• The week that ended January 17th was a favorable one for stocks, bouncing back from a sell-off the week before. The Nasdaq Composite Index closed the week up 2%, the Standard and Poor’s 500, up even better, 3%, and the TL20 group of twenty great companies whose shares are worth considering was up the same as the Nasdaq, 2%. 

• What did well: quantum computing names continue to soar, with Rigetti Computing up 62%, Quantum Corporation up 56%, Quantum Computing up 51%, and IonQ up 40%. That’s on top of some amazing gains in several of these, with Rigetti in the last twelve months up 873%. 

• Wolfspeed, the leader in silicon carbide chips, up 28% percent for the week on a feeling, a hope, that maybe the SIC market is coming back with some return to health of some the EV automotive market. 

It’s been a long time since I can remember anyone making a case for investing in Sony with any great enthusiasm, so, I was intrigued when I noticed on Thursday Bernstein’s David Dai initiating coverage of the stock with an Outperform rating, and a price target of ¥3,900 Japanese Yen, which would be about twenty-two percent from Thursday’s price on the Tokyo exchange of ¥3,100.

The shares just edged out the Tokyo Topix index in the past year, and the three-year return was miserable, a little under five percent versus thirteen percent for the index.

But, now, Dai sees better times ahead thanks to profit growing faster than revenue for the first time in a while. “In contrast with an indifferent 2022-2024, we expect a Sony more focused on efficiency to deliver earnings surprises,” writes Dai.

Taiwan Semiconductor Manufacturing, the world’s biggest contract chip maker, is the main supplier to Nvidia, and, when it comes to spreading the ebullience about all things artificial intelligence, Taiwan Semi is Nvidia’s partner in crime.

Almost every quarter, the remarks from TSM’s CEO, C.C. Wei, are the soundbite that backs up Nvidia’s CEO, Jensen Huang. Back in October, Wei spoke of “insane” demand for AI chips. For the latest earnings report, on Thursday morning, Wei didn’t need to use any hyperbole: The number speak for themselves.

AI-chip-related revenue was a “mid-teens” percentage of all of Taiwan Semi’s revenue in 2024, noted Wei, as that revenue tripled, year over year, driving Taiwan Semi’s total revenue for the year up a very healthy thirty percent. But, this year, he said, AI-related revenue will double again. Clearly, that means AI-related revenue is going to be even higher this year as a percentage of Taiwan Semi’s business.

I noted Tuesday that expectations for Microsoft’s “co-pilots” have met with disappointment, as folks such as Morgan Stanley’s Keith Weiss have had to revise down their revenue forecasts amidst lackluster demand for the stuff.

But hope springs eternal, and the enthusiasm that once attended co-pilots has now attached to “AI agents.” Some observers are starting to dream big, big numbers, just as they once did for co-pilots.

Meta Platforms is Exhibit A in that respect.

Meta is probably one of the companies that has had a better reception on the Street as far as AI usage, and the company is getting the benefit of the doubt it will make a lot of money off of AI agents.

As you may have heard, the White House on Monday issued a decree by outgoing president Biden placing restrictions on which countries can have access to U.S. artificial intelligence chips, which it claims was necessary to “ensure” that “U.S. technology undergirds global AI use and that adversaries cannot easily abuse advanced AI.” 

The announcement is being referred to as the “AI diffusion rule.

There’s been very little response, probably for exactly the reason I mentioned on Friday, which is that the Street sees little prospect that such a measure will be upheld by the incoming administration of Donald Trump. 

The week ending January 10th was a pretty miserable trading week, the Nasdaq Composite Index falling 3%, the Standard & Poor’s 500 index down 2%, and the TL20 group of stocks to consider somewhere in between the two. 

The week's major decliners included Advanced Micro Devices, and, right behind AMD, Nvidia, whose CEO and co-founder, Jensen Huang, gave the keynote address at the annual Consumer Electronics Show in Las Vegas but failed to impress.

It was a poor showing across the board for Mega Tech, with major decliners including Dell Technologies, down 8%, Oracle, down 7%, and Adobe, down 6%. 

The semiconductor group of stocks has been dominated for quite a while by the artificial intelligence trade, with Nvidia and Broadcom seemingly the only things to which investors would commit themselves.

With the arrival of the Consumer Electronics Show this week, there’s a tiny bit of hope in the air that 2025 will be a better time for the humble chip makers serving the numerous parts of the industrial and automotive markets — the boring stuff, in other words.

Jefferies’s European semiconductor analyst Janardan Menon reiterates a view on Infineon as his “top pick” after the company said at the show that “they are not seeing any incremental weakening in auto-related demand, similar to the messaging during our Jefferies-hosted auto roadshow in early December.”

UPDATE: I’ve updated the details about HSBC’s note on AMD, and added more about Deutsche Bank’s software ratings.

It’s one thing to question a company’s goals, quite another thing to write off the company.

Adobe and Advanced Micro Devices are two names getting the axe on Wednesday, as analysts conclude that they’re just not going to benefit from the ongoing AI spending wave. 

In the case of Adobe, Deutsche Bank’s Brad Zelnick cut his rating on the shares to Hold from Buy, and slashed his price target to $475 from $600, as part of a broad review of software names for 2025.

As noted last month, Adobe’s lack of a financial lift from AI has been testing investors’ patience. Zelnick writes that he is waiting to see “more tangible evidence of AI monetization,” but the tone of his write up gives you the sense he’s not optimistic. 

Apple stock has the lowest growth rate of the “Magnificent Seven” tech giants, but the highest stock multiple, one of the reasons Craig Moffett of the eponymous MoffettNathanson research house on Tuesday cut his rating on the stock to Sell, and cut his price target to $188 from $202, which would be twenty-two percent below today’s $242.

Moffett had initiated coverage of the stock in August with a cautious stance, writing that the shares were already then pricing in a very big payoff for Apple’s version of artificial intelligence on the iPhone, “Apple Intelligence.” Since that time, in the past six months, Apple stock has slightly outpaced the Nasdaq, up seven percent, and Moffett finds it illogical that the stock has gained while “there has actually been a steady drumbeat of bad news.”

It’s just the second day of a new year, but it’s back to business as the Street recommends for and against some of the top names.

First, the positives. Nvidia is a keeper, according to Merrill Lynch’s Vivek Arya, who reiterates a Buy rating, and his designation of “sector top pick” for semiconductors. The main “catalyst” for the stock in the near-term is CEO Jensen Huang’s customer press conference at the Consumer Electronics Show, starting next week in Las Vegas. Arya predicts the important stuff will be about robotics, no surprise; updates on the state of the “Blackwell” chips, after the headlines were all about delays last year; and maybe, but not for certain, Huang talking about Nvidia’s “entry into the AI PC,” meaning, a personal computer chip, long speculated

2024 was the year everyone’s broker, and maybe their friend’s friend’s broker, told them to get into Palantir stock.

Lots of names soared this year, and lots of names got more expensive, but Palantir stands out among the five hundred or so U.S.-listed stocks I follow. It is what I call a “six-time offender,” meaning, it got egregious this year by six measures.

Palantir, in case you didn’t know, is an artificial intelligence software maker, although, in truth, no one really knows what the company sells, given that Palantir keeps it shrouded in mystery. That includes those who own its shares and have seen their portfolio swell from holding it.

As you’ll notice in the accompanying table, Palantir’s stock has among the highest price-to-sales multiples of any stock, almost fifty-six times the next twelve months’ estimated revenue, and among the highest forward price-to-earnings ratios, 165 times. Those are strike one and two.

Watching the initial public offerings of tech companies is one of my favorite activities throughout the year. It’s a steady stream of different approaches to the idea of a tech company, some thrilling, some ridiculous, and all of them interesting. 

Because there’s no immediate pressure to buy these things — remember, historically, most IPOs trade down their first twelve months of trading — one can have an amusing time perusing them without feeling compelled to take action. 

My advice is to file these names away for potential action down the road. Some of them could be the next great companies, although some of them were already among the greatest stock trades this year.

Overall, the IPO market came back a little bit from the doldrums of 2023, when there were so few offerings that I stopped counting. This year, there were thirty-four U.S.-based listings, raising a total of $5.3 billion. The number of IPOs was split evenly between the first half and the latter half of the year. (I wrote about the first half back in June.)

• Welcome to almost the last podcast of the year. The week ending December 27th was generally a good one for stocks with the Nasdaq Composite Index up half a point, the S&P 500 up 0.7%.

• We saw a number of the biggest names rallying into this close, including Broadcom, up almost 10%, Advanced Micro Devices, up 5%, Intel, up 4%, Alibaba Group, up 3%, Qualcomm up almost 3%.

• That adds to some of the biggest gainers of the year, including Broadcom up a 116%. Among major losers for the year are Intel, down 60%, Samsung down 31%, Adobe down 25%, and Advanced Micro down 15% for the year.

• Amongst the biggest gainers last week were the quantum computing stocks, with D-Wave Quantum up 54%, Quantum Corp up 54%, and Rigetti Computing up 82%.

“Right now, we have zero revenue inside the data center […] We think there’s an intersect point, and that’s what we’re focused on.”

The world needs alternatives to Nvidia in the training and inference of artificial intelligence, but it is hard to see it coming from Masayoshi Son, the chairman of Japan’s SoftBank Group conglomerate. 

I’ve written in past that Son is a bit of a charlatan, inclined to woo the market with ridiculous promises, such as his 2020 “Vision Fund” excesses, or his 2018 insistence that “artificial general intelligence” had arrived (it still hasn’t.)

Perhaps I should be using the word vain now, as in, consumed with a grandiose sense of one’s own importance. 

Monday morning, Bloomberg’s Min Jeong Lee and Ian King report on Son’s latest escapade, to build a chip that will compete with Nvidia in AI.

Shares of Micron Technology are down fifteen percent in early trading Thursday morning, at $88.12, the response to a disappointing revenue forecast for the current quarter offered Wednesday night. The fall was doubtless made worse because the report followed Wednesday nearly three-percent sell-off of the broader market following “hawkish” states from the U.S. Federal Reserve Board.

This is a blip, however, as reasons for the downbeat outlook are hardly epic and all too familiar: There continues to be a build-up of DRAM and NAND inventories among customers of Micron’s in the smartphone and personal computer market, and the automotive and industrial equipment markets for chips continue to grind through a long period of recovery before their appetite for new chip purchases turns around.

“What we're essentially doing is we're attempting to repeat a playbook that we executed with our first large customer at these others.”

Semiconductors stocks will look a lot next year like this year, at least for the first six months, according to Merrill Lynch’s Vivek Arya, who on Monday offered up the firm’s 2025 outlook on chips.

Arya writes that the year is set to be one of “two varying trends,” with the first half of the year all about cloud computing and artificial intelligence, and the second half producing a rebound in the more mundane business of chips that are “cyclical,” namely, industrial and automotive applications.

That focus on AI will probably yield a lot of gains for existing winners, especially as Nvidia is Arya’s biggest proposed gainer, with a potential upside in 2025 of forty-two percent.

I mentioned on Monday that it’s worth considering alternatives to Nvidia in artificial intelligence chips, the most prominent two options being Marvell Technology and Broadcom.

Shares of Broadcom soared on Friday as investors got a new round of enthusiasm for that AI opportunity. The stock closed up twenty-four percent on Friday, for a gain of a hundred and one percent this year.

The company’s earnings report Thursday evening was good even if it didn’t present stunning upside. Fourth-quarter revenue was actually slightly below consensus, despite rising forty-four percent, and the forecast for this quarter’s revenue was merely in line with Street expectations.

But the stunning part was the two hundred and twenty percent growth in revenue from AI chips for the full year, to over twelve billion dollars.

Some recent developments suggest the beleaguered software group of stocks may be coming back into favor, as I noted a couple weeks ago following positive responses to both Snowflake and Elastic. 

The interest in software is showing up again this week, with initiations and upgrades and downgrades on a raft of names.

NOT EVERYONE’S A WINNER

First, the bad news: Some software stocks may be even less desirable in a rebounding software market. 

Wednesday, JP Morgan’s Pinjalim Bora rejiggered ratings on three software stocks, C3.ai, which just reported positively on Monday; PagerDuty, which sold off last month on its rather mixed earnings report; and Sprinkler, which also had a downbeat report last week. 

In earnings news Thursday, shares of Adobe are sinking while Ciena is surging, as the payoff from artificial intelligence is playing out very differently for the two.

Both stories, however, require a leap of faith.

Adobe shares are down over thirteen percent after its forecast offered Wednesday evening for this quarter came in below expectations: Adobe sees revenue rising eight to ten percent, the Street was looking for eleven percent. 

The big picture is that the business is not speeding up from AI as investors had hoped it would, something that had been dismissed in September by the bulls as merely a blip but now cannot be ignored. 

Tuesday’s assessment of Oracle was wild: numerous bullish analysts raised their expectations on the stock following the earnings report Monday evening, even though they cut their estimates for this year and for next year by a few hundred million dollars.
The analysts’ enthusiasm is in contrast to the general investor disappointment, as the stock sold off by seven percent Tuesday.

The disconnect is that the sell-side analysts believe that, no matter what’s happening at the moment, Oracle is a company reborn, with higher revenue growth ahead. Investors are saying, Show me.

The week that ended December 6th was a good one for stocks. The Nasdaq Composite Index rising 3%, the Standard & Poor’s 500 up 1%, and the TL20 group of stocks worth considering up 5% for the week.

It was a good performance from mega cap technology companies. The top gainer for the week was Tesla, rising 13% amidst enthusiasm for the potential payoff to Elon Musk and company from the incoming administration of U.S. president-elect Donald Trump. Second highest, 11% for Broadcom. Taiwan Semiconductor up 10%. Salesforce up 10%. Amazon up 9%. Meta up 9%. ServiceNow, a TL20 company up 7%, and Arista Networks, up 7%.

As I noted in a think piece Monday, Broadcom and Marvell Technologies could be a good alternative to Nvidia for investors looking for AI exposure. And both companies are also an increasing threat to Advanced Micro Devices, according to Merrill Lynch’s Vivek Arya, who cut his rating on AMD stock Monday to Neutral from Buy.

Arya warns that AMD is behind Nvidia in terms of the total GPU buffet, and now Broadcom and Marvell’s work on custom chips for Amazon and others could limit AMD’s AI chip sales.

“AMD's pipeline remains 1yr+ behind NVDA's (which is accelerating) and lacks a competitive networking (switching, optics) portfolio,” writes Arya.

It can be treacherous to bet against the shares of Nvidia, now the world’s greatest chip maker.

For example, the GraniteShares 2x Short NVDA Daily ETF, run by the New York-based ETF company GraniteShares, which promises to give you two-times the inverse of whatever NVDA stock does, has lost ninety-four percent of its value this year as Nvidia stock soared a hundred and eighty percent.

You would have done much better betting on GraniteShares’s other speculative instrument, the GraniteShares 3x Long Nvidia Daily ETP, which is up almost eight hundred percent. (Although, read up on why the ETP is not quite what it appears.)

If, however, you are concerned about Nvidia’s rising stock valuation, it’s worth considering alternatives.

It turns out that betting on artificial intelligence is a bet that can flip wildly from quarter to quarter.

Both Dell Technologies and Hewlett Packard Enterprise had high expectations heading into this earnings season. But Dell’s report last week left investors somewhat disappointed with the level of AI server sales, while Hewlett on Friday is rising by over nine percent after its report Thursday evening prompted a newfound faith in its AI abilities.

The jump in the stock is especially striking given that CEO Antonio Neri had to hold off on forecast revenue for the new year because the company is in the midst of buying networking firm Juniper Networks.

The gains may be short-lived, as the AI business is a precarious one, with lower margins for both Dell and Hewlett as the technological innovation is increasingly dominated by their supplier who is also their competitor, Nvidia.

We are once again in the dregs of earnings season, as some smaller, more obscure names turn in results.

We heard from Credo Technologies on Monday and Pure Storage on Tuesday, who each had news of stunning progress.

And Wednesday night, we got a glimmer of hope from Verint Systems, the cloud-based “customer-engagement” software maker, and a disappointing forecast from Synopsys, the chip intellectual property vendor that hasn’t yet shown the kind of interesting prospects I discussed with CEO Sassine Ghazi in October.

Verint is an example of how trading the earnings bounce can be tricky.

Update: Pure Storage and Zscaler news and views.

Bandwidth chip maker Credo Technology is a small-cap that has been riding the rise of artificial intelligence networking all year long, with numerous analysts singing its praises (see the sidebar.) It tends to take wild swings on its earnings reports, and Tuesday is a big, big up day for the stock, rising forty-five percent.

The surprise in Monday evening’s earnings was a very large one: the revenue outlook for this quarter, $115 million to $125 million, came in forty percent above consensus. That’s quite an achievement considering that all of the analysts covering the stock have a Buy rating on it.

Credo has products that overlap, to an extent, with what Astera Labs does, and Astera has been an absolute darling this year, but so has Credo, up two hundred and seventy percent versus Astera’s gain of a hundred percent since its March IPO.

Shares of Intel are up almost a point on Monday after the company announced that CEO Pat Gelsinger, who took the job almost four years ago, is retiring, to be succeeded by a team of CFO David Zinsner and current head of product Michelle Johnston Holthaus as co-CEOs as the company searches for a replacement.

It’s hard to see this moment as other than negative, and, in fact, Bloomberg’s Ian King, Liana Baker, and Ryan Gouldreport that Gelsinger report what I suspect, namely, that Gelsinger “was forced out after the board lost confidence in his plans to turn around the iconic chipmaker,” citing multiple unnamed sources.

Gelsinger was “given the option to retire or be removed,” they relate.

Sure: If things were going great, Gelsinger would not be retiring now, just shortly before all the work he’s done is supposed to return Intel to preeminence in chip manufacturing. It’s also questionable that he would retire without having already begun a selection process for his replacement.