I have a confession to make… sometimes I’m too quick to filter through the insane piles of junk email that stack up on my desk each morning. So many email teaser ads from newsletters have a long shelf life, running over and over for years, that I sometimes see the intro to an ad, say “covered it!” and toss it aside.
But sometimes those ads change. So when a reader persisted in following up with me about a Paul Mampilly ad, I thought I should double-check it. This is an ad that has Mampilly’s typical presentation “video” and talks about those “secret warehouses” that contain the future. Here’s how the ad starts:
“These Warehouses Hold the Key to a Potential Multi-Billion Dollar Revolution…
“With a $15 stock, early investors could take advantage of the ‘most important technology since the internet itself.'”
And I thought that was yet another repeat of Mampilly’s pitch for Advanced Micro Devices (AMD) that ran almost nonstop for two years with a similar “in these warehouses” pitch and the same stock photos used as tinsel on the tree… but, it turns out, I was wrong. There have been some little tweaks, like referring to this now as a “$15 stock” in that headline instead of promising “huge profits as this brand new industry surges up to 77,400%.”
So the general push is unchanged, this is still all about how massive data centers around the world hold that “secret” that is blockchain technology, and how that will change the world as it streamlines payment technologies and much of the rest of the internet… but now he’s pitching a different company.
So we wanna know what it is, right? AMD did awfully well, of course, will Mampilly capture lightning in a bottle again?
In fact, the clues in this pitch now specifically rule out the chipmakers who make blockchain possible (which was mostly AMD and NVIDIA in the first crypto boom, though now mining-specific chipsets have been built by others), and he’s hinting at an entirely different kind of company. Here’s a bit from the pitch:
“These processors are made by a few different companies— like Intel and others around the world. But these chipmakers are already expensive and that’s not where you should put your money.
“I’ve identified one company based in Sunnyvale, California that’s been overlooked by Wall Street—it doesn’t make the chips themselves, but the interfacing software that enables these chips to communicate with each other.
“This technology is as obscure as it is essential. And the market is oblivious to this company right now— its trading at only $15 per share.
“This is a huge opportunity.
“Historically, this stock has performed well, gaining an average of 22% every year for the last 23 years.
“And I believe it’s about to explode much higher.”
So… what other clues do we get about this little company?
“This company finalized a historic deal with Visa last October… the credit card giant acquired this firm’s blockchain integration software.
“This tech is about to be integrated with roughly 90% of Visa’s total payment volume…
“That’s nearly $11 trillion worth of global payments every year.
“When this integration is complete, other companies are likely to come calling…”
And there are some numbers attached, too, which always makes the Thinkolator’s job just a little easier… and it helps us to confirm those answers we already suspect:
“…the company is predicting a massive revenue increase over the next year—from $224 million at the end of 2019 to a whopping $350 million just over one year from now.”
So what’s our story stock this time around? Those clues point the Thinkolator to one specific company, Rambus (RMBS), which has been around for about 30 years and is best-known for developing memory interface technologies that speed up the movement of stored data within a computer (that’s an oversimplification, I’m sure). They’ve never been a chipmaker, really, but their goal has always been to license their designs, including RDRAM memory, and over the past 20 years they’ve been asserting that they own important patents in SDRAM and DDR memory technology. For a while they were derided in some quarters as a “patent troll,” just trying to squeeze money out of other companies like Micron and NVIDIA, but they have been somewhat more collaborative in recent years and are trying to do more licensing and joint development of memory technologies. At least, that’s the way their history appears to me — I have not followed the company in years. Here’s how the company describes itself:
"reveal" emails? If not,
just click here...
“We are dedicated to delivering first-to-market, high-quality memory, SerDes, and embedded security solutions for our customers.
“Throughout our 30-year history, Rambus has led the industry with innovations and IP solutions that solve the fundamental challenges faced by leading-edge computing systems. We have done so by predicting and addressing the major disruptions in technology development, building a portfolio of foundational IP and offering solutions that enhance the performance and security of the most demanding applications.
“Leveraging our semiconductor expertise, Rambus solutions speed performance, expand capacity and improve security. From data center and edge to artificial intelligence and automotive, our interface and security IP, and memory interface chips enable SoC and system designers to deliver their vision of the future.”
This is a small company, but a relatively high-margin one — not unlike some other names in technology that count on other people to develop and manufacture their products and rely on license agreements and royalties, like Universal Display (OLED). Rambus over the past year has had gross margins of about 74%, which is a little short of OLED’s 80% but still very high — even the fabless semiconductor companies who don’t own factories, like NVIDIA, tend to be down around 60%. The challenge for Rambus is that their high margins don’t provide much room for improvement or greater efficiency as they grow… and they’re not growing super fast.
They are growing, to be clear, and they are generally profitable — though their cash flow is much better than their earnings, and the recent growth has largely been a “snap back” from the awful numbers they put in late 2018 and 2019. They believe that they’ve successfully refocused the company on growth areas in security, data centers, and 5G networks, partially through acquisitions, and that they’ll be able to build a larger company on the foundation of their predictable royalty and licensing deals, including a big four-year renewal of their agreement with Micron.
Here’s a quote from their last earnings announcement, which was for a relatively weak third quarter on the revenue front but included a significant buyback authorization and some positive comments about the future, and helped drive the shares a bit higher than the $14 level they had before earnings (in early November):
“The Company’s memory interface chip business had a solid quarter, continuing to significantly outpace market growth with a 39% increase in quarterly revenue year over year. This growth is driven by ongoing increases in market share in DDR4 and continued demand in cloud and data center. For the industry transition to DDR5, Rambus is in a leading position for qualification with the memory ecosystem and CPU partners in next-generation systems.
“Growing complexity in SoC design across data center, AI and 5G markets continues to drive customer engagement for the Rambus Silicon IP business, with an increasing number of design wins in interface and security IP. Designed to meet the needs of the most demanding data center AI/ML workloads, Rambus leads the industry with the fastest, silicon-demonstrated HBM2E memory interface solution capable of running up to 4 Gbps.”
And to check the boxes on their other clues — Rambus used to be headquartered in Sunnyvale, and it might be that their move is incomplete due to COVID, or that they still occupy that building in Sunnyvale that has a big Rambus sign on it, but their headquarters officially moved down the road a few miles to San Jose recently. And yes, they did report $224 million in revenue for 2019, and over the summer the average estimate for 2020 revenue was about $360 million (though that number is up to $415 million now — a huge recovery from the $200 million or so that analysts foresaw at the beginning of the year).
And yes, Visa did acquire Rambus’ “blockchain integration software” late in 2019, though the actual language was that Visa “completed the acquisition of the token services and ticketing businesses, formerly Bell ID and Ecebs LTD, from Rambus.” I don’t understand why Mampilly teases this as a reason that others would “come calling” for this blockchain/payments technology, because it wasn’t a licensing deal or a royalty deal, Visa bought that business from Rambus. Rambus does still offer some crypto-focused products, particularly their “crypto accelerator cores,” but the payments business they sold to Visa is gone, and was not large enough to really impact their earnings in a visible way even in the quarter that the deal closed (the terms weren’t specifically disclosed, but that indicates to me it was a pretty small deal), and Rambus doesn’t generally talk much about blockchain or cryptocurrency applications in their earnings releases or conference calls.
It doesn’t strike me that blockchain is a particularly big driver for Rambus, but perhaps that will change in the future. Certainly data centers (or as Mampilly hints at them, “Warehouses”) are a big part of their market and a big expected driver of future growth (as they are for most of the big chipmakers, including AMD, NVIDIA and Intel). If you’d like some more color on how things are going, it’s worth checking out the transcript from their quarter conference call.
Rambus has turned things around to at least some extent, and is expected to post earnings growth of 5% or so over the next couple years, on revenue growth in the 5-10% range, which is not bad for a company that’s primarily in licensing and royalties and might be sustainable, but I don’t really understand the long-term potential or strength of their asset portfolio so I’m personally not ready to jump on board yet. There is certainly potential, and compared to much of the world of semiconductor IP and design they’re relatively inexpensive, at only about 5X next year’s sales and with a forward PE, assuming they hit analyst estimates, of only about 14 — which used to be “average” and now counts as almost “cheap” … so maybe Mampilly will end up being right, we’ll note down the name and track it over time as we do with all our other teaser “reveals” for you.
The big potential, I assume, is in bouncing back to become a beloved growth stock at some point on the strength of a new product portfolio and some huge growth, which is what happened to AMD after a long period of stagnation — most analysts don’t see that yet for Rambus, but one never knows. Here’s what Rambus has done over the past five years, just for some context on why this restructuring and reorganization of theirs was necessary — that’s Rambus in blue, the S&P 500 in red, and the Philly Semiconductor Index (through the SOXX ETF) in orange:
They’ve posted a strong and levered move over the past couple months, thanks to a surging tech market and some strong moves by chip names in general, though perhaps Mampilly’s attention also helped (we don’t know when he might have first recommended it, or whether this is a “special report” stock or a regular position in his model portfolio, but it is small enough for a big newsletter like Profits Unlimited to have a meaningful impact on trading volume) — this is that chart over the past three months:
Rambus remains a relatively small company with a market cap of about $2 billion, and it’s coming out of an earnings trough with some expectations of more solid revenue and earnings in the next few years, so any change to those expectations, good or bad, will likely continue to cause gyrations for a stock that has been around for a very long time and failed, so far, to generate any long-term growth in shareholder value. Maybe that changes in the years to come, maybe not, I don’t have enough info to feel comfortable betting either way at this point… so you’ll have to make your own call, as always, with your own money.
P.S. Readers always want to know what subscribers to investment newsletters think — if you’ve ever subscribed to Paul Mampilly’s Profits Unlimited, please click here to share your feedback and opinion with your fellow investors. Thanks!