A quick update on this account @wesr08
I’ve spent the past period building Ohr Investment Intelligence, an independent research platform focused on structural trends, earnings inflections and company fundamentals.
I’ve now converted this account over to Ohr.
The research is built around three questions:
What structural change is occurring?
Which listed companies are genuinely exposed?
And is that exposure actually showing up in financial performance?
I’ll be sharing selected company research, sector comparisons, structural-theme work and earnings analysis here.
No price targets. No Buy/Sell calls.
Evidence first.
https://t.co/yf3Sl1jTtm
Something important is changing in cybersecurity.
This week alone:
PHP patched two critical CVSS 9.2 flaws capable of unauthenticated RCE.
Node.js patched a high-severity vulnerability that could also permit remote code execution.
Railway disclosed possible exposure of customer environment variables, including API keys and database credentials.
Attackers used stolen Updraft access to create admin accounts and install malicious plugins.
And Reuters reported a hacker using AI agents to autonomously compromise more than 100 websites with minimal human direction.
I don't think the right conclusion is that we're seeing one coordinated global cyber incident.
The more important signal is speed.
AI is starting to compress the attack cycle:
vulnerability discovery
→ exploitation
→ credential harvesting
→ movement across systems
→ repeat
That materially reduces the time defenders have to react.
And I think that changes where some of the cybersecurity opportunity sits.
The areas I'm watching most closely are:
Application security
Identity + secrets management
Cloud workload protection
Software supply-chain security
AI-agent security
Zero-trust access
That puts names like $PANW, $CRWD, $CHKP, $NOW, $ZS, $NET, $CYBR, $OKTA, $S and $NTSK across different parts of the opportunity.
The key question isn't whether cyber risk is increasing. That's obvious.
It's whether AI makes attacks fast and scalable enough that security spending shifts toward the systems capable of controlling applications, machine identities, credentials and autonomous agents in real time.
The product opportunity looks increasingly real.
The next confirmation needs to come through ARR, billings, customer spend and cash generation.
I like how $OCTV is building this IPO base. There’s also a bit more happening underneath the chart.
Octave only separated from Hexagon in May, so it’s early days as a standalone listed company. Its Q2 numbers caught my attention:
• ARR of $1.143bn, up 7% YoY
• SaaS revenue up 21%
• Free cash flow of $93m
• Adjusted operating margin of 29%
Total revenue fell 4%, which needs watching. Management points to the subscription transition as the reason ARR and recurring revenue give a better read on progress. I’d still want to see that translate into stronger overall growth.
The AI angle is what makes me want to dig further.
Octave’s software sits across industrial design, construction, operations and security, with operational context covering more than 50 million assets. Customers using three or more workflows generate around 15 times the ARR per customer of those using just one.
That doesn’t prove adding workflows causes higher spending, but it does show how valuable the broader customer relationships can be.
CoLabs is already working with customers including Bechtel and Fluor on agentic workflows using their operational data. If agents are going to perform tasks inside industrial systems, access to that data, the surrounding context and permission to act could become valuable.
That’s the connection I’m watching with our broader work on securing AI agents, applications and data.
There’s still plenty to prove. We haven’t seen evidence that these agentic products are adding materially to earnings, and revenue remains soft.
For now, I like the setup. If the base keeps tightening and ARR, SaaS growth and cash generation hold up, $OCTV deserves a closer look.
While everyone is focused on $TSLA, $SPCX and the AI selloff, a different part of the market is still showing strength.
The shared theme is increasingly about securing AI agents, applications and data.
That problem is showing up across several layers:
Identity + permissions
$OKTA, with related governance exposure at $NOW
Agent behaviour + runtime protection
$S
Application, cloud + network protection
$NET, $NTSK, $CHKP, $FTNT
Finding + fixing vulnerabilities
$TENB, $QLYS
And the price action is starting to reflect it.
$NOW, $NTSK and $CHKP closed at 6-month highs on 21 October, while several others remain close to recent peaks.
What interests me is that this isn’t one product category.
As AI agents become more capable, they need access to identities, systems, applications and data.
That creates a broader security problem around:
who can act
what they can access
what they can execute
and how that activity is monitored and controlled
The product opportunity looks increasingly real.
The earnings proof is still developing.
The next confirmation is whether this shows up in ARR, billings, customer spend and cash generation.
AI spending is flowing through to semiconductor equipment earnings.
Lam Research just reported a record quarter:
• Revenue: US$8.22bn, up 22% QoQ
• GAAP operating margin: 40%
• Net income: US$2.75bn
What stands out to me is the combination of revenue growth and margins. AI demand and increasingly complex chips are driving equipment spending, and that spending is translating into earnings.
AI demand → more complex chips → equipment spending → earnings.
This is what I look for when tracking a structural trend: evidence that companies exposed to it are actually capturing the value.
Semiconductor equipment is still cyclical, and valuation matters. These results strengthen the operating case for Lam. Whether the shares offer value is a separate question.
$LRCX
The next bottleneck for AI agents may not be intelligence. It may be permission.
Agents are becoming capable of searching, buying, booking and moving through commercial workflows.
But that doesn’t mean companies have to let them in.
Shopify is opening product catalogue, cart and checkout-related functions to compatible AI agents.
Others are taking a different approach:
Amazon is restricting unauthorised shopping agents.
Yelp is pushing toward licensed access.
Delta is restricting external agent bookings.
JPMorgan and Schwab are limiting certain agent account access.
That leaves three possible models:
Open infrastructure — agents transact through APIs.
Closed ecosystems — transactions stay inside the incumbent platform.
Licensed access — agents pay for access to proprietary data or services.
The question is becoming less:
Can the agent do it?
And more:
Who controls the permission layer between the agent and the transaction?
That could become a valuable control point across commerce, payments, identity, consent and fraud prevention.
$SHOP $AMZN $BKNG $YELP
🇨🇳 CHINA’S DRONES TURNED THE SKY INTO A PIXAR FEVER DREAM
Chongqing lit up the night with more than 10,000 drones doing sky art, like a cat messing with a cicada, to a dragonfly chilling on a lotus.
The best part? Literal glowing jellyfish floating midair like it's SpongeBob time.
It looked like a Pixar movie got dropped into real life.
Fireworks could never.
Source: @SpoxCHN_LinJian
BREAKING: Trump administration has allegedly located between 75,000 and 80,000 of the over 300,000 missing migrant children, according to Harris Faulkner of Fox News.
The lady in red has a story:
Her husband abandoned her and the family. She struggled to make ends meet. She battled drug addiction for years. But she didn't give up.
Today, she watched her 40 year old son become the Vice President of the United States.
What a country.
HIGH-DOSE VITAMIN C is the REAL solution, NOT vaccines.
There’s a good reason why Big Pharma hates it:
There's no profit in healthy people.
Vitamin C is cheap, safe, and effective against cancer, infections, and chronic diseases.
Here's the truth: 🧵