June’s jobs report is not “lower unemployment = stronger labor market.” It looks more like a good headline masking weaker underlying momentum.
Nonfarm payrolls rose only 57k, roughly half of the 100–120k consensus range, while unemployment fell from 4.3% to 4.2%.
The key detail: labor force participation dropped from 61.8% to 61.5%, and the employment-population ratio also fell to 59.0%. That means some people likely left the labor force, so they are no longer counted as “unemployed,” which can mechanically push the unemployment rate lower.
One bright spot: Professional & Business Services added 36k jobs, but overall job growth was soft and not broad-based.
My read: labor demand is cooling, while the headline unemployment rate is being flattered by lower participation.
Next, watch whether payrolls keep missing expectations, participation keeps falling, and job gains stay concentrated in only a few sectors.
#JobsReport #NonfarmPayrolls #LaborMarket #Macro #Fed #USEconomy #Employment #Investing
BREAKING: Commercial shipping through the Strait of Hormuz has surged to 10 million barrels of crude oil per day, per Bloomberg.
Details include:
1. Traffic has increased sharply since the 60-day "Memorandum of Understanding" was signed
2. Oil volumes through the Strait of Hormuz still stand ~50% below pre-war levels
3. Uncertainty still exists around what will happen with the Strait of Hormuz once the MoU expires
US oil prices are down over -40% from the recent high.
@aleabitoreddit Is Optimus Gen 3 at risk of missing its expected summer launch window?
If so, how much would that delay matter for the broader Physical AI / robotics investment narrative?
$META saying it’s open to renting out AI data center capacity is being read as a super-cycle warning.
I think that’s mostly noise — and likely a dip to add memory exposure. $DRAM $MU
The real issue is Meta-specific: unlike AMZN/MSFT/GOOG, Meta doesn’t have a native cloud business to monetize excess AI capex.
AWS, Azure, and GCP can sell compute directly. Meta has to justify the spend through ads, recommendations, Llama, or new products.
That makes any hint of spare capacity look like “overbuild,” even if underlying AI infrastructure demand is still intact.
Also worth noting: Meta has been early to almost every frontier-tech cycle — social, VR, AI, open-source models — but productization has been uneven. The market is not just questioning capex. It’s questioning execution.
So I wouldn’t read this as “memory cycle is over.”
I’d read it as: Meta has the weakest AI capex monetization backstop among hyperscalers, and the market is using that to de-risk the whole AI infra basket.
For memory, that’s probably opportunity, not exit.
This is legendary for investment research. Bloomberg should be worried. X has the best “value discovery” of investment research materials because people comment, react, and shit about all the latest reports.
I personally place a lot of weight on X’s trendy discussions for my market screenings. It’s perfect for people who learn fast from discussions.
Announcing the hosted X MCP.
Agents now have access to the best real-time information source in the world.
Connect Grok, Cursor, or any MCP-compatible AI tool to the X API without any setup!
Check it out here: https://t.co/5MzPYwGFzD
$HSAI is trading at the Jan 2025 GS target price. They have been raising the TP to $30+. Much cheaper compared to other AI stocks.
Like if you want to see a deep dive report on $HSAI fundamentals.
$HSAI is leading the rise of LiDAR stocks, together with $OUST.
The robotics sector is consolidating at the correction. Let’s wait for the next dip and load up.
$HSAI is leading the rise of LiDAR stocks, together with $OUST.
The robotics sector is consolidating at the correction. Let’s wait for the next dip and load up.
Markets rallied hard on the MoU + VIX collapse, pricing in 70-80% regional peace.
But signed MoU ≠ resolved Strait of Hormuz. The real path is long and the tape will stay choppy.
I'm using every drawdown to add AI. MLCCs are still my highest-conviction lane. (as I laid out in the earlier thread: https://t.co/7o5QmaxCQz)
You adding on weakness or waiting for "resolution"?
$MURATA $TAIYO
#MLCC #AI
1/
Murata just hiked AI server & high-end auto MLCC prices 10-40%, effective July 1.
Stocks moved instantly:
$6981.T (Murata) +9.4%
$6976.T (Taiyo Yuden) +10.85%
The #1 MLCC maker is finally using real pricing power on the exact components AI data centers need most.
This is bigger than a headline. Here's why 🧵 1/5
#MLCC #Semiconductors #AI #PassiveComponents #FinTwit #Investing #AIStocks #SupplyChain #JapanStocks
Strong breakdown on the harmonic reducer + actuator layer and why LeaderDrive has such strong positioning for humanoid scaling.
Another angle on the same bottleneck: inside the harmonic drive itself, the flexspline. Steel still works, but the weight penalty and machining cost at high volume are real constraints.
PEEK carbon fiber composites (~30% lighter) + precision injection molding could change the economics for scale. Thread on the steel vs PEEK path and what to watch in Tesla’s summer prototype reveal:
https://t.co/kAErARL8Gp
1/
Inside a humanoid robot joint, the harmonic drive flexspline cycles thousands of times every minute.
Steel still handles it today. But the weight penalty and the machining cost at scale are the actual constraints everyone’s racing to solve.
#Optimus#Tesla#PhysicalAI #EmbodiedAI #Humanoid
Goldman’s note on Harmonic Drive getting more constructive on the humanoid side and a key North American customer entering mass production lines up perfectly with where the real bottlenecks are.
The next unlock inside those reducers is the flexspline itself. Steel still works, but PEEK carbon fiber composites could bring ~30% weight reduction and finally make injection molding viable at scale.
Thread on the steel vs PEEK path, Tesla’s staged roadmap, and what to watch in the summer prototype reveal:
https://t.co/kAErARL8Gp
1/
Inside a humanoid robot joint, the harmonic drive flexspline cycles thousands of times every minute.
Steel still handles it today. But the weight penalty and the machining cost at scale are the actual constraints everyone’s racing to solve.
#Optimus#Tesla#PhysicalAI #EmbodiedAI #Humanoid
8/
If the PEEK path clears durability and the molding tech matures, it meaningfully changes the cost and performance curve for building millions of dexterous humanoids.
Steel iteration buys time. Material + process innovation buys the slope.
What direction do you expect Tesla to signal in the summer updates?
Continued steel refinement and geometry wins? Hybrid PEEK introduction? Or something else on the joint side?
NFA. Just connecting the technical constraints of high-cycle harmonic gearing with the realities of scaling humanoid production.
1/
Inside a humanoid robot joint, the harmonic drive flexspline cycles thousands of times every minute.
Steel still handles it today. But the weight penalty and the machining cost at scale are the actual constraints everyone’s racing to solve.
#Optimus#Tesla#PhysicalAI #EmbodiedAI #Humanoid
7/
This sits at the exact intersection of material science, precision manufacturing, and embodied AI execution.
It’s why precision reducer names (Harmonic Drive Systems, Leader Drive) and advanced forming/material specialists keep surfacing in serious physical AI supply chain discussions. The bottleneck isn’t just the reducer. It’s what goes inside it at scale.