“There are a lot of things that feel like they're full. There are other companies that feel decidedly overvalued. And then there are some names that feel very reasonable, and some things that are just downright cheap.”

Stocks continue to hit new 52-week highs, and some are hitting new all-time highs, including Broadcom, Microsoft, SAP, and Nvidia.

The last of these, Nvidia, hit a new all-time Wednesday morning of $164.42, a return since it started trading in January of 1999 of 432,000 percent. Not bad.

As you can see in the accompanying chart, adjusted for splits, the past decade for Nvidia stock really stands out. It was 2015 that was the year that things really changed, as Nvidia started to become an AI stock. (The first sixteen years were not bad, though, a return of 181,327 percent.)

This new high is special because it moves Nvidia into a market capitalization of four trillion dollars, the first to make it that far, and well ahead of the next-closest, Microsoft, at $3.8 trillion.

I would imagine no one is particularly moved by this event, as it is not very different from $3.999 trillion, but Dan Ives of Wedbush Securities, one of the biggest fans of Nvidia, remarks today that it “speaks to the AI Revolution hitting its next stage of growth,” which just means that, yes, the AI trade is alive and well. Ives thinks Microsoft “will also hit the $4 trillion market cap club this summer and then over the next 18 months the focus will be on the $5 trillion club,” without making any predictions.

Shares of recently public, debt-laden artificial intelligence darling CoreWeave are down for a second day in a row after the company on Monday morning announced it will purchase data center provider Core Scientific for for nine billion dollars worth of stock, a deal that had been rumored for a month or more.

The acquisition has prompted two downgrades this morning, the main issues being large dilution for CoreWeave shareholders, and the transformation of CoreWeave from what had been an “asset-light” business — not owning too much stuff outright — to now asset-heavy, owning data centers.

CoreWeave is essentially buying its supplier, as Core Scientific makes up almost forty percent of the total data center “footprint” that CoreWeave is using to run GPU chips to process AI.

Last Wednesday brought word that small-cap software maker Datadog is going to have its shares added to the Standard & Poor’s 500 Index this Wednesday, July 9th.

The stock soared the next day, Thursday, by fifteen percent, which is typical of an S&P index announcement of this kind: people buy the news. The importance of index inclusion is obvious: fund managers who want to match S&P index performance tend to have to follow the new component by buying shares of it, which is an automatic lift.

In addition, there’s a PR element. Merrill Lynch analyst Koji Ikeda, on Monday articulated why announcements of this kind tend to be seen as favorable. “Now that it's going to be a part of the S&P 500, we believe investor awareness of this DevSecOps category leader will increase meaningfully,” writes Ikeda, referring to the software category into which Datadog is usually included, “DevSecOps.”

That makes sense, as DevSecOps is rather obscure, so index inclusion can tend to shine a light.

I was curious to know just how much index inclusion actually helps a stock, so I did a little research.

Adobe is one of the most interesting stock debates these days, given that it is a storied name, and given that the debate about its future hinges for the moment on what people think about artificial intelligence.

As I’ve noted before, the discussion is about whether Adobe can thrive, or even survive, as non-professional types — hobbyists, people who are amateurs at design, and lots of others who are not the “core audience” — are lured away by AI tools such as ChatGPT making their image for them.

Last month, Adobe’s quarterly report was deemed very “solid” by most analysts, but it hasn’t helped the shares much, with the stock down ten percent since then, at $377.36, a fifteen-percent decline so far this year.

On Wednesday, a skeptic turned more negative, as Omar Sheikh of Rothschild & Co.’s Redburn division cut his rating on the stock to Sell from Neutral, and slashed his price target to $280 from $420.

What is the Street doing during these summer months?

They’re driving around the country looking at artificial intelligence data centers for tea leaves.

Macquarie Research’s Paul Golding, an analyst covering payments technology, including crypto-currency, lead clients on a bus tour through Texas last week, meeting with crypto firms such as Hut 8, Riot, Galaxy Digital, and a twelve-year-old, privately held firm named Aligned Data Centers based in Danbury, Connecticut.

The connection of AI to crypto is that a lot of companies that were mining Bitcoin for years suddenly turned to AI to fill their data centers as mining demand cooled.

Hut 8 and Riot and Galaxy, and others such as Cleanspark, are selling their data center capacity to parties hungry for ever more racks of equipment to run GPU chips for AI.

That includes large deals by CoreWeave lately to rent capacity from Cleanspark, Applied Digital, and Core Scientific that made headlines.

Stocks are hitting new highs. The week ending June 27th saw several of the biggest names at new highs on Friday, including Broadcom, Cisco, Taiwan Semiconductor, Microsoft, Intuit, Uber.

It’s very broad-based, with a ton of other names were hitting new highs: Advanced Energy Industries, SK Hynix, Aeva Technologies, Amphenol, Booking Holdings, Celestica, DoorDash, Diebold, Flex, Jabil, Just Eat Takeaway, Cloudflare, Roblox, Sanmina, Rocket Lab, Spotify, Seagate, Zscaler, etc.

Reasons include the tariff pause put in place by U.S. president Donald Trump restored risk-on attitude, and there was a relatively benign earnings season, with just a couple of tariff-related warnings from Ciena, First Solar.

You can debate the virtues of Bitcoin, but it’s certainly been the case that the crypto-currency has been a better bet than the companies working on crypto, companies such as Argo Blockchain.

Back in February of 2022, I wrote a note about Argo and others that were in that group of “blockchain” exchanges, crypto mining facilities, data center providers, etc. Their stocks were already down at the time, but they’ve turned out in the intervening three-plus years.

While Bitcoin has more than doubled since February of ’22, to a recent $107,266 per coin, Argo’s shares in that time have lost ninety-eight percent of their value. Bitfarms, another hopeful, is down almost eighty percent in three years. Many have trailed the Nasdaq Composite’s fifty-five percent return, such as Cleanspark, up just twenty-eight percent. 

Just a couple have been good bets, including Coinbase, the largest crypto exchange, up eighty percent since 2022, and Galaxy Digital, the firm run by former Fortress Investment principal Mike Novogratz, whose shares are up seventy-six percent. 

Most of the best action of late has been in the crypto ETFs, not stocks.

However, hope springs eternal, and Circle Internet Group, which went public on June 5th in an offering lead by JP Morgan, Citigroup, and Goldman Sachs, has more than doubled to a recent $182.56. 

The stock received mix reviews Monday as several underwriters and other analysts initiated coverage of its shares. The dilemma for analysts is simple: Lots of potential, hugely expensive stock.

As I had previewed in this week’s podcast, there was a rising tide of good feeling about memory-chip maker Micron Technology heading into Wednesday evening’s report, and it was paid off with what is being described as a “stellar” or “blow-out” report.

It wasn’t enough for the shares, however, which are up forty-five percent this year, and which have more than doubled since hitting a fifty-two-week low in April. The stock sold off a point on Thursday to $126.

The issues, if there are any, are probably less profound than the opportunity.

The results for the fiscal third quarter were strong all around. The quarter’s revenue and profit were the strongest upside in five and four quarters, respectively, and the forecast for revenue and profit were both the strongest in five quarters.

In these summer weeks between earnings reports and tariff cliffs, the Street turns to its current obsessions, things such as artificial intelligence agents and stablecoins.

An emerging fixation of the Street is the notion that “agents,” AI programs that plug into databases and other existing programs, will help companies better monetize AI.

Merrill Lynch’s software analyst Brad Sills on Tuesday took a shot at sizing the market for agents, which he concludes could be a “compelling” $155 billion total addressable market (TAM) by 2030.

“To our knowledge, this would make our estimated 2030 TAM the Street/third-party research high, and by a significant margin,” he writes. Other figures he’s seen from firms such as Boston Consulting Group, in the neighborhood of $52 billion, “are definitely too low,” he avers.

Elon Musk on Monday introduced a few of Tesla’s robo-taxis to Austin, Texas, a maiden voyage that has been anticipated since last fall’s unveiling of the self-driving technology.

The rides were offered to select “influencers,” according to Reuters’s Norihiko Shirouzu and Abhirup Roy, and the whole experiment involves “a small trial with about 10 vehicles and front-seat riders acting as ‘safety monitors’,” they relate.

The response from the Street is rather muted, and it’s clear there are many more fans of Google’s Waymo service than of Musk’s company.

Ben Rose of the boutique Battleroad Research estimates that Waymo is already making a quarter of a billion dollars in annual revenue on millions of self-driving rides in multiple markets. Google is probably also losing hundreds of millions of dollars on the service, he estimates, which shows you the power of being able to subsidize such a project.

Welcome back from the U.S. Juneteenth Holiday weekend. 

Stocks continue to make gains through Friday's close, June 20th, with the Nasdaq Composite Index closing the week up fractionally. 

The Nasdaq is now up 2% for the month of June so far, the Standard and Poor’s 500 Index up 1%. The TL20 group of stocks to consider is up 6% for the month. 

Both indices are up slightly, year to date, building on May's big gains.

Welcome to a quiet summer Friday of a holiday week, when yours truly seeks to catch up on some missed coverage of stocks earlier in the week.

First stop, a large initiation of coverage on chip equipment stocks, which have been very mixed this year, the group roughly flat. As you can see in the table, most are pretty far from their fifty-two week high prices, so the group has been generally out of favor.

David Dai of Bernstein Research on Monday took a fresh look at the matter and came up with what I think is an intriguing contrarian view: Dump ASML, buy BE Semiconductor.

Dai appreciates the strengths of both companies (both of which are in the TL20 group of stocks to consider), but he’s focused on what happens to two great franchises as we move out in time.

He thinks ASML becomes less important than it is now, and BE more so. He starts ASML at Neutral and BE Semi at Outperform.

Sometimes the zeitgeist wins out over feelings about individual stocks.

Shares of chip-maker Marvell Technology have been a dog for a while now, down thirty-two percent this year at a recent $75.47, and up just three percent in twelve months, as quarterly results have failed to show the kind of upside that investors wanted.

But all is forgiven this week as the company Tuesday gathered analysts for an upbeat discussion of “AI infrastructure,” chips for running artificial intelligence.

The analysts came away delighted, and the zeitgeist of the AI-chip trade is now working in the shares’ favor: the stock is up Wednesday by seven percent, bringing gains for the week so far to almost twelve percent.

Marvell, you may recall, competes with Broadcom in helping the tech giants make custom chips for AI, known as “ASICs,” with Amazon being its most prominent customer. 

Well, that was quick: I interviewed Fastly CEO Todd Nightingale last month, and late yesterday it was announced that he’s departing to become chief operating officer of Arista Networks. In his stead, Fastly’s head of products, Kip Compton, is taking over as CEO. 

This often happens to reporters: They get played, in a sense, when an executive has a big change coming up and is seeking to build momentum by making media appearances. Oh, well, live and learn!

Nightingale replaces Anshul Sadana, who left Arista a year ago after eighteen years at the company. 

Investors seem to hate both of these new appointments, bidding down Arista by almost five percent today, to $90.83, and bidding down Fastly by almost nine percent, to $6.60. 

The analysts covering the stocks seem to be more content with the switch. 

I tend not to write about healthcare or biotech names because I don’t have a background in health or life sciences, and I believe the factors that drive those companies and their stocks are very different from those that drive areas such as computer networking and semiconductors, where I’ve spent over thirty years reporting.

However, I need to make accommodations more and more as technology seeps into every field and every industry. And some of the tech analysts I regularly follow are themselves expanding their coverage into life sciences for the right names.

Case in point, my curiosity was peaked with the initiation of coverage Monday of newly public Hinge Health, an eleven-year-old startup in San Francisco that claims to be changing the nature of outpatient care using artificial intelligence.

The appeal of the stock for many is that it is a kind of software stock, but trading at a cheaper valuation despite high growth at the moment.

Everything was “solid” for Adobe in Thursday evening’s fiscal second-quarter earnings report — that’s the term all the analysts used.

But solid results are not convincing anyone who has been worried about Adobe’s future. The stock fell five percent on Friday, the third quarter in a row that the shares have declined on the report. Adobe is down fifteen percent in the past twelve months. And the three-year, annualized return is a paltry three percent.

As I explained in this week’s podcast, the Street is concerned with whether creative professionals, those who grew up on Photoshop and the rest, are enough to offset what is expected to be a gradual migration away from Adobe of the less-dedicated users to artificial intelligence tools, things such as ChatGPT.

The question seems to be whether Adobe’s power in the visual creation market will be eroded by AI. 

There are two versions of the database giant Oracle, the good one and the bad one. Thursday is seeing the good version triumph, the stock up thirteen percent at $199.74, following Wednesday evening’s fiscal fourth-quarter earnings report.

Price targets are zooming, with Brad Zelnick of Deutsche Bank having what I see as the highest target, at $240.

The bad version of Oracle, which we saw the prior two quarterly reports, is a version of Oracle with sales growth that is underwhelming. The good version of Oracle is a company with tons of potential, in the form of the key non-GAAP metric “remaining performance obligation,” or, RPO, a measure of its backlog.

The numbers, actually, were not bad, with the revenue upside the biggest in two years — but only after brokers had cut their own estimates in the past six months. The forecast of $15.05 billion, at the midpoint, is just fractionally above the average estimate for $14.991 billion.

Shares of Nvidia are down fractionally Wednesday as the Street seems relatively unmoved by the keynote in Paris of CEO Jensen Huang at his company’s first-ever Euro version of its “GTC” event, which had taken place in the U.S. in San Jose back in mid-March, and which was also received rather tepidly at that time. 

I think Nvidia is in some ways such a broadly understood story that these kinds of massive conferences by Huang don’t have the ability to surprise much. 

And, to be fair, a lot of what was announced sounded to me very familiar. For the actual news items, you can check out Nvidia’s official blog post. 

Among the points discussed by Huang were that the company’s top of the line machine, the “GB200 NVL72,” is being produced by Nvidia’s partners at a rate of 1,000 a week, which is a nice reassurance that business is moving along at a steady clip. 

Perhaps the most buzz=worthy quote from Huang was his enthusiasm about quantum computing. Huang said the field is “reaching an inflection point,” and that “We are within reach of being able to apply quantum computing, quantum classical computing, in areas that can solve some interesting problems in the coming years.”

The saying is “sell in May, and go away,” and they haven't done that yet. For the week ended June 6th, the Nasdaq Composite Index rose 2%, as did the S&P 500 Index, and the TL20 group of stocks to consider was up 4%.

Helping the positive sentiment was a bullish report from Broadcom, and very positive remarks on Micron Technology.

But there were also reminders that the tariff issue has not gone away. Shares of Ciena fell sharply following its earnings report on Thursday, June 5th. Ciena's CFO, James Moylan, who is retiring, said that Ciena's profit was partly impacted by having to absorb an expense related to tariffs totaling mid single digit millions of dollars last quarter. Bulls advised that you look past that to all the brilliant opportunities for Ciena with fiber-optics and data centers.

Shares of Intel are higher by almost nine percent Tuesday afternoon, at $22.27, on no particular news — at least, not news that seems significant. 

The company said that Imperial College London has chosen Intel’s Xeon chip for its next supercomputer. Intel is apparently working with SoftBank on a new kind of memory chip for AI. And tomorrow, Wednesday, Intel’s CEO, Lip-Bu Tan, will discuss the roadmap for the company’s chips at an event called the “supplier summit.”

None of that seems eight-percent worthy. I’m left with a couple of random items that could have gotten someone excited. 

Will the third time be the charm for Oracle? 

The stock has lagged the software group this year, up only six percent, at a recent $177.15, albeit up a very healthy forty-one percent over the past twelve months. The underperformance is the consequence of two quarters in a row of disappointment from a revenue standpoint, the most recent one in March.

Oracle reports Wednesday evening, and a chorus of analysts arose Monday with some mixed views: Jefferies’s Brent Thill, BMO Capital’s Keith Bachman, Morgan Stanley’s Keith Weiss, and Citigroup’s Tyler Radke.

The most bullish is Thill, who has a Buy rating on Oracle stock, and who raises his price target to $200 from $190.

Thill writes that his survey of twenty of Oracle’s software resellers shows that two thirds “either hit or exceeded their plans” in the quarter, and almost half “cited q/q pipeline improvement.” 

“We need to change the mindset of forty years, fifty years of how data is maintained and stored. This is very hard.”

Things keep humming along at Broadcom, which on Thursday evening reported results that topped expectations, and for its outlook as well. 

However, the disappointment that has sent the stock down three percent in early trading, to $254.60, appears to stem from the company failing to talk about what’s in store for 2027. 

The bulls are putting that aside at the moment, figuring there’s plenty of time for Broadcom to talk about fiscal 2027 given that we are only halfway through fiscal 2025 (ending in October.)

The growth, meantime, is off the charts. Broadcom beat its own target for artificial intelligence-related chip sales last quarter, which had been forty-four percent growth, year over year, instead delivering forty-six percent, at a total of $4.4 billion. And for this quarter, that rate is going to speed up to sixty percent.

It was a report very much consistent with the positive trend seen in March, only more so..

Kudos to Raimo Lenschow of Barclays, who on March 30th took the side of database vendor MongoDB, arguing the shares were too cheap after what had been a twenty-five percent decline since the start of the year. 

Thursday, the stock is up twelve percent at $222.73, after the company’s earnings report Wednesday evening topped expectations. 

Lenschow takes a victory lap today, raising his price target to $270 from $252, noting that the principle element in the results was the company’s “Atlas” database, which makes up over two thirds of quarterly revenue, seeing its sales growth speed up for the first time in over two years, from twenty-four percent sales growth last quarter to twenty-six percent this time around. 

That’s a big relief for investors, as the product had been experiencing a profound slowdown for over two years now, the principle worry for the stock for a long time. 

Returns for investing, as opposed to random speculation, come over many years, as the best names are given a chance to work.

The TL20 group of stocks to consider is up 190% Thursday since its inception on July 15th of 2022, setting a new high for the group, surpassing the 187% achieved on January 22nd. That’s across five rebalancing events since the original group. 

That return is 2.7 times the return of the Nasdaq CompositeIndex in that same time. 

• The week that ended May 30th, an abbreviated week with Monday’s Memorial Day holiday in the US, was a good one for stocks, with the Nasdaq Composite Index rising 2%, lifted by most of large cap tech, with Oracle being the best performer, up 6% for the week, and other strong names such as Broadcom, Samsung, Texas Instruments, Meta, Nvidia, Apple, etc.

• We’re past the thick of earnings season, the tariff talk has cooled a bit, and tech has come roaring back. As just one indicator, the TL20 group of stocks to consider ended the week up four percent for the year versus the one-percent decline of the Nasdaq and a half-point gain of the Standard & Poor’s 500. 

Equity research types love puns, in good times, and in bad. 

The news Tuesday is very good for shares of Credo Technology Group, which is soaring by over eighteen percent, at $72.47, the exact opposite of the plunge in the stock back in March.

With a pun at the ready, TD Cowen’s Josh Buchalter declares that “CREDOn’t have anything to pick on.” 

In the March earnings report, the problem was that among Credo’s three biggest customers, Amazon, Microsoft, and Elon Musk’s xAI, Amazon made up an amazing eighty-six percent of revenue last quarter, raising the fear of customer concentration. 

This time around, Credo not only beat expectations, as it usually does, it said Amazon dropped from eighty-six to sixty-one percent of revenue.

Buchalter and others rejoiced.

“Our business plan is to drive growth both from this broadening customer footprint, and from our largest customer, in parallel.”

The White House’s budget proposal, which goes by the much-disputed name of the “One Big Beautiful Bill,” has some areas of concern for renewable energy names such as First Solar, whose stock is already down fifteen percent since the start of the year.

But, have no fear, writes Mizuho Securities’s Maheep Mandloi, who reiterates his “Top Pick” designation on First Solar on Monday, after concluding that the bill is “less dire than consensus fears.” 

The fear, as mentioned by The Washington Post’s Theodoric Meyer on Monday, the bill “would scrap credits for [renewable energy] projects that don’t start construction within 60 days of Trump signing the bill into law,” an effort to swiftly roll back tax credits that were given to the group by Democrats in 2022.