Morgan Stanley sees the robot bearings market growing ~300x by 2050. The nine bearing stocks that matter — $RBC, $TKR, $RRX, SKF, NSK, MinebeaMitsumi, Schaeffler — mapped with live data.
More Macroplane supply-chain guides · Baskets · Investment theses · Macro trends
Bearing stocks are shares of the companies that make precision bearings — the friction-killing rings inside every motor, joint, and rotating axis of every machine on earth. A Morgan Stanley robotics note circulating this month put a number on why they suddenly matter to investors: the robot bearings market grows roughly 300x through 2050, and no robot design ships without them.
That second clause is the whole thesis. GPU roadmaps change, actuator architectures compete, and nobody knows whether the winning humanoid comes from Tesla ($TSLA), Figure, or Unitree. But a small quadcopter drone needs 8–12 bearings, a humanoid needs 70 or more, and nothing on any drawing board removes them. Bearings win regardless of which robot wins.
👉 Track all nine names live: the Precision Bearings basket on Macroplane — an equal-weighted index of the global bearing makers, up 79.2% over the trailing two years.
The note (Morgan Stanley Research, July 2026) makes four claims worth taking seriously:
The obvious caveat: 2050 is a long way off, and the steep part of that demand curve sits in the 2030s and 2040s. What you are buying today is profitable, dividend-paying industrial companies with a call option on robots attached.
The Precision Bearings basket holds the nine listed manufacturers that matter for high-precision segments. The industry is consolidated — the top six global manufacturers control over half of the roller bearing market — and most of the leaders are conveniently public.
| Company | Ticker | HQ | Market cap | Robotics angle |
|---|---|---|---|---|
| RBC Bearings | $RBC | US | $20.2B | High-precision and thin-section ball bearings; aerospace/defense grade |
| Timken | $TKR | US | $7.0B | Engineered bearings plus Cone Drive and Spinea precision drives |
| Regal Rexnord | $RRX | US | $12.3B | Bearings, miniature servo motors, precision motion control |
| SKF | $SKF-B.ST | Sweden | $10.9B | The reference bearing maker; super-precision and magnetic bearings |
| NSK | $6471.T | Japan | $3.5B | Super-precision bearings, ball screws, megatorque motors |
| NTN | $6472.T | Japan | $1.3B | Bearings and actuators; lists robots among its end markets |
| JTEKT | $6473.T | Japan | $4.1B | Toyota-group bearings, machine tools, and gears |
| MinebeaMitsumi | $6479.T | Japan | $10.5B | The world leader in miniature ball bearings |
| Schaeffler | $SHA.DE | Germany | $9.2B | Bearings and planetary screws for humanoid actuators |
$RBC — RBC Bearings is the precision pure-play. Fiscal 2026 revenue of $1.56 billion came with a 43% gross margin and a 22% operating margin — extraordinary numbers for a metal-parts company, earned by selling into commercial and defense aerospace, submarines, and semiconductor equipment where qualification cycles keep competitors out. The market knows it: at roughly $598, the stock trades near 90x trailing earnings, the richest multiple in the group by far. You are paying up for the aerospace/defense compounder; the robotics option costs extra here.
$TKR — Timken is the value end of the same trade, at about 25x trailing earnings on 2025 revenue of $4.77 billion and $394 million of net income. The underappreciated part is what Timken bought over the past decade: its Industrial Motion segment includes Cone Drive and Spinea, both precision gear-drive makers whose products go into robot joints today. Timken was also the best performer in the basket over the past two years, up 113%.
$RRX — Regal Rexnord carries the bearings heritage of Rexnord plus a dedicated Automation & Motion Control segment: miniature servo motors, precision motion control, linear actuators. The complication is the balance sheet. 2025 operating income of $377 million was almost entirely consumed below the line — the residue of the debt-funded Altra acquisition, with long-term debt still around $4.8 billion — leaving a small reported net loss for the year. Q1 2026 swung back to a $57 million profit. This is a deleveraging story with a motion-control business attached.
$SKF-B.ST — SKF is the name the rest of the industry benchmarks against, from Gothenburg. Its catalogue spans rolling, super-precision, slewing, and magnetic bearings plus the condition-monitoring layer on top. If a Western humanoid OEM dual-sources precision bearings, SKF is on the shortlist by default. It was the basket's laggard over the past two years, up 31% — which also makes it the least robot-hyped entry point.
$SHA.DE — Schaeffler is Germany's bearings-and-motion group and the most explicitly humanoid-positioned European name. In our Unitree IPO supply-chain thesis, Schaeffler shows up as the bearings and planetary-screws layer of the humanoid actuator stack — the layer every Western OEM that doesn't verticalize like Unitree still has to buy.
Japan is where precision rotating hardware lives, and four of the nine names trade in Tokyo.
$6471.T — NSK makes super-precision bearings and, just as relevant for robotics, ball screws and megatorque motors — the linear-motion and direct-drive pieces of automation. $6472.T — NTN is the one that names robots outright in its served markets, alongside its electric actuator line. $6473.T — JTEKT, the Toyota-group company, pairs bearings with the machine tools and gears that build everything else. $6479.T — MinebeaMitsumi may be the sharpest robotics fit of the four: it is the world's largest maker of miniature ball bearings, the small-diameter parts that go into compact motors, fans, grippers, and drone rotors — exactly the form factors the Morgan Stanley demand curve is made of.
The basket is up 79.2% over two years, but the dispersion tells you what's actually being priced. $TKR (+113%) re-rated as an industrial with hidden drives content. $RBC earns a 90x multiple for aerospace and defense, not robots. $SKF-B.ST (+31%) still trades like a European industrial cyclical. Nobody in this group carries a humanoid premium the way rare-earth or actuator names do — which is the point of looking now rather than after the first disclosed humanoid design win.
👉 Compare the constituents yourself: open the Precision Bearings basket and flip between the 1Y, 2Y, and 5Y windows — the per-name divergence is the story.
A humanoid joint is a stack: a rare-earth magnet motor, a precision reducer, and bearings supporting every rotating interface in between. We've mapped the other two layers before — the magnet chain in rare earth stocks and the full actuator stack in the Unitree IPO supply-chain teardown. The reducer layer has its own listed champion in Harmonic Drive Systems ($6324.T), whose strain-wave gears are the joint-precision moat.
Bearings are the least glamorous layer of that stack and the hardest to disrupt. Chinese OEMs undercut Western motor prices, and Chinese gearbox makers ship at a third of Harmonic Drive's pricing — but high-precision bearing manufacturing at aerospace tolerances has resisted commoditization for a century. That's why the Robotics & Industrial Automation trend needs a components leg, not just robot OEMs.
| Broad robotics ETFs (ROBO, BOTZ) | Precision Bearings basket | |
|---|---|---|
| What you own | Robot OEMs, software, semis, some components | The bearing layer only |
| Robotics exposure today | Direct but crowded and richly priced | Indirect — profitable industrials with a robotics option |
| Downside if humanoids slip | Multiple compression across the fund | You still own mid-cycle industrial earnings |
| Expense ratio | 0.68–0.95% per year | Free to track on Macroplane |
The honest framing: an ETF gives you robotics beta now; the bearings basket gives you companies that get paid either way and re-rate if the Morgan Stanley curve starts showing up in order books.
Robotics revenue is approximately zero today. These companies earn their money from the 40/30/30 industrial/auto/distribution split. If humanoid timelines slip five years, you own industrial cyclicals at whatever point of the cycle you bought them.
China holds about a quarter of the market and is moving up. Chinese manufacturers already make up roughly 25% of the global bearings market. A Chinese humanoid supply chain — Unitree's model — can source domestic bearings and never touch the names above. The Western/Japanese basket wins on ex-China supply chains and high-precision grades, not on unit share.
Valuation dispersion cuts both ways. $RBC at ~90x trailing earnings prices in a lot of aerospace execution before any robot ships. $RRX still has $4.8 billion of debt to work down.
The 300x is a 24-year forecast. Morgan Stanley's own chart is nearly flat until the early 2030s. The bearings trade is a compounding-with-optionality position, not a catalyst trade.
Bearing stocks are shares of companies that manufacture precision bearings — the components that reduce friction and support rotating parts in every motor and machine. The main listed names are RBC Bearings ($RBC), Timken ($TKR), Regal Rexnord ($RRX), SKF ($SKF-B.ST), Schaeffler ($SHA.DE), and the Japanese makers NSK ($6471.T), NTN ($6472.T), JTEKT ($6473.T), and MinebeaMitsumi ($6479.T).
Morgan Stanley's July 2026 research puts it at 70 or more bearings per humanoid, versus 8–12 for a small quadcopter drone. The count scales with degrees of freedom: every joint, wrist, and finger adds motors, and every motor needs at least one bearing.
It depends on the robot form factor. MinebeaMitsumi ($6479.T) leads in the miniature ball bearings that small drones and compact joints use. Schaeffler ($SHA.DE) is positioned in humanoid actuator components. Timken ($TKR) owns precision drive makers (Cone Drive, Spinea) that ship into robot joints today. RBC Bearings ($RBC) and SKF ($SKF-B.ST) own the highest-precision grades that next-generation robotics will qualify.
No dedicated bearings ETF exists. Broad robotics ETFs like ROBO and BOTZ hold robot OEMs, software, and a handful of Japanese component makers, so the bearing exposure inside them is diluted. The Precision Bearings basket is a free way to track the pure component layer as one index.
Because bearings are required regardless of which robot architecture wins, face very low substitution and obsolescence risk, and scale in content as robots get more complex. The firm forecasts roughly 300x growth in the robot bearings market through 2050 — about 40 billion units and over $250 billion a year — and notes that OpenAI listed precision bearings as one of six critical components in its US robotics manufacturing RFP.
This is not financial advice — it's research and education. Bearing makers are cyclical industrials, forecasts to 2050 carry enormous uncertainty, and several of these names trade on foreign exchanges with different disclosure standards. Do your own work.
Bearing stocks are shares of companies that manufacture precision bearings — the components that reduce friction and support rotating parts in every motor and machine. The main listed names are RBC Bearings ($RBC), Timken ($TKR), Regal Rexnord ($RRX), SKF ($SKF-B.ST), Schaeffler ($SHA.DE), and the Japanese makers NSK ($6471.T), NTN ($6472.T), JTEKT ($6473.T), and MinebeaMitsumi ($6479.T).
Morgan Stanley's July 2026 research puts it at 70 or more bearings per humanoid, versus 8–12 for a small quadcopter drone. The count scales with degrees of freedom: every joint, wrist, and finger adds motors, and every motor needs at least one bearing.
It depends on the robot form factor. MinebeaMitsumi ($6479.T) leads in the miniature ball bearings that small drones and compact joints use. Schaeffler ($SHA.DE) is positioned in humanoid actuator components. Timken ($TKR) owns precision drive makers (Cone Drive, Spinea) that ship into robot joints today. RBC Bearings ($RBC) and SKF ($SKF-B.ST) own the highest-precision grades that next-generation robotics will qualify.
No dedicated bearings ETF exists. Broad robotics ETFs like ROBO and BOTZ hold robot OEMs, software, and a handful of Japanese component makers, so the bearing exposure inside them is diluted. The Precision Bearings basket is a free way to track the pure component layer as one index.
Because bearings are required regardless of which robot architecture wins, face very low substitution and obsolescence risk, and scale in content as robots get more complex. The firm forecasts roughly 300x growth in the robot bearings market through 2050 — about 40 billion units and over $250 billion a year — and notes that OpenAI listed precision bearings as one of six critical components in its US robotics manufacturing RFP. This is not financial advice — it's research and education. Bearing makers are cyclical industrials, forecasts to 2050 carry enormous uncertainty, and several of these names trade on foreign exchanges with different disclosure standards. Do your own work.