Everyone's Watching AI. They're Missing This

[01:00:00:02 - 01:00:14:16] When most people think about   investing in robots, they picture Tesla's Optimus,  or Boston Dynamics Atlas, or of course one of   those new humanoids from Figur, where that market  is growing more than 30% a year and heading past   $100 billion within the decade. [01:00:14:16 - 01:04:54:22]   Because the demand for humanoid robots hasn't  even begun yet. And these robots are made up of   a lot of individual parts, like actuators, motors,  gearboxes, and wiring. Where about two-thirds of   that cost is in making the robot move, in its  joints and in its hands, while the AI brain is   what everybody obsesses over is only about 6%.  And that machine has quietly become a national   security problem. And if you build one without  Chinese parts, then the cost roughly triples from   about $46,000 to $131,000. Because China controls  close to 90% of the world's magnet processing. And   a powerful magnet sits inside every single motor  that's in those machines. And that is precisely   why the United States is stepping in directly.  Where the DoD is now the largest shareholder in   America's leading magnet maker, and with its price  locked in for at least a decade. So for me, this   is a smarter play. Because instead of guessing  which company is going to build the winning robot,   instead you want to own the suppliers that  every one of them depends on. In this case,   it doesn't matter which brand comes out on  top because all of the parts makers are going   to win. And that is exactly what I'm going to  walk through today. The full bill of materials,   where some of the names are already running really  hot and they belong on a watch list, and then   there's a few that are still priced really well  where you might want to get in. But once again,   the build-out and the demand for humanoid robots,  it hasn't even begun yet. This is still the ground   floor at the very beginning. My entire goal  in this video is to give you the framework and   the important names that you're going to want to  watch out for. Where many of them you might want   to invest in when the timing is right. Because  as I said at the beginning, humanoid robotics   has an estimated 30% annual growth rate. And  the reality is, we haven't even seen these yet.   And the demand hasn't even kicked in. We'll kick  this off with Tim Ken, with the ticker symbol TKR,   which is a 125-year-old bearings maker, through  two businesses called ConeDrive and Spinea,   where they build both the precision gears at the  heart of a robot's joint. Then there's the strain   wave reducer and the cycloidal reducer. Nearly  every advanced robot joint on Earth runs on one of   those two types of gears. And Tim Ken is the only  company on a US exchange that makes both under one   roof. Those gears already ship in real volume into  industrial robots, the arms working on factory   floors, along with medical devices, defense  systems, and solar trackers. So this is a proven,   profitable business already today. But a humanoid  needs those exact same two gears in its hips,   its shoulders, and its knees, which means as  humanoid scale, Tim Ken already makes the part   that they're going to need. Whichever brand once  again comes out on top. And over the last decade,   Tim Ken's revenue has climbed about 60% up  to $4.6 billion. And the engine behind that   growth is its industrial motion division, the side  that makes these robot gears. Now running at $1.5   billion a year, about 34% of the whole company.  And while the business grew, management shrank   the share count by about 28% from $98 million  to $70 million. So every dollar of profit,   it's now going to land on far fewer shares. With  the adjusted earnings set to climb almost 20% to   hit towards $7.21. So after a run like this,  the stock is truly sitting near the top of   its 10-year range, so it does belong on a watch  list given its current demand. But once again,   it's going to probably kick in here shortly once  the rest of robots start to pick up momentum. So   when it comes to the single, scarcest part of  the robot, the joint itself, Tim Ken is the one   American name that is already selling it. Now the  only two companies in the world that build these   gears even better than Tim Ken are Japanese,  and you can own both over the counter. Which   simply means they trade here as foreign shares.  Thinner and wider spread, so you may want to use   a limit order and keep your size sensible and  buy through brokers that offer over the counter   stocks. And unfortunately that does not include  Robinhood. The first is Harmonic Drive Systems,   which invented the strain wave reducer and still  dominates the high end. Where a single humanoid   can use 20 to 30 of them. The second company is  Nabtesco, which owns about 60% of the cycloidal   reducer market and they already supply the biggest  robot arm builders in the world, Fanuc and ABB.   They are the best on the planet at this, which  is exactly why Tim Ken matters. It is the one   American name that puts both gears, the scarcest,  highest value parts of the entire robot, into your   portfolio on at least a normal exchange. [01:04:54:22 - 01:06:09:02]   Now, before I reveal more of my stocks, I want  to show you how I pressure test them. And that   brings us to today's sponsor, Investing Pro, by  Investing.com, where they're the world's leading   financial markets platform and Investing Pro  is the premium tier. I do my own homework on   every stock that I cover, but before I commit, I  run my pick through Investing Pro to see if the   data and the experts line up with me, starting  with ProPix AI. So, it runs the numbers on the   whole tech sector every month and it comes back  with its top picks. Currently around 15 stocks,   based on more than 50 different signals. So, when  the data shifts, the list continually updates. So,   it's only holding its highest conviction names.  So, and over the past five years, Tech Titan   has returned almost three times what the S&P 500  has over that exact same stretch. So, right now,   one of the names that it's holding is Marvell. So,  I pull it up and fair value on it specifically,   it tells me right now it's undervalued, fairly  priced, or it's already stretched. Look,   if you're putting thousands into the market,  you shouldn't be just doing it blind. Right now,   you've got 24 hours left to grab the lowest prices  on investing.com has ever offered. Click my link   right now for 60% off, plus my code BWB for an  extra 15%. Like, don't miss out. It's the best   deal that you're going to see all year. [01:06:09:02 - 01:20:25:03]   Next on the list is Regal Rexnerd with the ticker  symbol RRX, which builds the entire joint under   one roof. Through its motion brands across several  names, one company is making the servo motor,   the micro motor, the ball screw, and the precision  gear that together form a robot's joint. So where   every other pure play is selling just one piece,  Regal can sell a humanoid the whole thing from a   single supplier. That stack earns its living today  inside factory automation, aerospace, medical,   and packaging. So it is a real cash generating  business rather than just a promise. And it is   the identical content that a humanoid joint is  going to need next. And here's where most people   might misread the stock. On its trailing earnings,  Regal looks expensive, close to 49 times profit.   But that number is an accounting illusion which I  like to uncover. When Regal bought Altra for about   $5 billion, the way that deal gets written down  buries the company's real earnings. So the scary   figure is paper and it's not exactly on cash. And  on next year's actual profit, the stock trades   closer to 15 times. So the growth underneath is  what I really want to highlight. Over five years,   Regal has more than doubled its revenue, now at  $5.9 billion. And its free cash flow has more   than doubled right alongside it, reaching $893  million. And it's wisely using its cash to pay   down its debt fast, which drops straight into  its earnings. And it's why profit is set to   climb from where it is today at 30%. So that  mix, growing earnings against a modest price,   gives Regal a peg ratio near a one. So I want to  point out that Regal is a rare company that can   build a humanoid, its whole joint and it still  trades right about where you want to own it. Now   we'll move from the moving joint to knowing where  it is. And that brings us Allegro Microsystems,   which is the chip maker whose sensors tell every  joint in a robot exactly where it is and how fast   it's turning. Every joint needs one chip to sense  its position and another to drive its motor. And   a humanoid has dozens of joints, which is where  Allegro's story gets real interesting. The company   lays out a simple ladder for how much of its  silicon rides inside a machine, as that machine   gets more capable. For example, a robot vacuum has  about $5 of Allegro's chips. And a factory robot   carries about $55 worth. And a full humanoid?  Roughly $150. So as robots go from vacuums to   walking machines, the value that Allegro earns  from each one climbs about 30-fold. And it has   already won a design that puts close to 90 of its  chips inside a single robot's joints. So yes, this   is a cyclical chip business that just came through  a downturn brutal enough to briefly push it into   a loss. And what pulled it back out is AI data  centers. Now a tenth of all of its sales after   more than quadrupling, where revenue jumped 23%  in a single year, and its free cashflow multiplied   almost six times over a record $125 million. The  real catch here is that the market is already   starting to fall in love with it. It's pushing the  stock to its richest valuation in history, and its   first robot wins are with Chinese manufacturers  rather than American ones. So in just keeping   you aware, this is another one to just kind of put  it on the watch list and wait for your discounts.   Now I do want to point out that there's a foreign  name doing the same job over the counter. That's   Mlexus, a Belgium chip maker that makes these  same magnetic position sensors and rolled out a   version built specifically for a robot's joints.  Precise enough to telemotor exactly where it's   turning. Once again, I just want to give you the  full picture of these OTC stocks to do a little   bit extra research on your own. Now I'm not going  to list every single OTC name in today's video,   but I will leave a link to my Patreon where I am  going to go a little bit deeper in the research   so you can check up and see what all I have there.  Now the next area to cover with robots is touch,   and it might be the most overlooked sense of all.  A robot that can't feel force is really useless in   the real world because it would crush an egg, it  would lose its balance, or break your hand when it   shook it. So every humanoid needs force sensors  in its wrists, in its ankles, its fingers, and   this is the one layer the big supply chain studies  call the most wide open opportunity that's left.   Because nobody's built the standard part yet.  Right now one of the best options is Novanta with   the symbol NOVT, which builds the leading six axis  force sensor. The exact part that gives a robot   arm its sense of touch, and it also owns Solera  Motion, which makes the controllers that turn   the sensing into precise movement. So it sells  two of the hardest to source pieces in the whole   robot. And over the past decade, its revenue has  grown more than 160% from $374 million to nearly   a billion dollars. And it has now teamed up with  NVIDIA to drop its sensors straight into NVIDIA's   robot platforms. Now one thing I don't like is  that Novanta does not break out its robot revenue,   and its biggest near term growth is actually a  medical deal. Now if you want the purest bet of   them all, there is Vichay Precision Group, ticker  symbol VPG, which makes the foil strain gauge,   a hundred year old piece of precision engineering  that is the actual sensing element inside most of   the force and load sensors going into robot  wrists. It may be the only American pure play   on force sensing that you can buy, but the  fundamentals tell a little bit of a cautious   story. Its revenue has sat flat at around $300  million for years, and its earnings just slid to a   cyclical low, from 2.6% a share at 2022 peak down  to about 40 cents, yet the stock still re-rated to   an all-time high on humanoid hope alone, even  though that humanoid revenue is under 1% of   its sales. Now we'll move on to the site for  a robot with Cognex, the ticker symbol CGNX,   which is the world leader in machine vision. The  technology that gives a robot its eyes, and what   sets it apart is the AI that's inside. Its edge  learning software trains the system to tell a   good part from a flawed one off just a handful  of images, with no data scientists required. And   its newest cameras put that AI directly onto  the camera itself, one line running on ships,   and another on Nvidia. That turns a Cognex camera  from a simple sensor into an edge computer that   thinks at the exact spot it sees. And once one  of those systems is wired into a production line,   ripping it out means re-validating the whole  line, so it stays for years. It's the kind of moat   that four decades of vision software and a brand  synonymous with the category has really built up   for itself. And over the past decade, Cognex has  roughly doubled its revenue up to nearly a billion   And while like many of these, it moves in cycles,  it is climbing out of its 2023 low, with revenue   up more than 24% and earnings more than doubling  from the year before. And the entire recovery? It   comes from a single end market, logistics and  warehouse automation. Now its largest sector,   growing double digits for nine quarters straight  as e-commerce keeps building out. And none of   that has even touched humanoid robots yet. Every  one of these machines still answers to a brain.   And even though the brain is only about 6%  of the robot's cost, it is crucial to making   them mainstream. And here, Nvidia wins on both  ends. On the robot itself, it sells Jetson Thor,   a small onboard computer that lets the machine  see, think, and react in real time. And back in   the data center, it sells the second brain. It's  the one that trains these robots and coordinates   whole fleets of them. And those data centers?  Well they already run on Nvidia chips. Of course,   everybody is chasing a cheaper brain and the  fight has really started, with Qualcomm building   a humanoid specific chip called Dragonwing. And  the big automotive names, NXP, Texas Instruments,   and Infineon, all now shipping silicon aimed at  robots. But if you want the fastest learning,   most efficient brain on the market, well, Nvidia  still has the goods. And I have been banging that   drum for a very long time, so I'm not going to  get too much into it here. Now when it comes to   the brains of this robot revolution, Nvidia  collects on both ends, while everyone else   is just racing to catch up. Now strangely, one of  the most important features of a robot is the most   basic items you can get. Well, that's a magnet.  Which brings us to the raw material itself of MP   materials with the ticker symbol MP. And it's the  only American company that owns the entire rare   earth magnet chain under one roof, from a mine  in California all the way to a finished magnet.   Magnets are the hidden heart of the whole robot,  because every high torque motor, every joint   that's in a humanoid, and every electric car,  they're all going to rely on permanent magnets.   And if you recall at the beginning, I shared that  China controls roughly 90% of the processing. And   that is precisely why the Department of Defense  stepped in, and they bought a stake big enough to   become MP's largest shareholder, guaranteed the  company a floor price on its key material for 10   years. And they agreed to buy the entire output of  its new plant. I don't know, in my mind, when the   Pentagon co-signs your business, you're really no  longer just a mining stock. Now I do have to warn   you, this isn't a stock that you just buy based  on its earnings, because there really aren't any   yet. Even as revenue has climbed from under $70  million to about $275 million, MP still loses   money running their business. It's burned through  roughly more than $300 million in cash, building   all of this infrastructure. But what you really  are buying is a factory, and its existing plant   in Texas makes about a thousand tons of magnets  a year. The one going up right beside it is built   for $10,000. That's literally a tenfold jump,  with production targeted for 2028. Like I said,   Washington guaranteed the price and agreed to buy  all of it. And Apple's already signed a deal worth   around $500 million for American-made magnets.  Now moving on, there also has to be somebody   willing to assemble millions of these machines.  And the early leaders are starting to line up,   where J-Bull has become the worldwide production  partner for Aptronics Apollo humanoid. And the   car world is pushing into it too, with auto parts  giant Magna taking stake in robot maker Sanctuary   AI and opening its factories. The companies that  already know how to build complex hardware by the   million tend to win the mass production phase.  These two look to be planting their flags early.   Now we've gotten to the point where our framework  is left with one big hole to fill. And it happens   to be the scarcest, highest value parts in a  robot, and it's the gears and the screws inside   the joints. And this is where America happens  to be the thinnest. On the gears, you can at   least own the best in the world over the counter  with Harmonic Drive and Nabtasco that I mentioned   earlier. But the roller screw, which can run up  to a third of the cost of a linear actuator robot,   is one of the tightest choke points in the entire  machine. Now the ultra precision specialists,   Rolvis and GSA in Switzerland, they happen to be  privates, you can't invest in them. But two of the   giants that make them do trade over the counter,  and both are chasing this market today. And one is   SKF, the Swedish bearings maker building roller  screw modules purpose built for humanoids. And   also Schaeffler, the German giant whose business  makes them and just signed a major deal to supply   the actuators for a whole fleet of humanoids.  Now the same holds true for the rails that those   actuators slide along. It's linear guides, led  by Taiwan's Haiwin and Japan's THK. Which both of   them you can also reach over the counter where  I'll try to provide some deeper analysis and   information as I mentioned down below. So is there  an answer to this part here in the US, or at least   on the more open market? And the answer is, sort  of. From my point of view the name to watch here   is Moog with the ticker symbol MOG.A. Where they  do actually build world class actuators today,   but they go into fighter jets and spacecraft and  its robot revenue, well it's essentially zero.   What it does have is the real engineering to break  into this space. So you put it on the watch list,   well not because it has a robot story today, but  for the day that it lands an actual humanoid wind,   which we're all watching for. To me that would be  the signal that America is finally filling this   one part of the robot that it can't quite build  at scale today. So let me take a second and pull   this whole thing together. So we just walked  through the entire machine and the framework   that I've given in many of my videos. We've  got the gears and the motors that move it,   the sensors in the eyes that let it feel and see,  and the magnets it is built from, and the brain   that runs it all. And the point is the one that  we started with. You don't really have to guess   which robot is going to win. Sometimes the smarter  play is the sum of the parts, where you just want   to own everything that goes into them. So when I  look at all of these together, here's my read of   this basket. A lot of these names has already been  running up high because of hype, so you're going   to need some discipline in the game. Regal Rexnerd  is the one that's still reasonably priced and it's   sitting right about where I'd want to own it  today. Then we have Timken, Allegro, Cognex,   and Navanta, and they're all items that I have on  my watch list. And it's true, I feel like they're   terrific businesses whose prices have just gone  up a little bit quicker than the momentum is   actually justified. Next we have MP materials and  Viché precision. And there's speculative swings,   but MP is really betting on America and America  is kind of betting on it. And then lastly we   have Moog, and it's the one to watch to see if  the US can finally begin to close the gap on   actuators. Then of course we have the foreign  names that you can only buy over the counter,   and the reducer makers, the roller screw  and the linear guide makers, and the sensor   names. The data is a little bit harder to pull  together. So I did try to get my complete list,   and of course like I said, I'll share it down  below. I hope you're able to get some value   out of today, and of course, please, I do this  for educational purposes, and in no way is this   financial advice. My whole intent is to try  to help everyone get ahead of the robot boom,   and part of it is just knowing which ones  to watch when the right timing begins to   kick in. So as always, thanks so much for  watching, and I'll see you on the next one.