESS PROBLEM. Stock got cut to $31.92 (-10%) after the print, but the needle mover is sector-wide multiple compression, not the quarter. CEVA beat and raised: FY26 revenue growth now 13-14% (from ~12%), GM sitting at an absurd 87% (pure IP licensing economics). Q2 EPS $0.08 vs $0.07 est, revenue $29M +13% y/y, ahead of the $28.14M bogey. Then the stock dropped anyway. That tells you everything about positioning — the +78% y/y run left no room for a beat-and-raise that was already in the tape.
Needham trimmed 2026 numbers slightly, left 2027 mostly untouched. So the story is intact, they just want to pay less for it. Bull case: royalty inflection in H2 is driven by mix, not macro — IoT volume is holding up because low memory content insulates it from memory pricing, plus OEMs shifting volume from handsets to IoT. Bottom-tier handset royalties remain strong on the 4G→5G transition, and management has guided for more iOS in-house modem volume in H2. That last one is the real optionality — Apple in-housing modem content could pressure the royalty line, but here management is explicitly modeling growth from it. Worth watching.
"Royalty growth in the second half of the year likely stems from IoT volume being less affected by memory pricing due to low memory content and potential handset-to-IoT volume shifts by original equipment manufacturers."
Bear case: the market is pricing this fine. $43 PT is still ~35% upside, but the stock sold off on a beat-and-raise — that's not a great sign for near-term momentum. Semis/IP compressing across the board, and CEVA trades at a premium multiple for a business that only returns to profitability this year (consensus $0.56 EPS). The 87% GM is beautiful, but it's a royalty story at the mercy of handset volumes and OEM in-housing decisions. If H2 royalty acceleration disappoints, the multiple compression gets worse before it gets better.
Net: business and guidance both fine, but the risk/reward has shifted. $43 target is a multiple repricing, not a fundamental reset.
The take: Three PT hikes on one print — DA Davidson $34 (Neutral), RBC $36, Piper $39 — but bulls and the hold agree on mechanics: TOST is compounding the metrics that matter, and the debate is only about price.
Q2 beat where it counts. Non-GAAP FinTech & Subscription gross profit and adjusted EBITDA came in 4% and 13% above DA Davidson's model. Management raised 2026 midpoints on both by 2% — modest, but the direction is the signal. Revenue printed $1.91B, ~2% over consensus. The headline grower: RECORD 9,500 NET LIVE LOCATIONS ADDED in the quarter. That's the forward engine.
The warts: EPS of $0.26 missed the $0.32 bogey. That's why DA Davidson sits Neutral at $34 despite the beat — the multiple already captures a lot of the good news. Piper is the bull at $39 on margin strength; RBC at $36 cites the guidance revision. The Google Maps AI ordering integration is the optionality — ordering directly from the discovery surface is where the long-term r/r lives, though it's early innings.
Argus upgrades SNDK to Buy, PT $1,600 (stock at $1,212) — the $500 pullback from coverage launch is the entry. FQ4 print crushed: REVENUE $8.97B VS $8.42B CONSENSUS, EPS $39.25 VS $34.59 ESTIMATE, and FQ1 guide points to continued exceptional growth. Data center AI spend, client SSDs, enterprise solutions all accelerating. LTM revenue +175% keeps the multiple reasonable.
"The $500 pullback from its coverage launch price, coupled with accelerating financial momentum that is keeping valuations reasonable, creates a compelling price point for an upgrade to Buy."
Bulls are loud: Bernstein at $3,000, Evercore at $2,800, Lynx at $2,550. Even Evercore admits revenue guidance came in light — but gross margin expectations stay strong. Bear case lives at Jefferies: $1,750 PT on moderating NAND pricing and lower GM guidance. Same quarter, two reads. Argus sides with the bulls — 175% growth makes the valuation look cheap, and the profit-taking created the r/r.
Litchfield Hills hangs a Buy and $9 PT on this micro-cap defense name — stock at $1.44. That's roughly 5x implied upside, so either they see something the tape doesn't or this is a lottery ticket with a bow on it. (InvestingPro flags high volatility; at this price, PMs should size accordingly.)
The thesis is full-spectrum: Quantum Cyber assembles an AI-powered, quantum-accelerated autonomous defense platform covering drone warfare, counter-UAS, autonomous naval mine countermeasures, EMP shielding, anti-drone ammunition, C2, and quantum antennas. That's a lot of boxes to check for a company trading at $1.44.
The real driver isn't the tech — it's the procurement shift. The Department of War's Drone Dominance program, SOCOM, and the Marines have ALL launched drone fast-tracking initiatives this year. The military is prototying alongside operational units now, which structurally favors smaller agile vendors over prime contractor inertia.
"Defense industry procurement is changing in ways that favor companies like Quantum Cyber."
The June Bridgeport facility acquisition is the credibility anchor. 50,000 sq ft, $2.3M, tool-and-die capability for mass-producing drone airframes — plus $2.09M in purchase orders through Quantum Drones Corp and a provisional patent on a canisterized modular launch system. They've also thrown a proposal into a U.S. Army competitive selection process. That's real operational motion, not just a pitch deck.
Valuation math: peer-average sales multiple on 2028 revenue gets you to ~$4.50, roughly 3x the current print. The $9 PT is DCF-driven. Either way, you're buying on execution risk clearing between now and 2028.
We wouldn't chase a single initiation on a sub-$2 name — coverage starts are cheap to print. But the manufacturing asset and order flow are concrete, and the Drone Dominance narrative is the right tailwind. Watch the Army proposal for the next catalyst.
The 14% selloff is a guidance haircut, not a demand break. Stifel's OpenAI forensics show Q3 guidance sits 300-400bps below the 32-33% "steady OpenAI" target — and management always sandbags 300-400bps — so the entire gap is just a ~$30-35M sequential step-down in OpenAI spend, not a lost relationship.
Stifel keeps Buy / $305 PT. New model: ~$65M/quarter of OpenAI revenue for the next four quarters, ~$260M total vs. $340M prior — a modest 2% NTM headwind. The key claim: OpenAI has quarterly minimum commitments with Datadog, which de-risks management's own commentary on the relationship. In other words, the biggest customer can't just walk; the dip is timing.
The base business is still ripping: Q2 revenue $1.12B (+36% y/y), $42M beat, fifth straight quarter of acceleration, 79.5% gross margin, NRR stable. Needham went to $300, Macquarie to $260, both positive. Market heard "usage decline from largest customer" and puked; Stifel's framing says at $233 you're paying for the panic, not the problem.
THE 44% POP ISN'T THE STORY — THE REACCELERATION IS. Cloud growth hit 31% (from 29% last qtr), RPO accelerated to 44% from 37%, and Atlassian posted its first clean GAAP operating profit in two years. The market finally paid up for a quarter that had everything: revenue beat by 6.4%, EPS by 23%.
The Street responded en masse — FBN to $170 from $110 (Outperform), BofA upgraded to Buy at $175, BTIG to $180, Jefferies to $200. A collective $170-200 PT cluster off one print. The bull case is simple: record enterprise deal activity, seat expansion and cross-selling driving cloud — not the one-time data center migrations that muddied past narratives.
Cloud revenue acceleration was driven by seat expansion and cross-selling rather than one-time data center migrations.
The catch is the guide. FY27 subscription ARR growth ~18% (from 23%) and non-GAAP op margin ~25% — below the 28.3% consensus and a real step down from the 30.4% delivered in FY26. That's a deliberate growth investment tradeoff, not demand cracking. But it keeps the stock at a reasonable 3.7x FY27 revenue instead of a melt-up multiple.
CEO also threw in a $250M share purchase plan. Small, but optics matter after a 44% week.
DA Davidson is the first credible crack in the NBIS execution story, and they brought better diligence than most — site visit, city council hearing (the whole nine). Cut PT to $175 from $250, kept Neutral. Stock trades at $191.53, so the new target now sits ~9% BELOW the tape. That's a bold call, especially with no rating change.
The issue is physical, not financial. Vineland — the flagship facility — may not complete this year. That's a direct threat to 2026 upside guidance and the core narrative that NBIS is the best neocloud. The stock's +173% over the past year isn't multiple expansion; it's an execution premium. NBIS hit their buildout milestones while CoreWeave and the rest stumbled. Break that chain and the premium compresses fast.
"may remove upside from this year's guidance and risk the narrative that Nebius is the best neocloud"
That's the bear case in one sentence, and it's the whole ballgame.
This is one firm's view, to be clear. Baird is Outperform at $250, Piper Sandler Neutral at $224 — the latter flags NBIS's asset-light scaling model and custom server designs but won't chase the valuation. So the Street is still mostly constructive; DA Davidson just happens to be the only one who went to look at the site.
Earnings print AUGUST 12 — two days out, options implying a 13% move. This cut lands right before the catalyst, which tells you the timing is deliberate. The bar for management is now simple: prove Vineland is on track. If they can't hand-wave it on the call, today's PT is the floor, not the target. If they can, this becomes a buying opportunity on a $48.1B name with the AI inference narrative intact. The r/r here is actually asymmetric — dirt cheap optionality into the print.
Both analysts land on $7 PT — Guggenheim from $13, Oppenheimer from $20 — and both stay Buy/Outperform. That's the street re-underwriting this as a survival story, not a growth story.
The thesis break: Serve slashed 2026 revenue guidance 63% at the midpoint to $9-10M from $26M. Uber volumes declined for the FIRST TIME IN 17 QUARTERS. Serve is publicly walking away — no 2027 renewal — citing frustration with Uber's demand generation as Waymo and other autonomous partners crowd in. Management admitted results didn't meet internal expectations. That's not a trim; that's a strategic reset.
"The reduction reflects ongoing utilization challenges with Uber, which saw delivery volumes decline in the second quarter for the first time in 17 consecutive quarters."
The bull case: DoorDash volumes +50% sequentially. New partner announcement coming this fall (Summer Announcement Event Aug 17). Direct merchant integration is the new angle — less platform dependence, more owned demand. If the pivot works, $7 is the floor, not the ceiling.
The bear math: $170M LTM negative free cash flow. Cash > debt today, but at this burn rate the balance sheet does the talking. Q2 revenue of $3.2M (+400% YoY) still produced a -$0.59 non-GAAP loss.
Stock at $5.09, 73% off the $18.64 high. Most of the bad news is in the price. The question is whether $7 is a recovery target or a tombstone. The opex cut (-12% at midpoint) buys time, but Serve needs that fall partner announcement to be real, not vaporware.
Bernstein says the datacenter scare is overpriced. PT to $660 from $647 (Outperform) — modest bump, but the framing is the signal. The bear case: MSFT is throwing money at AI capacity with no demand visibility, and the market's pricing a bubble bust. Bernstein's rebuttal: MSFT is building BEHIND demand, not ahead of it. Datacenter square footage grows slower than cloud revenue. Lease obligations spread across multiple years. Purchase obligations for power, cooling, hardware barely reach the 12-month mark ($3.77T cap, 27.85x P/E — the bill is manageable).
The kicker that kills the stranded-asset thesis — even if AI demand collapses, MSFT flips the capacity into its existing CPU-based cloud business.
"Even if AI demand were to stop, Microsoft's future datacenter capacity could be shifted to support growth of its existing CPU-based cloud business."
That's the steelman, and it holds. MSFT closed FY26 Q4 with $19.6B cash and positive FCF — the ONLY major US cloud player printing cash while Alphabet and Amazon bleed on AI infra. 17 analysts just revised estimates up, and Tigress went $690 Buy on Azure acceleration + Copilot monetization. Net: the capex fear is a feature, not a bug, when your downside is redeployment into a business growing ~18%.
THE SETUP: STIFEL CUT PT TO $170 FROM $185 (HOLD) BUT RAISED 2026/27 ESTIMATES IN THE SAME BREATH. That's the whole story — this is a valuation reset, not a fundamental breakdown. Camtek keeps printing, and the market is telling you it's already paid for.
Q2 was a clean beat: REV $399.7M vs $382.7M est, ADJ EPS $2.34 vs $2.23. Q3 guide of $158-160M blows through the $149M consensus — a ~7% top-side surprise at the midpoint. Orders >$600M YTD, backlog covers H2 2026 and extends into H1 2027. Advanced packaging runs ~80% of sales by year-end 2026, growing ~30% YoY, with AI/HPC up mid-30s%. This is as good as the demand narrative gets.
The rub: stock +87% over the past year, trading at $157.63 (right at the new PT). Shares slipped on the print despite the beat — investors are now focused on valuation, margins, and what comes AFTER this growth burst. Flat non-advanced packaging and a slower H1 dilute the headline ~20% FY26 growth number to something less magical. Stifel sees Q4 up low-to-mid-teens sequentially, so the trajectory is right, but the multiple has no sprint left unless execution gets even better.
Bull case: AI packaging is still early, backlog visibility into 1H27 de-risks the model, and capacity adds in H2 give revenue upside. Bear case: you're paying a premium multiple for a company decelerating from recent pace into the 20s — and the market just watched a beat-and-raise get sold. Not enough to chase here. Hold is the right call — wait for a better entry or a clear reacceleration signal.
Truist trims to $51 from $55, stays at Hold — but this is mostly peer multiple compression doing the heavy lifting, not a thesis break. Stock sits at $44.86, DOWN >30% IN SIX MONTHS, and Truist's Fair Value screen flags current levels as rich. Fair enough.
They're not blind to the good stuff. Groves hit criticality less than a year after breaking ground — that's real execution. Acquisitions look like they compress the deployment timeline. Centrus LOI starts lighting the fuel supply path.
"The quarter showed Oklo reducing execution risk and building capabilities across power, fuel, and isotopes."
So why the cut? First-of-a-kind economics still unproven, and nobody can see the pathway to nth-of-a-kind cost compression yet. That's the entire debate in one sentence. Truist needs clarity on that curve before leaving the sidelines. Same seat I'm in.
Q2 was a mixed bag. Revenue $1.2M vs $83.8K consensus — that's a rounding error at this valuation, don't kid yourself. Adjusted EPS -$0.28 vs -$0.16 expected (the miss matters more). The actual headline: $3.0B CASH AND MARKETABLE SECURITIES and a raised 2026 spend guide to accelerate project work. They can afford the burn. The question was never survival — it's whether the cost curve bends enough from unit one to unit ten to justify the multiple. No profitability expected this year. You're buying a story, not an income statement.
Freedom Broker sees another leg here. PT to $38 from $30, Buy maintained — implied ~31% upside to the $28.93 close. The thesis is simple: AI infrastructure demand is accelerating ATEN's product business, and the guide raise proves it's not just narrative.
Q2 beat: $0.25 adjusted EPS on $80.1M revenue vs $0.24/$77.3M consensus. TOP LINE +15.5% YOY. GROSS MARGIN 79% — the software-ification of a hardware-rooted business is real. Product sales accelerating is the tell — that's the AI spend showing up in new bookings, not just recurring revenue. Management now guides FY26 to 12-14% revenue growth and 14-16% EPS growth.
The AI angle has three legs:
Stock is +56% over the last year. The $30-$45 PT cluster means the low-end bogey is basically the current price — someone out there thinks fair value is already here. 12-14% top-line growth is solid, not explosive, and this isn't a cheap multiple on those numbers. That shelf registration / prospectus supplement filing is worth monitoring too — probably pre-emptive, but an offering would overhang the name.
Net: rate of change is positive, margins are elite, guidance is heading the right direction. But this is a "size it small, let the momentum work" name — good setup, not a screaming bargain, and thin analyst coverage today means the Street hasn't fully crowded in. That's either opportunity or signal. I lean opportunity, but not a high-conviction 3p.
The beat was the bait. Q2 printed solid — non-GAAP EPS $3.38 vs $3.14 est, revenue +4.5% to $1.42B — and the market barely blinked. That's because management CUT CY2026 revenue growth guidance by another 125bps at the midpoint. SECOND guide-down this year. Stock at $97.43, DOWN 52% YTD. The trajectory matters more than the quarter.
Guggenheim trims PT to $140 from $165 (keeps Buy) but quietly removes EPAM as "Best Idea." The culprit list: rising token costs, shifting priorities at a large customer, and broken North America go-to-market execution. Management's fix doesn't pay off until CY2027 at the earliest. That's an eternity when the tape is this unforgiving.
At ~7x revised CY2027 EPS, the multiple screams cheap. But cheap on a deteriorating sales motion is how you catch a falling knife. William Blair already bailed to Market Perform on sustained North America pressure and limited growth prospects through 2026.
"Investors will likely need more proof points around remediating execution headwinds and improved customer behavior before becoming more constructive."
Exactly. Show-me story until North America inflects. No rush to step in front of this.
DA Davidson nudges PT to $65 from $62, keeps Neutral. Stock's up 86% YTD — this is a "we see it, we just don't chase it" print.
The demand picture is genuinely better. Q2 adjusted EPS $0.45 (beat by $0.03), revenue $291M vs. $291.76M est — a rounding error. RECORD REVENUE, 8TH STRAIGHT QUARTER OF YOY GROWTH, adjusted EPS UP 80% YOY on margins, cost cuts, and operating leverage. Logistics, Consumer Electronics, Packaging, and Semi doing the heavy lifting. Auto remains the laggard.
The offset: memory cost inflation. That's the 2027 tension — top-line momentum vs. input cost creep. And with the stock already up 86% YTD, the easy r/r is gone. Balance sheet is spotless ($750M+ cash, no debt) so the story isn't broken, but Neutral says the multiple already banked the improvement.
"The revised forecasts reflect a more favorable organic growth outlook and ongoing operating expense leverage, balanced against inflationary pressure from memory costs."
15 analysts bumped FY estimates higher post-print — directionally the Street agrees. The fight is whether 86% YTD leaves any juice left. DA Davidson says no.
Needham upped PT to $135 (from $130) after spending time with CFO Dave Wood, and the story is simple: Marriott is the gift that keeps giving, and the model hasn't even captured the premium/luxury tiers yet. AGYS sits at $3.05B market cap, PEG of 0.34, and just printed its BEST SIX-MONTH BOOKINGS STRETCH EVER — visibility into growth acceleration and margin expansion is about as clean as it gets.
This is a classic "early innings of a platform conversion" setup. The 2,000+ Marriott select hotels already deployed are subscription revenue, but the premium/luxury rollout is OFF the financial model entirely. That's option value the market is slowly starting to price. Oppenheimer ($118) and Cantor ($140) are on the same page — record bookings, growing backlog, FY27 guidance raised after a beat-and-raise Q1 (EPS $0.49 vs $0.39 est, rev $87.66M vs $85.98M).
"We believe the record bookings momentum (best six-month stretch) and the Marriott roll-out provides a clear line of sight to both growth acceleration and margin expansion."
The bull case writes itself: subscription mix shifts, margins inflect, Marriott premium tiers come next. The bear case? Not much in this tape — valuation is reasonable, execution has been clean, and the stock is up 24% in six months without getting ahead of itself. Light coverage today, but this one deserves a spot on the watchlist.
Verdict: The stock has run 91% in six months to $33.75 — and Wall Street is still chasing from BELOW the tape. DA Davidson lifts PT to $28 (maintains Neutral) while Rosenblatt goes to $32 (Buy). Both targets sit under where the stock already prints. That tells you everything about positioning vs. valuation.
The story is real, though. AI revenue accelerated to +78% YoY (~$39M), subscription growth clocked 14%, and the quarter beat on both top and bottom (rev $312.4M vs $306.43M est). Management raised full-year guide and — crucially — kept 2H26 unchanged with the callout that the outlook does NOT depend on winning new business. That's backlog visibility, not hopium.
The market still sold the print. Why? Margin pressure and a cautious profit outlook. That's the bull/bear crux right there: FIVN is investing hard in AI (new R&D and sales leadership), and the market wants to see the Rule of 40 path before paying up for the revenue growth. Company says 2027 is the target.
Results countered negative investor sentiment toward the contact center as a service category.
DA's math checks out — 9x 2027 FCF on an 8% current FCF yield — but a Neutral PT $6 below spot is not endorsement. AI growth is inflection, not yet profitability. At the 52-week high, the easy money's been made. Wait for the Rule of 40 proof point or a pullback into the $28-$30 zone before adding.
Stop pricing Micron like a memory trough company. UBS reiterates Buy, holds $1,625 PT (11x 2029 EPS of ~$165, discounted back one year) — stock sits at $872.88, roughly 20x trailing. They trimmed near-term numbers on a wider NAND pricing band into YE and 1H27. Irrelevant to the call. The thesis is structural, not cyclical.
Look at the EPS path: $74.13 (2026) → $184.89 (2027) → $265.65 (2028). All well above consensus. UBS modeling CUMULATIVE FCF >$450B THROUGH 2028 and EPS still >$160 in 2029. That's not a spike, that's a regime change. Industry NAND revenue goes from ~$300B (2026) to ~$495B (2027) on their numbers. If that's anywhere near right, the bears are fighting the wrong variable.
"This represents a structural reset in earnings power, where such durability and visibility should command a shift toward a broader semiconductor multiple."
The visibility piece is getting contractual, not hopium. Moody's upgraded senior unsecured to Baa1 on take-or-pay customer agreements covering a substantial chunk of revenue. That's the durability evidence. ThinkEquity pushed PT to $900 (same direction, way lower multiple — UBS is the bold one). Citi pulled the catalyst watch tag after mixed memory prints elsewhere — near-term timing caution, not a thesis break. Light tape, heavy conviction.
RBC stays Outperform into Wednesday's print, but the $520 PT sits BELOW the $539 tape — weird optics, though targets are lagging the tape, not the thesis. The firm expects a BEAT-AND-RAISE: CY26 Systems outlook bumped above the prior 30% growth guide, with visibility extending beyond the eight-quarter horizon. That matches the drumbeat from semi cap equipment peers — DRAM, HBM, and Advanced Packaging keep firming, and AMAT owns the lead position in all three.
"RBC Capital expects double-digit wafer fabrication equipment growth to sustain for the next three years driven by generative AI, extreme DRAM tightness, and re-emerging competition in Advanced Foundry."
Value-based pricing is a near-term revenue/margin tailwind. 4F² memory, backside power delivery, and panel-level packaging extend the cycle beyond the current memory upleg. NAND (~4% exposure) and ICAPS (10-15%) are modest kickers, not the narrative. The Street split says it all: Erste downgraded to hold on below-average relative growth while Stifel and Mizuho both push to $650. That's a valuation debate, not a fundamental one. Options imply a 7.5% MOVE on the print. With the rate of change in memory still pointing up, I lean with the $650 crowd.