Mizuho on Semis pullback: 'we see this as a buying opportunity with $Micron Tech(MU.US), $Lumentum(LITE.US), $Credo Tech(CRDO.US), $Sandisk(SNDK.US), $Broadcom(AVGO.US), $Dell Tech(DELL.US), and $Lam Research(LRCX.US) remaining well-positioned'.
"We look at the AI semis space after reports of Anthropic and OpenAI CEOs calling for AI development slowdown post Dario Amodei's essay, with the push for guardrails on frontier model adoption. Key points: 1) Anthropic and OpenAI seeking to potentially slow down AI development citing safety concerns, with pushback from US Chair David Sacks challenging their altruistic reasoning, 2) we believe imposing RSI limits combined with a lack of global AI consensus or coordination implies potential AI pacing ineffectiveness, as it risks ceding share to efficient Chinese competition (Ex-2) and, 3) Key to watch - No change to AI DC Capex/GW compute spend from AI labs and CSPs, while AI adoption, while early, remains positive for the AI Semis beneficiaries. With the SOX down ~6% this morning, we see this as a buying opportunity with MU, LITE, CRDO, SNDK, AVGO, DELL, and LRCX remaining well-positioned."
Share your thoughtsWhat's on your mind?
0 / 2,000
- A
GOOD MORNING & WELCOME TO FOMC DECISION WEEK.
In addition to a potential rate hike, the fears over the weekend caused by Dario’s essay have led to a long software/short semis premarket. Nvidia, Marvell, Sandisk, AMD, Nebius, Micron, CoreWeave, Intel are all red from 3-7%. Palantir, Atlassian, Crowdstrike, Service Now, Adobe, Intuit are all green from 2-5%. Personally, I think the market is having a knee jerk reaction to the headlines over the weekend. In the past 24 hours, Anthropic themselves have signed multiple new compute deals (one with Rumble for $14B yesterday) and are preparing for their IPO by officially picking the Nasdaq as their exchange. In the face of all the “pace the frontier” slowdown news, the company creating that news is not slowing down their compute obligations.So why is the market reacting like this? Well, the entire AI trade is rooted in capex. Until there is more clarity on what it means to slow down model training and recursive learning, the market is assuming that means there is a closer end to the growth in capex than expected which obviously is bad for the AI names. Great for software as more money gets allocated to the application layer and great for hyperscalers since they can generate more FCF which is why those names are green premarket.I think this fear could flip within days if not within hours, but if you match this fear with a rate hike that might be priced in but still is not good for high-beta growth (which is what the AI trade is composed of) then you start to see how the market can make sense of taking these names down further.Jensen speaks at the All-In Summit today, he’s likely either going to completely reject Sam/Dario’s fear-inducing message by giving a more optimistic approach to think about AI or agree with it and still make the claim that compute obligations, even if not needed to train new models as fast as possible, are not slowing down. Going to be a big week. $NVIDIA(NVDA.US) $Palantir Tech(PLTR.US) $Micron Tech(MU.US) $Marvell Tech(MRVL.US) $Sandisk(SNDK.US)Source: amit
- A
$Broadcom(AVGO.US)
Context: I like Broadcom because it sits right at the heart of AI infrastructure. While most people focus on GPUs, companies like Google and Meta work with Broadcom on custom AI chips and high-speed networking. Its software business also brings in steady cash flow, giving the company a strong foundation beyond AI.
My trade: My average cost is $355.172, and with AVGO around 362, I’m sitting on a small gain. I’m holding my shares today and not adding more. With strong demand for custom AI chips, networking, and solid software cash flow, I’m comfortable letting the position run while keeping my risk under control.
Takeaway: Broadcom has strong exposure to AI through both custom chips and networking. In your semiconductor portfolio, do you prefer GPU companies or custom ASIC players like Broadcom?
Trade Showcase: Trade, Show & Earn Rewards! - A
Google recovers its costs on TPUs in 1-year, while its overall payback period for AI servers is 2-years, media report, citing Google Cloud CEO Thomas Kurian, and adding that Google TPUs have found new customers across financial, biotech, government and scientific sectors in addition to AI labs. TPU strength is a boon to Broadcom and MediaTek, which both see their competition for Google’s chip design business as big enough for everyone to keep growing, even if new design rivals enter the market. $Broadcom(AVGO.US) $Alphabet(GOOGL.US) $NVIDIA(NVDA.US) #mediatek #semiconductors
Source: Dan Nystedt
- N
Context
Broadcom Inc. ($Broadcom(AVGO.US)) recently posted a blowout Q3 2026 earnings report, with overall revenue surging 86% year-over-year to $29.59 billion. The explosive growth was led by custom AI semiconductors, which skyrocketed 221% to $16.7 billion. Broadcom also vastly raised its long-term guidance, forecasting $115 billion in AI revenue for fiscal 2027 and a staggering $230 billion by 2028. However, the stock experienced a 13.74% pullback over the last month as Wall Street focused on a slight Q4 revenue guidance miss ($34.8 billion vs. $35.1 billion expected) and minor operating margin compression from shifting to full-rack AI systems.
My trade
-Strategy: Scaled in long via equity during the post-earnings dip.
-Entry Target: $360.00 – $362.00 (accumulating right at the current $361.99 support level).
-Stop Loss: $335.00 (just below the recent support floor).
-Price Target: $460.00 (in line with conservative analyst upgrades).
Takeaway
The market is missing the forest for the trees by punishing a minor Q4 guidance discrepancy. Broadcom’s underlying fundamentals are incredibly robust: it holds a highly visible custom XPU pipeline with mega-cap clients (like Alphabet, Meta, and OpenAI) alongside an enterprise software cushion from VMware. With a forward PEG ratio hovering near an attractive 0.4, this short-term correction offers a premier entry point into the structural backbone of global AI infrastructure.
Trade Showcase: Trade, Show & Earn Rewards! - T
😍
Apple's Ternus Era Begins: How to Position for the Next Chapter?Apple has delivered a total return of nearly 3,000% since Tim Cook became CEO in 2011, significantly outperforming both the Nasdaq 100 and S&P 500.After an extraordinary 15-year run, can $Apple(AAPL.US) ...
- D
$Broadcom(AVGO.US)
Post 4 Episode 22
Context: One reason Broadcom interests me is that the next phase of AI may not be dominated entirely by off-the-shelf GPUs. Large hyperscalers increasingly want custom accelerators designed around their own workloads, power budgets and economics — and Broadcom sits in a strong position to help enable that transition.
My trade: I see AVGO as a way to participate in AI compute without having to predict which individual custom chip ultimately wins. Its role in designing and connecting these systems gives me exposure to the broader trend toward specialised silicon while complementing my more direct accelerator holdings.
Takeaway: Sometimes the better investment isn’t betting on one winning chip. It is owning the company that can benefit as multiple customers build their own.
@Captain's Treasure
Trade Showcase: Trade, Show & Earn Rewards! - D
Bearish, the consumers need to digest the price increases first.
Then the new CEO will have to convince the consumers the value in the pricing.
And subsequently convince the market and investors that there’s still room for $Apple(AAPL.US)to grow in the smartphone space.
Apple's Ternus Era Begins: How to Position for the Next Chapter?Apple has delivered a total return of nearly 3,000% since Tim Cook became CEO in 2011, significantly outperforming both the Nasdaq 100 and S&P 500.After an extraordinary 15-year run, can $Apple(AAPL.US) ...
- P
Folderable is going to be interesting but due to the high price it might be putting off many of the fanbase. Let’s see how it goes
Apple's Ternus Era Begins: How to Position for the Next Chapter?Apple has delivered a total return of nearly 3,000% since Tim Cook became CEO in 2011, significantly outperforming both the Nasdaq 100 and S&P 500.After an extraordinary 15-year run, can $Apple(AAPL.US) ...
- N
nice good
Apple's Ternus Era Begins: How to Position for the Next Chapter?Apple has delivered a total return of nearly 3,000% since Tim Cook became CEO in 2011, significantly outperforming both the Nasdaq 100 and S&P 500.After an extraordinary 15-year run, can $Apple(AAPL.US) ...
