A message from our friends at Stansberry Research (sponsor)
Dear Reader,
Let me ask you...
Why did Buffett step down as the CEO of Berkshire Hathaway now?
You might have your own theory...
You see, Buffett didn't step down in 2008, when the banks collapsed (though he was already 78 years old by that point).
And he didn't retire in 2020, when the pandemic brought chaos to the markets (again, he was 23 years past retirement age by then).
No...
He's chosen 2026 as the time to finally hand over the reins.
How come?
Well, after analyzing over 3,000 stocks and 40,000 data points, our institutional research – which all of the top ten money managers in the world pay thousands to access – suggests the real reason the Berkshire CEO stepped away is because of a shockwave that is about to pulse through the market.
Make no mistake...
Massive gains – and huge losses – are coming in the next six months.
And it seems like Buffett knew.
...and learn about the one specific stock you should consider buying because of what's about to happen.
By the way, it won't cost you a cent to find out which stock it is – it's all revealed (including the ticker) in this presentation.
Best wishes,
Rob Spivey
Research Director, Altimetry
P.S. In his first letter to Berkshire shareholders, Greg Abel just cited one of the same sources of Buffett's stock picking inspiration as I do in this presentation – you'll be surprised by what it is.
BONUS ARTICLE
Marvell Just Broke the Chip Gloom — Here’s What Traders Need to Know
Bullet Summary
Marvell surged roughly 11.4%–12% after forecasting fiscal 2028 revenue near $15 billion, ahead of the analyst average around $12.92 billion.
Fiscal Q4 revenue came in at $2.219 billion, up 22% year over year, with non-GAAP EPS at $0.80, both ahead of expectations.
Fiscal Q1 revenue guidance is about $2.4 billion ±5%, above estimates near $2.27 billion–$2.30 billion.
Full-year fiscal 2026 revenue reached a record $8.195 billion, up 42% year over year.
Data center remains the engine: Marvell said fiscal 2026 data-center revenue topped $6 billion, and its custom business doubled in fiscal 2026.
In a mixed chip tape, Marvell is acting as a “rare bright spot” while the semiconductor ETF SMH remains softer on the session and peers show more uneven follow-through.
The key trader question is no longer whether AI spending exists. It is whether Marvell can convert custom silicon, optics, and connectivity demand into sustained operating leverage and trend confirmation.
Macro Context: Why This Move Matters More Than a Simple Earnings Pop
Marvell’s move matters because it is happening in a part of the market that has recently become more selective.
The semiconductor trade is no longer being rewarded simply for saying “AI.” The market has started separating:
compute leaders with visible demand,
connectivity and infrastructure names with measurable backlog,
and everyone else who still needs to prove that AI demand becomes durable revenue rather than thematic enthusiasm.
That is why Marvell’s print landed differently. This was not a marginal beat. It was a message that hyperscaler AI spending is still flowing into the “plumbing layer” of the data center: custom silicon, interconnect, optics, and storage. Reuters reported Marvell expects fiscal 2028 revenue to grow close to 40% and approach $15 billion, with management explicitly tying that outlook to AI-driven data-center demand.
That forward number matters because it does two things at once.
First, it gives traders a multi-year growth marker in a sector where near-term demand debates have become more intense. Second, it helps reframe the chip sector conversation away from “Is AI spending peaking?” and toward “Which companies are still moving up the value chain as hyperscalers keep building?” Reuters separately reported Broadcom now sees AI chip revenue exceeding $100 billion by 2027, reinforcing that the custom silicon market is no longer theoretical.
So the real importance of Marvell’s rally is not the percentage move. It is the signal:
AI infrastructure demand is still broad enough to produce upside surprises outside the most crowded names.
Sector Breakdown: Why Marvell Is a Bright Spot While the Rest of the Group Is More Uneven
The semiconductor space is trading in layers now.
Beneficiaries
The current winners tend to be names exposed to:
custom AI silicon,
networking and optical interconnect,
data-center switching,
and cloud-scale infrastructure buildouts.
Marvell fits directly into that basket. MarketWatch noted that investor enthusiasm centered on Marvell’s role in AI-driven demand, including custom chip partnerships, optical DSP exposure, and storage/connectivity demand tied to hyperscaler capex. It also highlighted that Amazon, Microsoft, Alphabet, and Meta are expected to raise 2026 capex materially, with total spending around $650 billion.
Broadcom also belongs in that winner’s circle, and its own update this week reinforced the same message: custom silicon demand is scaling, OpenAI and Anthropic-related workloads are becoming real commercial opportunities, and hyperscaler infrastructure budgets remain large.
Under pressure
The parts of the chip market still struggling are the ones where:
end-market recovery is less visible,
non-AI demand remains uneven,
or the revenue bridge from “AI exposure” to actual earnings power is still too vague.
That helps explain why Marvell’s move stands out today. Even with MRVL sharply higher, the VanEck Semiconductor ETF, SMH, is softer on the session, while AMD is down and Nvidia is roughly flat. That divergence tells you this is not a broad-based relief rally in chips. It is a selective bid into a name that just improved its forward earnings narrative.
That distinction matters. When a stock rallies against a mixed sector tape, the market is not just rewarding beta. It is rewarding company-specific information.
Stock-Level Analysis: What Marvell Actually Reported
Start with the quarter.
Marvell reported fiscal Q4 revenue of $2.219 billion, up 22% year over year, with GAAP gross margin of 51.7%, non-GAAP gross margin of 59.0%, GAAP diluted EPS of $0.46, and non-GAAP diluted EPS of $0.80.
That was enough to beat expectations, but the real spark came from the guidance.
For fiscal Q1, Marvell forecast revenue of about $2.4 billion, plus or minus 5%, which came in ahead of estimates around $2.27 billion–$2.30 billion.
Then came the longer-range signal.
Reuters reported Marvell expects fiscal 2028 revenue to grow nearly 40% and approach $15 billion, well ahead of the consensus expectation of roughly $12.92 billion.
That is the number that changed the stock.
Why? Because markets reward long-duration guidance only when it is tied to a believable product set and a visible end-customer demand engine. Marvell appears to have both right now.
Full-year scale also matters
Marvell’s fiscal 2026 revenue was a record $8.195 billion, up 42% year over year. That is not the profile of a company merely getting an AI halo. That is the profile of a company already converting the infrastructure cycle into real scale.
Data center is now the center of gravity
On the earnings call, management said fiscal 2026 data-center revenue surpassed $6 billion, while the custom business doubled in fiscal 2026.
That is the line traders should keep in mind.
Marvell is no longer just a diversified semiconductor name with multiple end markets. It is increasingly a data-center infrastructure name whose valuation is being reset around AI connectivity, interconnect, and custom silicon.
Strategic additions deepen the thesis
Recent acquisitions add more context. Investors.com reported management expects the Celestial AI acquisition to generate $1 billion annually by fiscal 2029, while XConn Technologies is expected to add around $100 million in fiscal 2028 revenue.
Those numbers are not the near-term driver of today’s move, but they do reinforce the longer-term idea that Marvell is building a broader AI scale-up networking platform rather than simply riding one product cycle.
Why the Market Is Treating This as a “Rare Bright Spot”
There are three reasons.
1) Marvell improved the sector narrative
When a stock in a weak or mixed group beats and raises, it often gets a higher-quality re-rating than it would in a euphoric sector tape. The market is effectively saying:
“If the group is not getting a free pass, then the names that still work must really be working.”
That is what happened here.
2) The growth is tied to capex that still looks durable
Marvell’s outlook is not being driven by vague hopes around enterprise AI adoption. It is being driven by hyperscaler capital spending, custom silicon roadmaps, and networking demand that sits directly inside large-scale AI clusters. Reuters said Marvell’s president and COO tied this demand directly to the company’s data-center business.
3) The stock is not priced like a pure moonshot
MRVL currently trades around $75.68, with a market cap near $80.8 billion and a P/E around 32.7. That is not cheap, but it is also not in the same valuation bucket as some AI-adjacent peers. For comparison, Nvidia trades around 45.6x, Broadcom around 71.7x, and AMD around 78.3x on the current tape.
This matters because it gives Marvell room to be re-rated if the market decides its earnings bridge is becoming cleaner.
Technical Framework: How Traders Should Read This Move
This is where discipline matters.
A stock can have a strong quarter and still become a poor trade if the opening move is chased without confirmation. The protocol here is straightforward.
Gap behavior
This is a classic earnings-gap setup. The stock is repricing into a new zone because the market’s prior framework for fiscal 2028 revenue was too low. That means the first thing traders should watch is whether the gap holds.
A strong earnings gap typically behaves in one of two ways:
Gap-and-go: price holds early gains, consolidates above VWAP, and builds a higher intraday base.
Gap-and-fade: price spikes on the open, loses VWAP, and turns the news into liquidity for sellers.
If Marvell holds the gap midpoint and reclaims or defends VWAP after the first pullback, that is usually a sign institutions are accepting the new valuation range.
Moving averages and support/resistance
Because this is a sharp repricing event, immediate support becomes:
the gap-day midpoint,
then the gap-day low,
then the nearest pre-gap resistance zone, which often becomes first major support if the move is real.
Immediate resistance is less important on day one because the stock is still in price discovery. Once the first session settles, traders can use:
the intraday high,
the prior all-time or multi-month swing zone,
and the volume profile around the gap.
RSI and momentum
A move like this will likely push short-term RSI into elevated territory quickly. That alone is not a sell signal. In true earnings re-ratings, RSI can stay overbought longer than traders expect. The more important tell is whether momentum stays paired with volume retention and VWAP support.
Volume context
This is not a quiet drift higher. It is an event move. High volume is supposed to be there. What matters next is whether:
follow-through days still attract volume,
or whether the first surge exhausts the demand.
The best bright-spot rallies in semis usually show at least one additional confirmation session, not just a one-day explosion.
Scenario Modeling
Base Case: Re-rating, then consolidation
Trigger: Marvell holds most of the gap, but the broader chip sector stays mixed.
Sector impact: Custom silicon and AI connectivity names stay bid, while generic semiconductor beta remains uneven.
Stock implication: MRVL consolidates in a higher range, with traders using VWAP and the gap midpoint as the first validation zone.
This is the most likely near-term setup. The move was big enough that some digestion is normal. The key is whether digestion looks constructive rather than distributive.
Bull Case: Marvell becomes a leadership rotation within semis
Trigger: The stock gets follow-through, analysts keep revising estimates upward, and the market starts treating Marvell as a cleaner second-tier AI infrastructure leader rather than a niche connectivity play.
Sector impact: Strength broadens into other custom silicon and networking names. Broadcom-style custom silicon enthusiasm spills over.
Stock implication: MRVL keeps building higher highs and higher lows, with valuation expanding as traders price in a stronger fiscal 2027–2028 ramp. Reuters and analyst commentary already point in this direction, with upward revisions to longer-term sales and EPS expectations appearing immediately after the report.
Bear Case: One-day bright spot, then sector gravity takes over
Trigger: The semiconductor group remains weak, macro pressure returns, and Marvell loses VWAP and the gap midpoint quickly.
Sector impact: Traders treat the move as isolated and sell into strength rather than rotate capital into the group.
Stock implication: MRVL retraces a meaningful portion of the move and turns into a range trade rather than a trend.
This is the key risk. A good quarter does not immunize the stock from macro or sector de-risking if rates, tariffs, or growth fears return.
Active Trader Strategy Framework
Use a defined-risk process.
If MRVL holds above VWAP after the opening range, watch whether the gap midpoint becomes support. That is the first sign institutions are accepting the new range.
If the stock loses VWAP and cannot reclaim it by mid-session, assume the first move is being sold and wait for a cleaner reset.
If Marvell outperforms while SMH stays soft, that relative strength matters. It suggests the market is buying MRVL specifically, not just buying chip beta.
If Broadcom and other AI infrastructure names continue confirming demand strength, watch for sympathetic rotation into Marvell on pullbacks.
If the stock breaks the gap-day low, consider that invalidation of the initial re-rating thesis in the short term.
That is the framework.
Not prediction.
Not excitement.
Process.
Conclusion
Marvell’s move matters because it is not just an earnings reaction. It is a challenge to the current pessimism around parts of the chip sector.
The company delivered:
a beat,
strong near-term guidance,
and a long-range revenue target near $15 billion that materially exceeded expectations.
In a more selective semiconductor tape, that is enough to matter.
The bigger takeaway is this:
The market is still willing to reward semiconductor companies that can prove they sit in the critical layers of AI infrastructure. Marvell just made a strong case that it belongs in that group.
Watch the gap.
Watch VWAP.
Watch whether relative strength persists even if the broader semiconductor tape stays uneven.
If that confirmation holds, traders may be looking at more than a one-day bright spot. They may be looking at the early stages of a higher-quality re-rating in one of the few chip names that just improved the sector narrative.
Preparation beats prediction.
Editorial Disclaimer
This commentary is for informational and educational purposes only and does not constitute investment advice. All market strategies involve risk, and past performance is not indicative of future results. Readers should conduct their own analysis or consult a licensed financial professional before making investment decisions.