The retailers that go with Amazon rather than fight Amazon will win, says Davidson
If you’re looking for a derivative bet on the increasing power of Amazon, you might consider the companies that just go with the Amazon flow, riding that river, rather than fight the current.
That’s the perspective offered Wednesday morning in a note to clients by Tom Forte of D.A. Davidson. Forte has for several years now followed Amazon’s every twist and turn as a “platform,” meaning, a set of resources that let other companies sell stuff, what Amazon terms third-party sales, or “3P.”
Third-party sales are rising as a percentage of Amazon’s business, and Forte thinks they’re going to really surge in future. He sees 3P going from fifty-four percent of total sales at Amazon in the third quarter of this year to seventy-five percent over the long term.
“Increasingly the company is building a platform for third-parties to sell merchandise on Amazon,” writes Forte, “from enabling them to advertise their products to holding their inventory in its fulfillment centers to shipping them to consumers (often via its own first-party delivery assets).”
The reasons are obvious. Amazon makes a higher margin on third-party sales, and can make lots of extra money for its businesses such as advertising and its AWS cloud service.
All you need to know to buy Qualcomm per Baird
Shares of wireless chip giant Qualcomm may have almost forty percent upside from a recent $146.34, according to one of the most pithy stock initiations I’ve seen, from Baird analyst Tristan Gerra.
Gerra has been in the business for a long time, and I had to marvel at the fact that his note, issued late Tuesday, was just a couple of pages. Usually, an initiation of coverage is a book-length tome that pores over every aspect of the business, and sometimes aspects of the meaning of life and the universe to boot.
But Gerra is to the point. Qualcomm, whose shares he starts at Outperform with a $200 price target, is expanding on multiple fronts, not just phones, including the personal computer, cars, and the Internet of Things, or IoT.
But first, phones. You may have heard Apple is poised to cut Qualcomm out of the iPhone, but Gerra doesn’t get into that. Rather, as he sees it, the numbers are encouraging. In the most recent quarterly report, management said that it expects smartphones that use 5G technology to expand by one hundred and fifty percent in 2021, in unit terms, to a range of 450 million to 550 million units. That’s good because Qualcomm has tons of design wins he points out. They have “700 5G designs announced or in development and over 110 5G agreements,” writes Gerra.
He also notes the company is seeing increases in its amount of content in smartphones. And China is a notable source of upside. He sees the company gaining share in selling chips into Chinese handsets as the percentage of phones with 5G in China is already high, at sixty-eight percent.
The smart way to buy expensive software stocks according to Canaccord
I know what you’re thinking: How will I ever get through another year of software stocks being insanely overpriced? Especially when people tell you that you will just have to get used to paying up for growth.
Among stock analysts, I’ve noticed that Canaccord Genuity’s software analyst, David Hynes, is one fellow who’s pretty down to earth about software valuations. So I was struck when I saw a note come over the transom from Hynes on Monday titled “Category Killers: the path to navigating an expensive space in 2021.”
Hynes’s point is that companies that come to dominate in a particular area of software, such as ServiceNow in IT, tend to have stocks that are “expensive the whole way up.” You don’t wait for them to become cheap.
“These are the types of names that you can buy, put away, and own almost irrespective of price,” writes Hynes. Clearly, he writes, investors are doing that with Snowflake, the most expensive stock in software anyone can remember.
With that in mind, Hynes proposes five companies that could be category killers. Four of these I had heard of: MongoDB, Veeva, Five9, and ZoomInfo. The fifth, Avalara, is new to me. The sixteen-year-old company is based in Seattle and provides programs that have to do with tax collection.
Wall Street licks its chops for security software
It’s a funny thing about computer security: while the solution, ultimately, might be to make better software, the industry response is generally to sell more stuff to fix the stuff that broke.
And so the massive computer breach that emerged last week, affecting Microsoft, as well as the Department of Defense, the U.S. Treasury, and many corporations and agencies, in the U.S. but also around the world, prompted excitement Monday among analysts of security software firms. The Street sees game on for stepped-up security buying in 2021.
Even as shares of SolarWinds sold off the past two days, shares of FireEye, the cloud security firm that first detected the breach, have shot up, rising forty-one percent in the past two trading sessions. FireEye is not the only beneficiary.
Analysts today were making their lists and checking them twice.
Michael Turits with KeyBanc Capital Markets today reiterated an Overweight rating on shares of Palo Alto Networks and Okta, and raised his targets on the stocks to $391 from $350 for Palo Alto, and to $313 from $282 for Okta. Turits has “increased confidence that cyber security spending will increase subsequent to the attack,” he wrote in a note to clients.
The SolarWinds hack: When fundamental software flaws are a problem
The massive hacking affair that surfaced last week, known as the SolarWinds hack, is being discussed by lawmakers as potentially an act of war. Before rushing to war, one might reflect on the fact that the attack has all the hallmarks of a problem that has been going on for decades, namely, poorly designed software.
The hack attack has compromised infrastructure at Microsoft as well as at the Department of Defense, the U.S. Treasury, and many corporations and agencies in the U.S. but also around the world. This is a high-class kind of cyber attack, top-drawer stuff.
But it also has a mundane side. FireEye, the cloud security company that first told the world about the attack, describes in a blog post the basics. According to FireEye, at the center of the attack is a file called a dynamic-link library, referred to by the shorthand “dll,” dee-el-el.
This is just a fancy name for a program that any other program on your computer can tap on the shoulder to ask for some help. Sometimes, vendors of software break up their programs into multiple programs, including one central program that does most of the work, and then a dynamic-link library that is kind-of a helper to the main program.
Why it’s suddenly Microsoft and everyone against Intel
It’s been a rough month and a half for Intel.
The report Friday by Bloomberg’s Ian King and Dina Bass that Microsoft is developing some in-house computer chips, perhaps to displace chips from Intel, is merely the latest defection. Two weeks ago, Amazon announced its own new chips for its AWS cloud computing service that, again, could take over work that might otherwise be done on Intel chips. And, of course, Apple last month unveiled Macs that run on a new Apple-designed chip, the “M1,” which will reduce how many chips Intel can sell to Apple.
Headlines about competition come and go, but what is happening with Intel is part of a broader confluence of things that have been going on for many years.
The original sin for Intel, the sin for which it has been paying for over a decade, was having missed the mobile computing revolution. The very first iPhone, introduced by Steve Jobs in 2007, used a chip from Samsung. That Samsung chip used chip instructions created by a supplier that is Intel’s main rival, ARM Holdings, a thirty-year-old company that is currently in the process of being bought by Nvidia. The Apple M1 chip, and the Amazon and the Microsoft parts, also use ARM’s technology. ARM has been taking away business that might have been Intel’s for years now.
What I want to know is when Airbnb becomes the property manager for hotels
It usually takes about a month after an IPO for stock analysts to initiate coverage of a newly public company, but swift individuals can often get in ahead of the crowd, assuming they are not in the underwriting syndicate that created the offering.
Such is the case for Shyam Patil of Susquehanna Financial, who today initiated coverage of Airbnb, which went public December 10th, with a “Positive” rating and a price target of $180.
Airbnb shares are up seven percent since the first day of trading at today’s close of $157.30.
Patil is very positive, seeing numerous “vectors” of revenue growth for the company in the years to come, even though Airbnb has already racked up billions annually in revenue, despite a big plunge this year because of the pandemic.
The most interesting thing to me in Patil’s report is his discussion of hotels. The thing that I’ve been wondering about with this company is when they take over the hotel booking business and maybe even some of the functions of property management.
I think most would assume that sometime after the pandemic, some things will be different about travel. No one knows what that might be for certain, but one possibility is that leisure travel will snap back for a lot of people cooped up. At the same time, business travel may not snap back because a lot of offices will have shut facilities, and people will be working from home a lot more, on average, as a kind of new normal, to use a terribly over-used phrase.
Snowflake: At fifty times 2023’s revenue, math doesn’t add up, says William Blair
Snowflake, which came public September 16th, has a brilliant opportunity in its marketplace for data sharing, according to William Blair analyst Kamil Mielczarek, but the stock, at fifty times potential sales in 2023, is just too expensive to recommend, he thinks.
Mielczarek initiated coverage of Snowflake shares today with a “Market Perform” rating, noting that the company “is growing at a faster rate than almost every technology company in history” at this point in its life, but its stock “prices in a huge growth premium for at least the next five years and is not representative of the various risks.”
The report was prepared in conjunction with long-time William Blair analyst Bhavan Suri.
Mielczarek thinks the math is against the shares even assuming enormous growth. He reviews growth of comparable companies such as Salesforce and ServiceNow and Workday at comparable periods in their lives. It’s true, Snowflake beats them all, with sales growth of one hundred and seventy-four percent in its first year of life, one hundred and nineteen the next, then eighty-seven percent in year three and sixty-four percent in year four.
That’s well above growth of even ServiceNow, which was one hundred and sixty-three percent in its first year, and seventy-four percent the next.
All that means Snowflake is “in a class of its own.” However, even high growth is going to taper off because of the law of large numbers, and Mielczarek sees a tipping point in 2026.
Don’t FOMO your portfolio, buy Oracle, says Monness
Oracle chief technologist and chairman Larry Ellison loves to trash-talk other software companies, but so far this year, that posture is not helping his stock, which is up just twenty-two percent, vastly trailing most software names, including even established companies such as Microsoft, which is up thirty-seven percent this year.
So, perhaps Oracle needs a champion. Enter Monness Crespi Hardt analyst Brian White, who today wrote to clients that people are way too obsessed with missing out on the next big thing, which is why they are chasing over-priced, momentum software stocks.
White advised chilling out about expensive software names, and instead buying shares of a beaten down gem, Oracle, whose shares he raised to a Buy rating from a Hold, with a price target of $82, which would be $26 above a recent price of $65.
“With the fear of missing out on the next big digital theme with little regard for valuation, next-gen software multiples have been bid into the stratosphere this year,” writes White, “with a growth-at-any-price investment philosophy, a phenomenon that has begun to resemble the tech mania of the late 1990’s.”
White makes reference to an average price-to-sales multiple of twenty-one times next year’s revenue for high-growth software names, but mentions that two of them are trading at thirty-five times and forty-four times, respectively. White doesn’t mention names, but those multiples match two software names I can think of, Okta and Datadog.
After two roaring years, chips likely to underperform in 2021, says Deutsche
Chip stocks have been a star during the pandemic, rising fifty percent this year compared to a fifteen percent rise in the Standard & Poor’s 500 Index.
And for that reason, 2021 may not be as rich, according to Deutsche Bank analyst Ross Seymore. The Philadelphia Semiconductor Index, the most common measure for chip stocks, also outperformed the market in 2019, rising by sixty-one percent, double the S&P’s return. The year prior, 2018, the SOX, as the chip Index is called, declined by nine percent, more than the seven percent decline of the S&P.
That’s one hundred and forty-one percent in two years versus forty-nine percent for the S&P. And now it’s time the good times to end, argues Seymore. The SOX never beats the S&P more than two years in a row, he notes.
“We fear that post-Covid ‘normalization’ may make the semiconductor sector a source of capital rather than a destination for it,” writes Seymore, “and thereby continue the somewhat surprising, but nonetheless factual historical trend of the SOX never outperforming the S&P 500 for more than two consecutive years.”
“Since its inception in 1994, the SOX has never outperformed the S&P for more than two consecutive years, with the average underperformance in the third year being -6%.”
The battle that is shaping up between Dynatrace and Splunk
If you like the thrill of competition, an emerging battle to consider is the face-off between Splunk, the seventeen-year-old vendor of security software, and Dynatrace, the six-year-old vendor of application monitoring software.
Rosenblatt Securities analyst Blair Abernathy late Wednesday issued an initiation of coverage note of both stocks, and gave both a Buy rating, and observed that the two companies are increasingly moving into direct competition with one another because they are expanding their product offerings. (Abernathy has a long history at several prominent firms such as Stifel prior to joining Rosenblatt this year.)
Splunk was traditionally all about security, looking at log files to detect computer break-ins. The company’s morphed via a series of sixteen acquisitions in the last seven years. Its role now includes very broad monitoring of any corporate data, and also constant monitoring of IT operations. The total addressable market, according to Abernathy, is $81 billion for Splunk. By way of reference, Splunk has just $2 billion in annual sales.
Abernathy likes the company’s expanding ambition. “As it spreads across more buying centers, Splunk’s recurring revenue per customer increases significantly and, we believe, Splunk also becomes harder to competitively displace.”
And then there were two: Are Qorvo, Skyworks acquisition bait?
When markets are consolidating, it can be interesting to look around at things that might be future areas of acquisition for investible opportunities.
One such area could be the field of RF chips, which has dwindled to just two survivors, as far as “pure play” investments, Skyworks Solutions and Qorvo.
The market for semiconductors, of course, has been consolidating for years now, and in particular, the wireless market has consolidated a great deal. Apple last year bought the baseband modem business of Intel for a billion dollars in order to have an alternative to Qualcomm. Back in 2016, the chip company then known as Avago bought Broadcom, which put it in competition with Qualcomm for certain kinds of wireless chips.
The area that has remained a pure play in the mobile business is what’s called the “front end,” a gaggle of chips that include power amplifiers, switches, filters, multiplexers, and other arcane things. These parts shape the quality of a wireless signal as it comes into and goes out of the so-called transceiver chip, and the modem chip that sits behind the transceiver.
Upstart IPO: Is the personal loan business actually A.I. or something else?
Update 3: The stock closed up almost fifty percent at $29.47. Not bad!
Update 2: The shares opened at $26, now up about twenty-five percent at $25.11.
Update: The deal has priced at the low end of the expected range of $20 to $22.
Previously:
One of the great things about financial markets these days is that the term artificial intelligence is being thrown around a lot. That’s nice, because it’s an opportunity to have a frank discussion about the term AI and what we all think it means or should mean.
The term AI has no precise definition, and so it’s not surprising that some uses of the term may stretch the bounds of what’s fair use.
For example, I’ve written that C3.ai, which uses the ticker symbol “AI,” is not an AI company. They are an infrastructure software company. There’s a difference.
Outside of tech companies such as C3.ai, there are an entire cohort of companies coming to market now that claim to be using AI to transform industries in one way or another.
One that I’ve mentioned before is Rocket Loans, a thirty-five-year-old, Detroit-based company that went public in August and that claims to be using AI to re-invent the mortgage industry.
This week heralds another intriguing public offering, Upstart Holdings, an eight-year-old company based in the sleepy Silicon Valley town of San Mateo that says it wants to transform the personal credit market. It bills itself as “a leading, cloud-based AI lending platform.”
Cambium: Here comes the rural gold rush, according to Northland
There’s something of a gold rush that’s getting underway in rural America, a twenty-billion-dollar program that has bi-partisan support known as the Rural Digital Opportunity Fund.
The ten-year initiative, administered by the Federal Communications Commission, is giving billions in contracts to service providers to bring more connectivity to the countryside in order to “close the digital divide.”
The FCC last week announced the winners of an initial auction for the Fund, the first companies selected to receive federal money to roll out service.
I was glad today to see a favorite of The Letter get some favorable mention as a consequence of the Fund. Cambium Networks, a small telecom equipment provider I’ve interviewed here a couple of times, got the thumbs up Tuesday from analyst Tim Savageaux at Northland Securities.
The Fund represents an “opportunity that the company has previously sized in the $100s of millions,” notes Savageaux. With the announcement of the winners, that opportunity “begins to roll out,” he writes.
Micron: the PC rally is not over, says Wedbush
This has been a banner year for personal computers, and it looks like there’s no slowing down, Wedbush analyst Matthew Bryson declared in a note to clients Tuesday.
“While we have/had been expecting a better environment for memory next year, it certainly now appears the setup for PC demand will be much stronger than we would have anticipated a few months ago,” writes Bryson, “when we were looking for demand to slow following a work-from-home-related refresh.”
Personal computer shipments, in case you’ve forgotten, had their strongest quarterly sales result in five years this past September quarter, according to Gartner, with sales up 3.6%.
Bryson covers shares of memory-chip maker Micron Technology, which he rates Neutral. Despite that muted stock view, Bryson raised his price target on Micron shares to $82 from $55, based on the higher PC expectation.
Bryson’s comments are part of a trend of increasingly upbeat remarks about Micron’s outlook, such as comments by Deutsche Bank last month.
Snowflake, C3ai lose altitude: hard to control what you don’t understand
There is a concept in engineering called feedback control. An airplane, for instance, has to constantly adjust to wind patterns to maintain altitude.
Some engineered systems are so complex that they can’t be controlled by a person, they need to have a computer control them.
A brilliant example of that was given in a talk last week by Jeff Shamma, a professor at the University of Illinois at Urbana-Champaign who is an expert on feedback control systems. Shamma cited the X-29, a novel airplane developed by NASA in the 1980s. It was both highly maneuverable but also highly unstable. So unstable, in fact, was the X-29 that it had no human pilot.
“It could not be flown by humans,” said Shamma. “Rather, a human pilot issued commands to a flight control system, and then the flight control system issued commands to the vehicle.”
Something similarly inhuman is likely going on with some very expensive recent IPOs, Snowflake and C3.ai, two cloud computing companies that have run into some turbulence.
Qualcomm still has another billion modems to sell every year outside of Apple
Shares of Qualcomm sold off by seven percent Friday on a report Thursday by Bloomberg’s Mark Gurman that Apple, during a town hall meeting of its staff this week, discussed making its own cellular modem chips for the iPhone, presumably meaning it will dump some or all of the chips it buys from Qualcomm.
A modem is the chip that controls how communications are formatted to be sent over a wireless communications link. It is essential to all wireless devices, including smartphones.
As many fans of Qualcomm are saying Friday, this is probably a buying opportunity for Qualcomm stock.
The report by Gurman is not a new development: analysts knew at the time that Apple purchased Intel’s modem business in July of 2019, for one billion dollars, that Apple intended to make its own modems.
If Apple dumps Qualcomm chips from all iPhones in coming years, Qualcomm would still have a very vibrant market opportunity. Apple is only one vendor of smartphones; there are plenty of other manufacturers who need modem chips.
The potential of Snowflake, according to Larry Ellison
Oracle’s chairman, Larry Ellison, also chief technologist of the company, is not just a smart person with lots of strong opinions. He is a person who likes to opine about anything that might fall within his bailiwick. That’s one of the fun things about Ellison, that he just likes to share his perspective.
And so, Thursday evening, a fairly routine report by Oracle of quarterly results — revenue and profit better than expected, a forecast that was well-received — turned into an opportunity for Ellison to expound on the software landscape as he sees it.
And, in particular, a company called Snowflake, which went public September 30th, which has the most expensive stock anyone can ever remember, and which some think could kill Oracle’s business in some areas.
Analyst Brad Zelnick of Credit Suisse asked Ellison how Oracle will stay competitive given that technologies sometimes rise and fall as generational buying habits in information technology change.
Technology will solve Facebook, not regulation
The suit brought Wednesday against Facebook by New York attorney general Letitia James and forty-seven other U.S. State attorneys general, for allegedly abusing Facebook’s monopoly in social networking, has obvious parallels to the lawsuit brought against Microsoft by the U.S. Department of Justice twenty years ago.
And as in the case of Microsoft, technology, not regulation, is the ultimate solution.
To understand why that is, consider how the two cases are similar, and what ultimately happened to Microsoft.
Similar to the U.S. Department of Justice’s suit against Microsoft, James and her peers detail numerous ways in which Facebook allegedly sought to either co-opt companies that were perceived as threats, or to destroy them by cutting off their oxygen supply, as it were.
Referencing internal Facebook communications, the suit documents how Facebook CEO Mark Zuckerberg would discuss buying companies, which it ultimately did in the case of Instagram, or else forbidding them to use Facebook’s application programming interface. The description echoes allegations at the time of how Microsoft used its control of the Windows programming interfaces to try and throw obstacles at Netscape in the browser wars, and how Microsoft strong-armed PC makers to install only Windows and not other software on new computers. The suit brought Wednesday against Facebook by New York attorney general Letitia James and forty-seven other U.S. State attorneys general, for allegedly abusing Facebook’s monopoly in social networking, has obvious parallels to the lawsuit brought against Microsoft by the U.S. Department of Justice twenty years ago.
And as in the case of Microsoft, technology, not regulation, is the ultimate solution.
To understand why that is, consider how the two cases are similar, and what ultimately happened to Microsoft.
Similar to the U.S. Department of Justice’s suit against Microsoft, James and her peers detail numerous ways in which Facebook allegedly sought to either co-opt companies that were perceived as threats, or to destroy them by cutting off their oxygen supply, as it were.
Referencing internal Facebook communications, the suit documents how Facebook CEO Mark Zuckerberg would discuss buying companies, which it ultimately did in the case of Instagram, or else forbidding them to use Facebook’s application programming interface. The description echoes allegations at the time of how Microsoft used its control of the Windows programming interfaces to try and throw obstacles at Netscape in the browser wars, and how Microsoft strong-armed PC makers to install only Windows and not other software on new computers.
C3ai more than doubles in debut: they know not what they buy
You could say a company is successful when it sells a quarter of a billion dollars worth of stuff, any stuff. C3.ai, which went public Wednesday, by that measure is a successful company, having amassed roughly that much in future revenue as of October.
The fact, however, that C3.ai is successful appears to have gone to some people’s heads. The stock offering, after having been twice increased in price, from $34 initially to $38 and then, finally, to $42, proceeded to double in value by the end of Wednesday. In fact, the stock opened like thunder, at $100, one hundred and thirty-eight percent above the offer. It closed at $92.49.
What inspires people to pay more than double what the most plugged-in, most educated investors would pay?
It’s a pressing question, because the stock now fetches nineteen times what I could imagine the company making in a very generous scenario, a year from now.
I would bet money most people buying Wednesday have no idea what they’re buying. It’s a complex brew of software, as I’ve explained, and few understand it.