美股重點新聞

財韭
想問下你點睇mu 財報

主要關注記憶體報價，MU近排睇sell side偏負面，主要擔心DRAM 25H1會開始走down cycle，我已經冇咩有別於市場共識嘅諗法，AI HW主要得返NVDA TSM AVGO三隻了

好既 唔該

咁就慘，113買入，都唔知返唔返到家鄉

Mu個追唔追好

miss咗個機會
我自己就唔追了，估值冇話特別平

After Mark Zuckerberg showcased Meta's AI future at Meta Connect, Jefferies analyst Brent Thill says the keynote leaves the firm "more bullish" on Meta's consumer and enterprise AI opportunities. On the consumer, Jefferies is bullish on Meta.AI's impressive traction, with about 500M monthly actives, and sees potential for new image, translation and voice capabilities to boost core engagement. On the enterprise side, Llama is "becoming a serious contender" and the firm sees momentum building with its new multimodal capabilities, calling Meta "one of our top AI picks." The firm keeps a Buy rating and $600 price target on Meta shares.

財巴rklb會唔會追

10/10前賣啲TSLA，之後先入返

Zuckerberg comments on AI scaling and the thirst for continually training models:

"So that leads to this dynamic where Llama 3 you know we could train on you know 10 to 20,000 GPUs, Llama 4 we could train on you know more more than 100,000 GPUs, Llama 5 we can plan to scale even further and there's just an interesting question of how far that goes.

It's totally possible that at some point we just like hit a limit and just like previous systems there's an asymptote and it doesn't keep on growing but it's also possible that that limit is not going to happen anytime soon and that we're going to be able to keep on just building more clusters and generating more you know synthetic data to train the systems and that they're just going to keep on getting more and more useful for people for quite a while to come. It's a really big and high stakes question I think for for the company is because we're basically making these bets on how much infrastructure to build out for the future and this is like hundreds of billions of dollars of infrastructure so like I'm clearly betting that this is going to keep scaling for a while.

It's one of the big questions I think in the field because it is possible that it doesn't. You know that obviously would lead to a very different world where it's I mean I'm sure people still figure it out eventually just need to make some new fundamental improvements to the architecture in some way but that might be a somewhat longer trajectory for okay maybe you know the the kind of fundamental AI advances slow down for a bit and we just take some time to build new products around this or it could be the case, and that's what I'm betting on, that the fundamental AI will just continue advancing for quite a while and that we're going to get both a new set of products that are just really compelling in all these ways and that the technology landscape and what's possible will just continue being dynamic over like a 20-year
period and that's probably what I'd guess is going to happen but it I think it's one of the bigger questions in the industry."

Q4: revenue /EPS of $7.75b/36.5%/$1.18 vs. the Street at $7.65b/34.6%/$1.11

Q1 Guide: revenue/gross margins/EPS of $8.7b/39.5%/$1.74 vs. the Street $8.31b/37.6%/$1.5…

Large upside to rev and GM upside of 180bps vs street. Mgmt sounded very bullish on the call saying: “We expect the HBM total available market (TAM) to grow from approximately $4 billion in calendar 2023 to over $25 billion in calendar 2025." JPM: The beat is clearly being driven by healthy server demand, namely by AI server (though they’re also seeing replacement of older traditional servers, multiple of which can be replaced with only one, freeing up datacenter rack space), where they are ramping 12-Hi and diversifying customers.” They also said mid 30s % of revs for capex and reiterated HBM being 6% of CY25 bits and accretive to DRAM GM and Total GM. Mizuho: “MU sees much better pricing and margins from (1) strong and improving HBM yields, (2) stronger demand and pricing from cloud and server data center demand, (3) less bad PC and smartphone end markets and outlook into 2025 with acceleration expected in PC in 2H25.” AI-driven enterprise SSD (eSSD) now represents 50% of the NAND mid in the Q, up from only 20% two quarters ago.

Bulls will say commentary and pricing/gm upside point to DRAM cycle bottoming in this very short winter and HBMs will lead us into the glory days of AI summer very quickly. Annualizing Nov EPS upside gets you close to $10 on FY25, and my sense is that buyside might move even higher than $10. FY26 #s will now be around $14 (with the most bullish still at $15+). Bulls will start talking like they did for most of 2023: this DRAM cycle will be unlike any other so MU deserves a 13x+ multiple at $14 = $180+. Frankly its hard to argue with bulls when we just had a quarter when every DRAM datapt was negative and yet MU prints a q like this. So what happens when the datapoints get better? Bears have an uphill climb here and valuation framework shifts quickly from BV to P/E. Feels like game-on again in memory land.

Bernstein says MU “is building inventory now and expects better demand in 2HFY25 to bring inventory back to a healthy level exiting FY25. This implies Micron’s capacity is below demand, supporting its repeated comments on supply discipline. Demand is expected to accelerate into 2HFY25 to propel both volume and price higher.”

JPM says the stock should continue to outperform through 2024 and into 2025 as the market continues to discount improving revenue, margin and earnings power on top of "strong positive" earnings revisions. The firm believes accelerating high bandwidth memory and artificial intelligence mix along with favorable industry dynamics to continue into 2025 for Micron.

Rosenblatt who has a $250 PT and $18 in FY26 EPS brings his big bull energy: “Near-term smartphone/PC inventory weakness was trumped by big data center strength (AI and gen purpose), HBM mix, high-density DIMMS, and high-end SSDs. For months now, the Street has been lulled into the notion of an emerging vanilla down cycle (inventories in smartphones/PCs) while studiously avoiding the new, once-in-a-generation HBM dynamic and its disruptive impact on supply/demand on the 3-to-1 wafer trade ratio compared to commodity DRAM. Overall, DRAM S/D is expected to continue to tighten through calendar 2025, HBM is sold out, smartphone/PC inventories normalize into the spring, and interestingly, NAND is also exhibiting healthy S/D. During the Q&A last night, analysts seemed puzzled by the Bullish overall 2025 framework and skeptical of Micron's view of capturing over 20% of the 2025 DRAM share (TAM ~$25B), which, by implication, questions the visibility driven by negotiated orders (units and prices set).”

MS who just lowered #s and dbl downgraded Hynix said MU is maintaining their view that they will do several billion next year and that they will achieve their level of DRAM market share - mid 20s - in HBM, but the "market forecast seems too high," the analyst contends. Micron management is executing well, but the stock "remains expensive," according to Morgan Stanley, which says it sees better risk rewards elsewhere in AI and memory.

Citi says hile there is excess DRAM inventory in the PC and handset end markets, this should finish by spring and is being offset by strength from the data center end market Citi remains bullish on the DRAM recovery given supply/demand dynamics for 2025.

Google is working on artificial intelligence software that resembles the human ability to reason, similar to OpenAI’s o1, marking a new front in the rivalry between the tech giant and the fast-growing startup.

In recent months, multiple teams at Alphabet Inc.’s Google have been making progress on AI reasoning software, according to people with knowledge of the matter, who asked not to be identified because the information is private. Such software programs are more adept at solving multistep problems in fields such as math and computer programming