Most Marvell Technology (MRVL) coverage frames the debate as a question about AI demand. The analyst distribution says the real disagreement is about something else entirely. At $188.68 as of July 17, 2026, MRVL carries a consensus target of $252.56 across 43 analysts — but the range behind that average runs from a low of $110 to a high of $385 (StockAnalysis). That is a spread implying anywhere from a 41.7% loss to a 104% gain on the same company, over the same horizon, using the same public information. A 250% gap between the most bearish and most bullish target is extraordinary for a large-cap semiconductor name, and it does not come from disagreement about whether AI infrastructure spending is real.

It comes from customer concentration. Marvell derives roughly 45% of revenue through a single distributor and 82% from its top-10 customers. That single pair of figures explains the entire distribution. If those relationships are design wins — multi-year custom silicon programmes that are painful to unwind — then concentration is a moat and $385 is defensible. If they are purchase orders that can be re-sourced, then losing one top customer removes 8% or more of revenue in a quarter, and $110 stops looking like a panic number. Having covered custom-ASIC vendors through two hyperscaler procurement cycles, this is the distinction that separates the two camps, and almost no competing analysis states it as the actual crux.

Key Facts

• MRVL trades at $188.68 as of July 17, 2026, against a consensus target of $252.56 implying 33.9% upside — StockAnalysis
• Analyst targets range from $110 (-41.7%) to $385 (+104.1%) across 43 analysts, with a “Strong Buy” consensus rating
• Roughly 45% of revenue flows through a single distributor and 82% comes from the top-10 customers
• RBC Capital Markets models 40%+ revenue growth sustained for three years and data centre revenue rising 50%+ this year and next, at a $360 target
• UBS raised its target to $340 from $230; KeyBanc issued a $400 target on July 14, 2026
• Marvell acquired Celestial AI, disclosed alongside its fiscal Q3 2026 results — CNBC
• The company is integrating with NVIDIA’s ecosystem via NVLink Fusion, extending beyond its independent custom-silicon business

What Marvell actually sells, and why concentration is structural

Marvell is not a merchant chip vendor in the way Nvidia is. Its core business is custom silicon: it co-designs application-specific integrated circuits (ASICs) for individual hyperscalers, alongside optical digital signal processors, silicon photonics and high-performance analog components that move data inside and between data centres.

The useful analogy is contract aerospace manufacturing rather than component retail. A company that machines a specific structural part for one airframe programme does not have thousands of customers, and would not want them. It has a handful of relationships, each worth enormous revenue, each embedded in a multi-year certification cycle. Concentration is not a bug in that model — it is the direct consequence of the business being hard enough that only a few customers can use what you make.

That framing matters because it changes what the 82% figure means. In a commodity business, 82% revenue from ten customers signals fragility. In custom ASIC design, it signals that you have won ten of the roughly fifteen programmes worth winning. The risk is not that customers are fickle; it is that each individual programme is enormous, so a single loss at the next design refresh is a step-function event rather than a gradual erosion.

Marvell’s own framing leans hard into the durability side. “We are in the early innings of a multiyear infrastructure buildout,” said Matt Murphy, Chairman and Chief Executive Officer of Marvell Technology, on the company’s earnings call.

What the customers and partners are actually doing

The most significant recent development is not a customer win but a partnership that changes Marvell’s competitive position. The company is connecting its silicon portfolio to NVIDIA’s ecosystem through NVLink Fusion — a notable move for a business whose custom-ASIC franchise has historically been sold as the alternative to buying Nvidia merchant silicon.

“By connecting Marvell’s leadership in high-performance analog, optical DSP, silicon photonics and custom silicon to NVIDIA’s expanding AI ecosystem through NVLink Fusion, we are enabling customers to build scalable, efficient AI infrastructure,” Murphy said.

Read commercially, that is a hedge. If hyperscalers keep building custom accelerators, Marvell wins on ASIC design. If they consolidate onto Nvidia platforms — the scenario laid out in our Nvidia $302 bull versus $152 bear breakdown, where Vera Rubin orders run to $1 trillion through 2027 — Marvell still supplies interconnect and optics into those racks. The partnership converts a binary bet into a position with two ways to win, which is precisely what a company with 82% top-10 concentration should be doing.

On the acquisition side, Marvell bought Celestial AI, disclosed with its fiscal Q3 2026 results. Optical interconnect is the emerging bottleneck in scaling AI clusters — as accelerator counts rise, moving data between them becomes the constraint rather than raw compute. Buying into that layer is consistent with the interconnect-and-optics hedge rather than a bet on winning more ASIC sockets.

What has not been disclosed is equally relevant. Marvell has not named the single distributor representing 45% of revenue, nor broken out per-customer exposure within the top ten. That opacity is legal and normal, but it is why the bear targets exist: analysts cannot model the downside precisely, so the conservative ones assume the worst.

Market impact: what the numbers actually support

Setting the inputs against each other makes the disagreement measurable.

Here is the synthesis neither camp states directly. RBC’s model of 40%+ revenue growth sustained for three years is not a demand forecast — it is a retention forecast. Sustaining that rate with 82% of revenue in ten accounts requires effectively zero programme losses across three consecutive design cycles. That is a demanding assumption, and it is not made explicit in the target.

Run the arithmetic from the other direction and the bear number becomes legible. If Marvell lost one meaningful top-10 customer and the associated revenue did not re-source, the revenue base contracts while the multiple compresses simultaneously — because the market would immediately reprice the concentration risk it had been ignoring. Revenue down and multiple down together is how a stock goes from $188.68 to $110 without the AI thesis being wrong at all. That dual mechanism is why the low target sits 41.7% below spot rather than at a modest discount.

The pattern is familiar from adjacent names. In our Micron $1,486 bull versus $740 bear analysis, the spread also turned on whether contracted revenue is structurally durable or cyclically flattering. Micron’s answer was take-or-pay contracts covering an entire year of supply, with purchase orders extending into 2028. Marvell has no equivalent public disclosure, which is a genuine informational disadvantage when investors are trying to price exactly that question. Two companies exposed to the same AI buildout, and the one that published its contract structure gets a narrower target range.

There is a third comparison worth making, because it isolates the variable. Our AMD forecast covering a $700 bull and $385 bear case describes a merchant vendor selling standard parts to a broad customer base. AMD’s bull-bear spread is driven by share-gain assumptions against Nvidia. Marvell’s is driven by retention within a customer list it already holds. Those are different risks that produce superficially similar-looking dispersion, and conflating them is the most common analytical error in this part of the market.

The practical consequence for anyone modelling MRVL is that the usual semiconductor inputs matter less than normal here. Foundry pricing, wafer allocation and end-market demand all feed the model, but none of them moves the needle the way a single procurement decision at one hyperscaler does. That is an uncomfortable position for a $188.68 stock with a “Strong Buy” consensus, and it is the reason the low target sits where it does rather than at a conventional 15% discount to spot.

Regulatory and geopolitical tension

Marvell sits at an awkward intersection of export controls and supply-chain policy. Its custom ASICs are designed in the United States and fabricated primarily at Taiwanese foundries, then deployed into data centres worldwide. US export controls on advanced accelerators restrict where the highest-performance parts can ship, and custom silicon designed for a hyperscaler’s global fleet must be architected around those restrictions from the start.

The more specific exposure is Taiwan concentration. Unlike Micron, which manufactures across the US, Japan, Singapore and Taiwan, a fabless designer carries geographic risk it cannot diversify away on its own timeline — foundry qualification for a leading-edge custom part takes quarters, not weeks. Any disruption to Taiwanese capacity is a direct revenue event, and it is not reflected in a growth-based valuation model.

There is also a quieter antitrust dimension worth watching. As custom silicon becomes the primary route for hyperscalers to avoid dependence on merchant accelerators, regulators examining AI compute concentration have an interest in keeping that route open. That is structurally favourable to Marvell — the policy incentive runs toward preserving alternatives to a single dominant supplier. It is one of the few regulatory dynamics in semiconductors that points in a company’s favour rather than against it, though no enforcement action has made it concrete.

What happens next: three predictions

First, the concentration disclosure becomes the swing factor. If Marvell begins breaking out customer or programme-level revenue with more granularity, the bear targets compress quickly, because the uncertainty premium in the $110 case is largely informational rather than fundamental. Watch the next 10-K risk factors more closely than the earnings headline.

Second, NVLink Fusion revenue shows up before new ASIC wins do. Interconnect and optics attach to racks being deployed now, whereas a new custom programme takes multiple quarters from win to revenue. Expect the partnership to contribute measurably ahead of any announced design win, which will make the growth look more diversified than the customer list actually is.

Third, the target range narrows without the stock moving much. A $110-to-$385 spread is unstable — it reflects genuine uncertainty rather than genuine disagreement about value. As the retention question resolves in either direction over the next two reporting cycles, expect analysts to converge toward the middle before price follows. Dispersion of this width tends to collapse from the tails inward: the $110 case is retired by a single clean quarter of retention, and the $385 case is retired by any disclosed programme loss. Both tails are more fragile than the midpoint, which is why convergence usually precedes direction.

The honest conclusion: at $188.68, MRVL is priced close to the consensus midpoint of a debate that has not been settled. The bull case to $385 requires three years of clean programme retention. The bear case to $110 requires one meaningful loss. Neither is remote, which is exactly why the spread is this wide — and why the concentration figures, not the AI demand data, are the numbers to track.

FAQ

What is Marvell’s (MRVL) stock price right now?
Marvell Technology traded at $188.68 as of July 17, 2026, against a consensus analyst target of $252.56, implying roughly 33.9% upside. The stock has been notably volatile through mid-July.

What is the analyst price target for MRVL stock?
The consensus is $252.56 across 43 analysts with a “Strong Buy” rating. Targets range from $110 at the low end to $385 at the high end, with RBC Capital Markets at $360, UBS at $340 and KeyBanc issuing $400 on July 14, 2026.

Why is Marvell’s analyst target range so wide?
Customer concentration. With roughly 45% of revenue through a single distributor and 82% from the top-10 customers, the difference between a bull and bear case is whether those relationships are durable design wins or re-sourceable orders. That single question produces a 250% spread in targets.

What does Marvell actually make?
Custom silicon — application-specific integrated circuits co-designed for individual hyperscalers — plus optical digital signal processors, silicon photonics and high-performance analog components used to move data within and between data centres.

How does Marvell relate to Nvidia?
Both competitively and cooperatively. Marvell’s custom ASIC business is an alternative to buying merchant accelerators, but the company is also integrating with NVIDIA’s ecosystem through NVLink Fusion, supplying interconnect and optics into Nvidia-based racks.

What would invalidate the bull case on MRVL?
The loss of a single meaningful top-10 customer or design programme. With 82% of revenue concentrated in ten accounts, one loss removes roughly 8% or more of revenue and simultaneously forces the market to reprice concentration risk — revenue and multiple falling together.

This article is informational analysis only and does not constitute investment advice. Semiconductor equities are highly volatile and custom-silicon revenue is subject to programme-level concentration risk. Prices and analyst targets quoted are timestamped snapshots as of July 2026. Conduct your own research and consult a regulated financial adviser before making any investment decision.

As the second-quarter earnings season of 2026 approaches its most critical stretch, the global equity market finds itself at a pivotal crossroads.

For over two years, a relentless, AI-driven bull run has propelled mega-cap technology valuations to historically elevated levels.

However, the narrative on trading desks has undergone a fundamental shift. The era of rewarding companies simply for uttering the words “artificial intelligence” is officially over.

As Alphabet, Microsoft, Meta, Amazon, and Apple prepare to open their books between July 22^nd  and July 30^th, Wall Street is demanding concrete evidence of monetization.

Investors are no longer grading on a curve; they want to see the receipts.

Big tech earnings ahead: the $725 billion arms race

The defining metric of this entire reporting cycle will undoubtedly be capital expenditure (capex).

The sheer scale of infrastructure investments being deployed by the four major US hyperscalers – Amazon, Microsoft, Alphabet, and Meta – has reached eye-watering proportions.

According to updated consensus data, their combined capex guidance now sits at an unprecedented $725 billion for the current year, representing a staggering 77% increase from 2025.

2026 projected capex commitments:

Amazon: ~$200 billion

Microsoft: ~$190 billion

Alphabet: $180 billion – $190 billion

Meta Platforms: $125 billion – $145 billion

This staggering allocation of capital into graphics processing units (GPUs), power grids, and massive data center footprints has triggered intense anxiety among institutional allocators.

While this structural build-out serves as a massive secular tailwind for hardware providers like Nvidia (which won’t report its data center metrics until August 26), it places immense pressure on the software and cloud giants to prove this capital is yielding high-margin returns.

A guidance cut this week would signal weak underlying enterprise demand – while an unbacked increase in spending without a corresponding bump in revenue could spark a sharp margin-driven sell-off.

The reporting calendar: key dates and battlegrounds

The heavy lifting begins next week, with the market tightly focused on three specific reporting windows:

• July 22, 2026 (Alphabet): Google’s parent company kicks off the gauntlet alongside Tesla. Alphabet’s Q1 results saw Google Cloud revenue expand by an astonishing 63% year-on-year to hit $20 billion, boasting a record 32.9% operating margin. Wall Street is looking for Q2 revenue to hit roughly $116.8 billion. The core focus will be whether Google Cloud can sustain its 63% growth crown or if aggressive new market entrants have begun eating into its enterprise pipeline.
• July 29, 2026 (Microsoft & Meta): Microsoft will present its fiscal fourth-quarter results, where any print for Azure growth below 35% will likely be treated as a severe deceleration. Simultaneously, Meta will need to prove that its $125 billion+ capex is continuing to optimize its ad-targeting engine and drive top-line growth to offset the massive cash burn of its infrastructure layer.
• July 30, 2026 (Amazon & Apple): Amazon is expected to print revenue near $196 billion, with the market hyper-focused on AWS margin expansion. Apple will report its fiscal third-quarter numbers with an estimated revenue of $108.9 billion. Apple presents a fascinating contrarian play; by leveraging an installed base of over 2.3 billion active devices to deploy “Apple Intelligence,” it is executing a capital-light AI strategy that insulates its margins from the data center spending war engulfing its peers.

Cloud growth: The ultimate litmus test

Because cloud infrastructure is where enterprise AI demand materializes first, the sequential and year-over-year growth rates of Azure, AWS, and Google Cloud will serve as the market’s ultimate truth mechanism.

Investors are highly attuned to the risk of a “margin squeeze” – a scenario in which heavy depreciation costs from newly built data centers kick in before corporate clients scale up their paid software seats and API usage.

A note of caution was already introduced to the broader tech sector following IBM’s earnings miss on July 14^th, which triggered a sharp one-day decline.

While analysts isolated that specific event to hardware supply-chain timing rather than systemic weakness in macro AI demand, it illustrated just how fragile investor sentiment has become.

With valuations priced for perfection, the upcoming multi-day stretch will decide whether Big Tech’s massive architectural bets can sustain the next leg of the macroeconomic expansion, or if the market is due for a harsh reality check on the actual velocity of AI monetization.

The post Big tech earnings outlook: Wall Street demands receipts on $700B AI spree appeared first on Invezz

The comforting story about SpaceX stock is that Thursday’s aborted Starship launch was just an engineering hiccup and the dip is a buying opportunity. Both halves might be true, but the framing misses what actually changed: for the first time in SpaceX’s history, a routine test-flight abort — the kind that happened repeatedly before the IPO with zero financial consequence — is now a market event that erased billions in public-market value. SPCX trades near $125.40 as of July 17, 2026 (MarketBeat), below its $135 IPO price and down 18.5% since the June debut, after the company scrubbed its 13th Starship flight when four Super Heavy engines failed to ignite. The analyst spread on the same stock runs from CFRA’s $115 bear case to Raymond James’ $800 bull case — a seven-fold range that is the widest we have tracked across this entire bull-versus-bear series.

That $115-to-$800 spread is the story, because it is not a disagreement about a quarter — it is a disagreement about what SpaceX is. Having mapped this series across HIMS, Ethereum and the AI-infrastructure names, SPCX is the cleanest case yet of a private-market valuation dream colliding with public-market price discovery in real time. In the private secondaries, SpaceX was marked toward a $3 trillion-plus valuation on the promise of Starlink and Starship; in the public tape, retail that bought the hype above $200 is capitulating, and a crypto prediction market’s odds of SPCX closing July higher collapsed from 61% to 32% in days, per Benzinga. The abort did not change the engineering. It changed who sets the price — and that transfer, from patient private capital to daily public sentiment, is the real event.

Key Facts:

• • SPCX trades near $125.40 (July 17, 2026), below its $135 IPO price and down 18.5% since the June 12 debut — MarketBeat, CNBC
• • Bull case: $800 — Raymond James, implying roughly +398% upside — MarketBeat
• • Bear case: $115 — CFRA, implying roughly −31% downside; Moffett Nathanson sits at $131 — MarketBeat
• • Average 12-month target: ~$234–$244 across 32–37 analysts, a Moderate-to-Strong Buy consensus (24 Buys) — MarketBeat, TipRanks
• • Starship Flight 13 aborted July 16 when four Super Heavy engines failed to ignite; two Raptors will be replaced — Euronews
• • Crypto prediction-market odds of SPCX closing July higher fell from 61% to 32% — Benzinga
• • SPCX was added to the Nasdaq-100 in early July, then fell below its debut price in a multi-day slide — CNBC

What’s actually happening: the first launch as a listed company

Flight 13 was the first Starship test since SpaceX became a public company, and that context turned a familiar engineering ritual into a governance-grade event. The launch window opened at 5:45 p.m. Texas time on July 16; the automatic abort fired before liftoff when the Super Heavy booster’s startup sequence left at least four engines unlit. The safety system did exactly what it is built to do — too few engines is a scrubbed launch, not a failed one — and pre-IPO, that distinction was academic to anyone but engineers. Post-IPO, it printed red on a Nasdaq-100 component.

Elon Musk narrated the abort in real time. “Some of the engines didn’t start, triggering an automatic launch abort. Now offloading propellant. Next launch attempt hopefully in a few days,” he posted, following up with the fix: “To be confident of a good flight, 2 Raptors will be removed & replaced. Most probable launch timing is early next week.” (Euronews) In hardware terms, replacing two Raptor engines and retrying within days is a routine turnaround. In market terms, it extended a six-day losing streak and pushed the stock further under its offer price — the disconnect FinanceFeeds flagged the moment the listing priced, in why Wall Street couldn’t let SPCX fall.

The deeper shift is temporal. SpaceX’s development culture is built on rapid iterative failure — blow up early prototypes, learn, refly — a cadence that made it the most capable launch provider on Earth precisely because aborts and explosions carried no financial penalty. A public listing attaches a daily price to every one of those events. The company that thrived on visible failure now answers to a market that punishes it.

Industry response: the valuation-versus-revenue argument goes public

The most consequential response is not from an institution but from the retail base now setting the marginal price. The prediction-market collapse from 61% to 32% is one signal; the capitulation on trading forums is louder. The single most-upvoted SPCX post this month is a retail investor documenting a $200,000 loss on the stock, with the top reply — 2,075 upvotes — pointing at the core bear thesis: “Crazy that people knew the insanely bloated valuation but still went long.” Another widely-shared comment did the math that anchors the skeptics: “when you see valuation of 3T and they only make like 50b off Starlink.” That is the entire bear case in one sentence — a spacefaring monopoly priced as though Starlink’s revenue already justified a multi-trillion-dollar market cap.

Institutional desks are, remarkably, on the opposite side. Of roughly 32 to 37 covering analysts, the consensus is a Buy, with Raymond James at $800, Arete Research at $401, Deutsche Bank at $255, Needham at $250 and Morgan Stanley at $225 — an average near $234–$244 that implies more than 85% upside from the July price. Only CFRA ($115) and Moffett Nathanson ($131) see downside from here. This is the mirror image of the HIMS setup we broke down in the $40-versus-$21 bull/bear case: there, the market price had run above the analyst average and the street was upgrading from behind; here, the market has fallen below the entire target range and the street is standing pat, insisting the tape is wrong. When retail and the sell side diverge this violently on a Nasdaq-100 name, one of them is about to be repriced.

The numbers: $800 bull, $115 bear, a seven-fold spread

Sources: MarketBeat and TipRanks analyst compilations (July 17, 2026). Table compiled July 17, 2026.

The synthesis no single target reveals: the SPCX spread is not just wide, it is the widest in this series — a seven-fold range from $115 to $800 on one stock, versus the roughly ten-fold on OKLO, but on a company already generating real Starlink revenue rather than a pre-revenue reactor. When we deconstructed OKLO’s $140-versus-$14 split, the disagreement was about whether a technology would work at all. SPCX’s disagreement is subtler and more dangerous for holders: everyone agrees the technology works — Falcon 9 is the most reliable rocket in history and Starlink is a genuine business — the fight is entirely about the multiple. A spread this wide on a proven operator means the market has no shared framework for valuing it, and stocks without a shared valuation framework trade on flows and sentiment, not fundamentals. That is precisely why a launch abort moved it. Until an earnings cadence gives the market a number to anchor on, every Starship event — success or scrub — will swing the price more than the underlying business warrants.

There is a crypto-native footnote to the price discovery that is more than a curiosity. Before SPCX ever traded on Nasdaq, tokenized SpaceX proxies on venues like Gate and pre-IPO platforms were already quoting the company, and those synthetic markets ran hot into the debut — FinanceFeeds documented how the SpaceX IPO crushed every tokenized stock on day one and how crypto rails priced SPCX toward $2.3 trillion first. Those same 24/7 markets are now the leading indicator on the way down: the Benzinga-cited prediction market repricing from 61% to 32% is exactly the kind of continuous, capital-backed sentiment gauge that traditional equity markets lack between earnings dates. For a stock whose fair value nobody agrees on, the prediction markets are functioning as the real-time consensus the analyst PDFs cannot provide — and right now that consensus is bearish into month-end.

The regulatory and structural layer: index flows meet founder control

Two structural forces now pull against each other. On one side, the early-July Nasdaq-100 inclusion forces passive index funds to hold SPCX regardless of valuation — a mechanical bid that should dampen downside. On the other, the same inclusion means every index-tracking retirement account is now exposed to a stock whose price reacts to rocket-engine ignition sequences, a concentration of idiosyncratic risk that has drawn scrutiny given SpaceX’s dual-class structure and Musk’s concentrated control. The tension is the governance version of the innovation-versus-caution push-pull: index membership demands the stability of a blue chip, while the company operates with the risk appetite of a venture-stage moonshot.

There is a live regulatory subplot too. Reporting that SpaceX may direct stock toward federal children’s savings accounts, alongside the company’s deep government-contract entanglement (NASA, the Space Force, and Starlink’s defense business), keeps SPCX in a category few Nasdaq-100 names occupy: a listed equity whose largest customer and primary regulator are frequently the same US government. For institutional allocators, that is both a moat and a headline risk — the kind of dual-use exposure that widens, rather than narrows, the plausible valuation band.

What happens next: three scenarios into the relaunch and August

Prediction one — the relaunch is the near-term binary. Musk targets “early next week” for the retry; a clean Flight 13 that reaches its Starlink V3 deployment objective likely snaps the six-day losing streak and lets the stock reclaim its $135 IPO line, because the abort narrative reverses fastest when the very next attempt succeeds. A second scrub does the opposite, hardening the “priced for perfection, delivering delays” story that the sub-IPO price already reflects.

Prediction two — August earnings are the settlement date, and the bar is brutal. As one retail skeptic framed it, the market wants to see whether a company with “high dollar expenses” can justify the multiple when the first public earnings arrive. Expect the print to matter more than any launch: it will hand the market its first real number to anchor the $115-to-$800 spread, and the spread will compress hard toward whichever end the Starlink margin trajectory supports.

Prediction three — the analyst-versus-retail gap resolves toward the middle, not the extremes. The street’s $234 average will drift down as targets get marked to the tape, while the retail capitulation overshoots and creates the bounce; the likeliest 2026 path is a stock that spends months proving it deserves a number between Moffett’s $131 and Morgan Stanley’s $225, with each Starship flight a volatility event around that grind. SpaceX will keep being the most important private company to go public in a generation — and, for now, the clearest live experiment in what happens when a valuation built in the private markets has to be defended, launch by launch, in the public ones.

FAQ

Why did SpaceX stock (SPCX) fall below its IPO price?
SPCX slid below its $135 IPO price for the first time in mid-July 2026 and traded near $125 by July 17, down 18.5% since the June debut. A six-day losing streak, the aborted Starship Flight 13, and a broad view that the IPO priced in a multi-trillion-dollar valuation the revenue does not yet support all compounded the decline.

What is the bull case for SpaceX stock?
Raymond James’ $800 target is the extreme bull anchor (+398%), with a street average near $234–$244 implying more than 85% upside. The case rests on Starlink’s subscriber and revenue growth plus Starship optionality — reusable heavy-lift enabling everything from satellite deployment to lunar and Mars logistics.

What is the bear case for SpaceX stock?
CFRA’s $115 target (−31%) is the published bear floor. The thesis: a roughly $3 trillion implied valuation against Starlink revenue estimated near $50 billion is unjustifiable, index inclusion forces holders into idiosyncratic launch risk, and every Starship scrub now carries a financial penalty it never did privately.

What happened at the SpaceX Starship launch on July 16?
SpaceX aborted Starship Flight 13 before liftoff when four Super Heavy engines failed to ignite and the automatic safety system scrubbed the attempt. Elon Musk said two Raptor engines will be replaced and targeted a relaunch “early next week.” It was the first Starship test since the IPO.

Is SpaceX stock a buy after the drop?
Wall Street says yes — a Moderate-to-Strong Buy consensus with 24 Buys and ~87% average upside. The market disagrees, having pushed the price below every target but two. The $115-to-$800 spread means there is no consensus framework; the August earnings print and the Flight 13 relaunch are the near-term catalysts that will narrow it.

Does a launch abort really matter to SpaceX’s business?
Operationally, no — replacing two Raptor engines and reflying within days is routine, and pre-IPO these events carried no financial cost. The change is structural: as a Nasdaq-100 component, SpaceX now attaches a daily public price to a development culture built on rapid, visible failure, so aborts move the stock far more than they affect the underlying company.

What is SPCX’s price target range?
$115 (CFRA) to $800 (Raymond James), with a ~$234–$244 average across 32–37 analysts — a seven-fold spread, the widest in FinanceFeeds’ bull/bear series, reflecting deep disagreement over how to value a proven space monopoly with early public-market revenue disclosure.

Wall Street’s main indexes opened lower on Friday as investors continued to pull back from semiconductor stocks, extending a broader reassessment of the artificial intelligence-driven rally that had powered markets to record highs earlier this year.

The Dow Jones Industrial Average fell about 486 points, or 0.9%, while the S&P 500 lost 1.1%.

The Nasdaq Composite dropped 1.7%, reflecting renewed weakness across technology stocks.

The latest decline followed another sharp selloff in semiconductor shares on Thursday, with investors questioning whether the pace of AI-related capital spending can be sustained after months of strong gains.

Semiconductor stocks extend losses

Chip stocks led the market lower in trading as the sector’s recent pullback accelerated.

Nvidia shares fell about 3.4%, while Applied Materials and Lam Research each dropped more than 5%.

Intel, KLA Corporation, Arm and Micron Technology also traded lower.

The iShares Semiconductor ETF (SOXX) and the VanEck Semiconductor ETF (SMH) both declined more than 3%.

The Philadelphia Semiconductor Index remained under pressure after hitting a nearly two-month low on Thursday.

The benchmark has fallen more than 19% from its late-June record high and was on track for its worst weekly performance since March 2025.

The weakness came despite strong quarterly results from Taiwan Semiconductor Manufacturing Co. (TSMC) and upbeat guidance from ASML, suggesting investors remain focused on broader concerns surrounding AI infrastructure spending rather than company-specific earnings.

The selloff was not limited to US markets.

Semiconductor shares also weakened across Asia-Pacific and European markets on Friday.

Chinese startup Moonshot AI also added to competitive concerns after unveiling a new artificial intelligence model that it said narrows the gap with leading US offerings.

Netflix drops after weak outlook

Technology stocks faced additional pressure after Netflix forecast third-quarter revenue and earnings below Wall Street expectations.

Netflix shares plunged more than 11% in trading despite reporting second-quarter results that were broadly in line with analyst estimates.

Elsewhere, Intuitive Surgical fell 11% after maintaining its da Vinci procedure growth forecast and warning that insurance-plan changes may be delaying patient care.

Investors also awaited the University of Michigan’s consumer sentiment survey and industrial production data later in the day, which were expected to provide further insight into the health of the US economy following a busy week of inflation reports and second-quarter earnings.

Although major US banks delivered solid earnings earlier in the week and recent inflation data came in softer than expected, those positives failed to offset mounting concerns surrounding technology valuations.

Geopolitical tensions remain in focus

Investors also monitored escalating tensions in the Middle East.

The US military said it completed its sixth consecutive evening of strikes against Iran, targeting military infrastructure, logistics assets and maritime capabilities.

Iran, meanwhile, said it had targeted US military forces in Syria and Bahrain, while Kuwait reported that an Iranian attack struck a power and water desalination plant.

The renewed conflict has further weakened the fragile truce reached last month and continued to disrupt energy flows through the Strait of Hormuz, a critical shipping route that normally carries around one-fifth of global oil supplies.

Oil prices moved higher amid the geopolitical developments. US West Texas Intermediate crude traded above $81 a barrel, while Brent crude rose above $86 a barrel.

The CBOE Volatility Index, often viewed as Wall Street’s fear gauge, also climbed to its highest level in more than a week as investors adopted a more cautious stance heading into Friday’s session.

The post Dow sinks 480 points as AI selloff deepens, chip stocks extend losses appeared first on Invezz

The lazy read on the ETH price at $1,576 — down roughly 46% this year — is that the analyst community has gone bearish with it. The published numbers say something stranger: the most institutional bull on the street, Standard Chartered, still carries a $7,500 end-2026 target, while Citi’s freshly cut $3,175 target — the reduced one — implies a double from here, and its recessionary bear case sits at $1,198 (CoinGecko’s expert survey, July 2026). The real bear number lives somewhere the sell side won’t print it: Kalshi’s regulated prediction market prices an 18% chance ETH trades below $750 in 2026, per Yahoo Finance. That, in plain terms, is the actual live spread on Ethereum right now — $7,500 hope against a $750 tail — and the market price sits closer to the tail than to any target on the board.

Having tracked this bull/bear series across equities all year, ETH presents a structure none of the stock names share: the widest gap between public conviction and private positioning. Fundstrat’s Tom Lee spent early July arguing Ethereum “could be a $5 trillion network” — implying six-figure percentage upside from today’s roughly $190 billion market cap — while reporting by Wu Blockchain’s research desk documents that Fundstrat’s own 2026 outlook privately modelled a first-half pullback into the $1,800–$2,000 range. The market has since undercut even the private number. When the loudest bull’s internal base case breaks and the price keeps falling, the question stops being “who is right?” and becomes “what would force either side to capitulate?” — and that question has three dates attached to it.

Key Facts:

• • ETH trades near $1,576 as of July 1, 2026 — down roughly 46% year-to-date — CoinGecko
• • Bull case: $7,500 by end-2026 — Standard Chartered’s institutional target, with $40,000 modelled by 2030 — CoinGecko expert survey
• • Bear case: $1,198 — Citi’s recessionary scenario, alongside a cut 12-month base target of $3,175 (from $4,304) — CoinGecko
• • Tail risk: Kalshi’s regulated prediction market prices an 18% probability of ETH below $750 in 2026 — Yahoo Finance, June 22, 2026
• • Fundstrat’s public base case for ETH reached $12,000, while its 2026 outlook privately modelled an $1,800–$2,000 first-half pullback — Wu Blockchain Substack
• • BitMine holds 5.62 million ETH (~$9.7 billion as of June 15) and bought 76,881 ETH in a single June week — Yahoo Finance
• • A former Ethereum Foundation contributor warns core development needs roughly $30 million annually amid a possible funding shortfall — Yahoo Finance

What’s actually happening: a 46% drawdown with a fight over the floor

Ethereum’s 2026 has been a repricing of everything the 2025 bull market believed. The year began with ETH as the institutional tokenisation play; it enters late July fighting to reclaim $2,000, with technical analysts framing that level as the “biggest obstacle” on any recovery path and community capitulation showing in real time — the most-upvoted ETH investment threads this month are from holders who sold near $1,100 asking whether to return. The proximate causes stack: crypto ETF flows spent eight weeks in net outflow before stabilising — with Ether funds still bleeding even as Bitcoin funds recovered, as FinanceFeeds tracked in the July 16 ETF flow report — while Citi explicitly tied its target cut to slow US market-structure legislation and weakening on-chain activity.

Beneath the price, a governance subplot is feeding the bear case: the proliferation of parallel organisations — Etherealize, the Ethereum Community Foundation, EthLabs, and now an “Ethereum Institutional” initiative — has left even core community forums asking who actually funds and steers protocol development. Trent Van Epps, a former Ethereum Foundation contributor, put a number and a timeline on it: “From recent conversations across all core development, there is a risk we will enter a slow-burning funding crisis within the next 3-9 months,” he warned, citing roughly $30 million in annual funding needs. (Yahoo Finance)

Industry response: treasuries accumulate while the ETF door revolves

The strongest counterforce to the drawdown is corporate: BitMine — the Ethereum treasury vehicle Tom Lee chairs — held 5.62 million ETH worth roughly $9.7 billion as of June 15 and added 76,881 ETH in a single week of the drawdown. That is the Strategy playbook mapped onto Ether: a listed balance sheet converting equity and debt into protocol ownership on every dip, price-insensitive by design. Set against it, the ETF channel keeps leaking — Ether funds posted outflows even on days the broader crypto complex recovered — and the split matters for microstructure: treasury buyers lock supply for years, while ETF flows mark sentiment daily.

Lee’s public posture through the drawdown has not moved an inch. “Should Ethereum be a trillion or $2 trillion or a $5 trillion network value? Yeah, I can easily see it in the next few years,” said Tom Lee, Co-founder at Fundstrat and Chairman at BitMine, on the New Era Finance podcast in early July, adding: “I think there’s a lot of upside, I’m pretty confident about the price upside.” (CCN) On the funding controversy, his rebuttal was categorical: “In my opinion, zero chance of this ‘crisis’ happening for ETH.” The asymmetry worth noting: the man making the $5 trillion argument also runs the vehicle whose 5.62 million ETH would be the argument’s largest beneficiary — conviction and exposure are the same trade here.

The numbers: $7,500 hope, $1,198 fear, $750 tail

Sources: CoinGecko expert survey (July 2026); Wu Blockchain Substack; Yahoo Finance/Kalshi (June 22, 2026). Table compiled July 17, 2026.

The synthesis the individual targets hide: ETH at $1,576 trades below every published sell-side scenario except the explicit recession case — a position none of the equity names in this series occupies. When we mapped HIMS’s $40-versus-$21 spread, the market price had outrun the analyst average; when we broke down APLD’s standoff, even the bear case sat above spot. Ethereum is the APLD setup at asset-class scale — the price has already paid out most of the bear thesis, so the live argument is not $7,500 versus $3,175 but whether the $1,500 floor the market keeps testing holds against the Kalshi tail. Prediction-market pricing is the honest tell precisely because it is capital at risk rather than a research PDF: 18% below $750 means the market assigns nearly one-in-five odds to an outcome no bank has published.

Quick Take: The ETH price at $1,576 sits below every published sell-side scenario except Citi’s explicit recession case. The bull argument is no longer about hitting $7,500 — it is about the market price merely closing the gap to Citi’s own reduced $3,175 target, a double. The bear argument is no longer about targets at all — it is Kalshi’s 18% odds of sub-$750, a tail that only fires if ETF outflows resume and the funding controversy escalates into visible developer attrition.

The extreme end of the bull distribution deserves its own line, because it reframes what “bull case” means here. In June, Lee told CoinDesk that ETH could ultimately reach $250,000 as corporate validators take over network security economics — an argument about Wall Street running validator infrastructure, not a price target in any conventional sense, per CoinDesk. Even the podcast version implies numbers the street will not print: a $5 trillion network value against roughly 120.7 million ETH in circulating supply arithmetically implies a per-coin ETH price above $41,000 — CCN ran exactly that division when the claim landed. The distance between $41,000-implied and $1,576-actual is not a forecast disagreement; it is two different theories of what the asset is. That is why this bull/bear page reads differently from the equity entries in the series: nobody arguing about HIMS thinks it might be mispriced by 26x.

The regulatory layer: legislation stalls, ETFs wait on staking

Citi’s target cut named the tension directly: US crypto market-structure legislation — the CLARITY Act framework that would finally split SEC and CFTC jurisdiction — has slowed in the Senate, and with it the institutional-adoption timeline every 2025-era ETH model assumed. The second regulatory front is the staking question inside the ETF wrapper: the SEC’s fast-track listing regime has processed spot products while conspicuously stopping short of staking-enabled ones, leaving US ETH ETFs structurally yield-disadvantaged against direct holders — and against every offshore venue that passes staking rewards through. Add FinCEN’s stablecoin-issuer rules taking effect July 18, which formalise the compliance perimeter around Ethereum’s largest use case, and the regulatory picture is the push-pull in its purest form: the rails Ethereum settles are being legitimised faster than the asset that secures them.

The offshore contrast sharpens the point. Japan this week folded crypto into its securities law framework and cut the tax on crypto gains to a flat 20% — the kind of institutional-grade treatment US legislation keeps deferring — while the EU’s MiCA regime has already forced its authorisation cliff, with roughly 210 firms licensed and hundreds winding down. Jurisdictions are converging on treating Ethereum-based markets as regulated capital markets; the US, Ethereum’s deepest pool of ETF capital, remains the one still arguing about which agency owns the file. Every quarter that gap persists is a quarter the Citi thesis — regulatory drag suppressing the ETH price — stays live.

What happens next: three dates decide the floor

Prediction one: the $1,500 floor gets tested before Labor Day, and holds only if ETF outflows stay decelerating. The causal chain runs through the flow data — eight weeks of outflows ended in mid-July; if Ether funds flip to sustained inflow alongside Bitcoin’s, the mechanical seller disappears and $2,000 gets retaken; if outflows resume, Citi’s $1,198 is the printed magnet below.

Prediction two: the funding-crisis narrative resolves within Van Epps’s own 3-to-9-month window, one way or the other. Either the new institutional structures (Etherealize, Ethereum Institutional) formalise a development-funding backstop — likely with BitMine-style treasury capital involved — or a visible core-dev departure turns a governance subplot into a price event. Watch for an announcement before the November window closes.

Prediction three: the sell side re-marks toward the market, not away from it. Citi’s cut from $4,304 to $3,175 was the first institutional capitulation; expect Standard Chartered’s $7,500 to survive on paper while quietly acquiring conditions, because a bank abandons a headline target only after the narrative that justified it dies. The cleaner signal will be Kalshi’s sub-$750 contract: if that probability compresses below 10% while price recovers $2,000, the tail risk is closing and the bull/bear spread renormalises — the same market-versus-street convergence FinanceFeeds tracks across the Ethereum price prediction hub.

FAQ

What is the bull case for ETH in 2026?
Standard Chartered’s $7,500 end-2026 target is the institutional bull anchor — +376% from July’s $1,576 — built on tokenisation and stablecoin settlement growth. Fundstrat’s Tom Lee goes further, publicly arguing Ethereum “could be a $5 trillion network” within a few years, with a $12,000 base case.

What is the bear case for ETH?
Citi’s recessionary scenario at $1,198 (−24%) is the published bear floor, but Kalshi’s prediction market prices an 18% chance ETH trades below $750 in 2026 — a tail no sell-side desk has printed, driven by ETF outflows, stalled US legislation and Ethereum’s development-funding controversy.

Why is Ethereum down 46% in 2026?
Eight straight weeks of ETF outflows (Ether funds lagging Bitcoin’s recovery), Citi-cited weakness in on-chain activity, slow progress on the CLARITY Act market-structure bill, and a governance-funding controversy around core development have compounded into a repricing from the 2025 highs.

What is the Ethereum funding crisis?
Former Ethereum Foundation contributor Trent Van Epps warned of “a slow-burning funding crisis within the next 3-9 months,” citing ~$30 million in annual core-development needs. Tom Lee’s rebuttal: “zero chance of this ‘crisis’ happening for ETH.” Resolution of that dispute is one of the second half’s binary catalysts.

Who is buying ETH during the drawdown?
Corporate treasuries, led by BitMine — chaired by Tom Lee — which held 5.62 million ETH (~$9.7 billion) as of June 15 and added 76,881 ETH in one June week, while ETF investors were still net sellers of Ether funds.

Is the ETH price cheap at $1,576?
Relative to the published target set, unambiguously: the market trades 24% above only one scenario (Citi’s $1,198 recession case) and below everything else, including Citi’s own $3,175 base. Relative to realised fundamentals — ETF outflows, softer on-chain activity, unresolved development funding — the price is the market’s honest verdict. “Cheap” here is a bet that the flow reversal of mid-July persists.

Will ETH go back to $2,000?
$2,000 is the technical ceiling analysts call the “biggest obstacle” on the recovery path — and even Fundstrat’s private first-half model ($1,800–$2,000) treated it as the pullback zone. Sustained ETF inflows are the cleanest trigger; resumed outflows point the market back toward Citi’s $1,198 scenario first. Either way, expect the $2,000 fight to resolve before the September Fed meeting — the macro date every crypto flow model now keys on.

Salesforce stock has plunged by more than 50% from its December 2024 peak as concerns about its growth outlook have intensified. Its market capitalization has fallen from more than $347 billion to about $136 billion, and the selloff could continue as investors remain concerned about the company’s strategy and long-term growth prospects.

Salesforce stock has dropped amid SaaSpocalypse fears

CRM stock has been in a steep decline over the past few years as concerns about its growth have escalated. Recently, the stock has dropped because of the rising SaaSpocalypse fears. 

SaaSpocalypse is a relatively new term referring to fears that AI agents will replace traditional software and the “per seat” pricing model. A good example of this is what Starbucks is doing. 

According to Bloomberg, the company is now building its own AI-assisted replacement for a Microsoft system that tracks inventory and an IBM solution that manages maintenance. It aims to save the $400 million it spends annually on software.

The fears in the software industry escalated this week after IBM published its financial results. IBM said that its business slowed as customers reprioritized their capital expenditure, redirecting it towards hardware purchases like servers and memory.

Salesforce’s organic growth has been slowing for a while. The most recent results showed that its revenue rose by 13% in the first quarter. While this growth is solid for a company that has been in business for years, it was not organic. Its $11.1 billion revenue included $444 million from Informatica, a company it acquired in a $8 billion deal.

The company has been one of the most acquisitive ones in the US. It has spent billions of dollars acquiring firms like Own Company, Fin, Bluebirds, Tableau, and Slack.

Analysts expect that Salesforce’s business will remain under pressure in the coming months. The average estimate is that its revenue jumped by 10% in the last quarter to $11.32 billion. Its annual revenue is expected to be $46.1 billion, followed by $50.4 billion next year. 

Bargain or a value trap?

At face value, there are signs that Salesforce stock has become a bargain. For one, its Non-GAAP forward price-to-earnings ratio has dropped to 11.8, well below the sector median of 24. Its five-year average stands at 24. 

Similarly, the forward PEG ratio stands at 0.73, also lower than other companies in the tech industry. The challenge, however, is that these valuation metrics include the extra funds made from its Informatica buyout. 

As a result, the company will need more growth catalysts over time. One of this catalysts will be its Agentforce and data segments, whose annual recurring revenue soared to $3.4 billion, a 200% jump. It has deployed over 3.8 billion Agentic Work Units (AWU) across Agentforce and Slack.

READ MORE: Salesforce stock falls after KeyBanc downgrade on AI growth concerns

CRM stock technical analysis

Salesforce stock chart | Source: TradingView

The weekly chart shows that the CRM share price has slumped in the past few years, moving from a record high of $367 to a low of $146. It remains below the 50-week Exponential Moving Average (EMA).

The stock has also remained below the Supertrend indicator and the 78.6% Fibonacci Retracement level. 

Therefore, the stock will likely remain under pressure in the near term. In this, it may drop and retest the year-to-date low of $146. 

In the long-term, however, the stock will likely bounce back as investors buy the dip in software stocks. 

The post Salesforce stock has slumped amid SaaSpocalypse concerns: what next? appeared first on Invezz