Rare earths don’t get much attention in AI debates.

But they power everything physical about AI: drones, EVs, robots, advanced weapons.

And one US company is building the entire supply chain — from mine to magnet.

$MP (MP Materials) could be foundational. 🧵 MP Materials is the only US firm that spans the full stack:

✅ Rare earth mining
✅ Oxide separation
✅ Metal production
✅ Finished magnet manufacturing

China dominates 90% of this. MP is bringing it back onshore.

1/
Zuckerberg hasn’t said a word about this.

But Meta ( $META) might be sitting on one of the most overlooked growth levers in all of tech.

It could reshape the future of advertising and knock Google ( $GOOGL) from its pole position — and hardly anyone is talking about it.
🧵
2/
Meta already has one of the best ad engines in the world. Billions of engagement signals across Reels, Stories, Feed.

But conversational AI changes the game:
Instead of inferring interest, Meta AI can capture intent — in real time, through chat.

Why AI Agents Could Reshape Enterprise Economics

1/
AI agents promise to both reduce the input resources needed for back-office tasks and expand output in revenue-generating roles.

In other words, fewer inputs for the same output, and higher output for the same inputs. Let’s see what that does to profit. 2/
Baseline Profit
P0 = R - (M + N + O)

R: baseline revenue
M: cost of revenue-focused employees
N: cost of non-revenue employees
O: other costs

The AI Value Chain

AI is reshaping industries, but to power this revolution, a vast value chain is forming.

From raw materials to AI agents, here’s where the biggest investment opportunities lie. 🚀👇 1/ Mining Materials:

Before AI models can run, we need materials to power the infrastructure:

🟠 Copper → Essential for AI data centers, EVs, and humanoid robotics. Key players: $BHP (BHP Billiton), $RIO(Rio Tinto), & $GLEN (Glencore).

⚛️ Uranium → AI’s electricity demands will drive nuclear energy growth. Key players: $CCJ (Cameco), $UEC (Uranium Energy Corp), $NXE (NexGen Energy)

🔋 Lithium → AI needs battery storage at data centers & edge devices. Key players: $ALB (Albemarle), $SQM (Sociedad Química y Minera)

1/5 AI apps will disrupt enterprise SaaS.

The traditional model of building software is about to be upended.

What does this mean for existing SaaS companies? Most SaaS apps are just CRUD databases with business logic on top. But AI agents can replace human-coded logic, rendering this model obsolete. The implications are huge. 2/5 AI agents can access multiple database backends without bias, leveling the playing field for new entrants. Existing SaaS moats will disappear, making it harder for companies to defend their market share. As AI takes over the logic layer, backend systems will be replaced with AI-native architectures. This will pave the way for a new wave of innovation and disruption in the SaaS industry.

Everyone’s talking about how platforms like $MNDY (monday) are the future of work management, but there’s one big gap holding them back from becoming true enterprise powerhouses.

Let’s dive into what it takes for MNDY to evolve into a system of record (SOR) like $CRM (Salesforce) or $SAP. 🧵👇 Legacy systems like Salesforce, $ORCL (Oracle), and $WDAY (Workday) are entrenched because they serve as SOR — the gold standard for validated, reliable business data.

MNDY? Not quite there yet.

Here’s why platforms like MNDY aren’t SORs today — and why it matters.

A revisit of the $PANW valuation exercise.

$CHKP, $ZS, & $CRWD are the perfect weighted peer group to compare to PANW. On aggregate the peer group matches well in terms of rev & FCF.

For those not interested in the details, just monitor the mkt caps. For more detail, carry on You can see the revenue of $PANW & the total of the peer group are quite close.

Th revenue-weighted P/S divided by growth estimate for the peer group is 0.83. For $PANW it is 0.40.

So, on a forward revenue basis PANW is trading at half the price of the peer group.

1/12 $PANW vs $FTNT.

Here are some interesting contrasts between these two businesses.

Both are great cybersecurity companies that will become $100bn+ giants within a few years imo. 2/12 $PANW’s biz model is built for agility. They have zero inventory management, zero PoP infrastructure management, & also lease all office space rather than invest in real estate. This helps keep operations light & nimble and the lower capex requirements increases cash on hand

1/ $PANW 3Q22 ER overview & thesis recap.

Very few bearish points & lots of bullish points in recent 3Q22 ER. 2/ 29% growth at $5bn+ rev shows $PANW is grabbing more market than pure cloud players growing higher but at a lower base.

Cont. 20%+ product growth highlights hybrid requirements of orgs post-pandemic.

NGS billings cont. hyper-growth trajectory, even as it surpasses $1.6B.

1/15. $PANW: There are lots of #cloudsecurity startups competing w/ PANW, & also several next-gen platform startups (e.g., Orca, Lacework, Wiz).

We think PANW will still prevail as the undisputed cloud security giant. Here are some thoughts why. 2/15. At a high-level we surmise $PANW’s previous and future successes in the cloud are due to their:

a) longstanding expertise in DPI & application-centric awareness/security.
b) container management knowhow that has been largely influenced by $GOOGL recruits.
c) smart M&A.

$FTNT's founder & CEO Ken Xie has demonstrated one of the most incredible long-term visions in cybersecurity history.

20+ yrs ago he envisioned standard CPUs would eventually be inefficient for processing security inspections coz of data volume & encryption. Furthermore he envisioned networking & security would need to converged for optimal efficiency.

1/14. In past 10 years the productivity space has been ripe with innovation.

We’ve seen Comm players (Teams, Slack, $ZM) & Collaboration players ($SMAR, $DBX, $TEAM).

The 3rd C of productivity space is Coordination ($ASAN, $MNDY, AirTable).

Here's a thread about the latter. 2/14. In the 20th century, Scientific Management was the main mode of management.

It entailed simplifying jobs & optimizing them into a predefined series of steps.

Micromanagement ensued & this necessitated rigid org hierarchies.

In 2017 CS & other IBs were bearish on $PANW. And then in 2018 Arora became CEO & totally turned things round, transforming the firm from a dead-end firewall to a cloud security, ZTNA, & XDR leader, with very aggressive M&A strategy.

Here are some thoughts as to why it worked. Usually such aggressive M&A (10 acq. in 2.5 yrs & spending nearly $3bn) destroys LT shareholder value.

1/11. Assessing who is the better at endpoint security - $S or $CRWD – is kind of academic coz both will sustain high growth for many years to come. 2/11. Though, if you agree that better AI/ML models will develop better security defenses, it’s interesting to consider what each vendor is doing to improve their AI.

No doubt the war & rising geopolitical tensions will lead to more ransomware & other cybercrime.

Here is a thread highlighting the key security areas & the associated strongest vendors to prosper.

$PANW, $FTNT, $S, $ZS, $CRWD, $OKTA, Illumio, Netskope War will escalate state-backed attacks >>> more funding going to skilled bad actors.

This funding will pay for increased automation to scan for openings across the web > therefore enterprises w/ global IT sprawl are most vulnerable.

$PANW has the best solution to fight this.

1/ $OKTA thesis. Not advice.

a) An opportunity to exploit coz of investor short-sightedness

b) Core biz is strengthening & could become the most valuable aspect of Zero Trust

c) The cloud-native approach will eventually smother out incumbents in newly entered markets

cont. 2/

d) Relative attractiveness compared to $NET and $ZS

e) Current price looks to be the lower bound of intrinsic range

f) High SBC but we calculate a 5% dilution charge going forward – not excessive if you believe in the long-term thesis

1/8 We’re big believers in the utility of #Blockchain – use cases such as real estate, B2B contracts, supply chains etc.

Though, with the philosophy pertaining to society and the human psyche, can decentralization reach the potential many believe? 2/8 In much of their work, philosophers such as #Nietzsche & novelists like #Dostoevsky (who is studied a lot by philosophers), theorized that utopia was unattainable due to our fundamental desires for greed, power, selfishness, gratification, jealousy, etc.

1/6 $PLTR's technological advantages can be described in a few ways, though we think it's their ontological skills that underpins most of what they can do.

An ontology is a shared vocabulary or language that maps out relationships between things. 2/6 The Internet has advanced thanks to various ontologies.

Without an ontology your search engine wouldn't be able to find & retrieve information from websites.

Ontologies are also needed to make social networks/ecommerce sites work - so people can find other people/products.

1) Here is a short thread outlining the investment case for $INTC. Not investment advice.

$INTC’s EV/EBITDA is 5.5x vs an industry avg of 21x – mainly due to expectation that INTC will disappoint yet again.

These 4 components might make you think differently. 2) $INTC has fallen behind the comp due to its stubbornness to sustain its rigid vertical integration whilst other semi players specialized (either in foundry or design).

The new IDM 2.0 strategy disentangles its foundry biz from its chip design innovation.

1/ Meta ($FB) has been out of favour for the past 5 months. Prospects of the Metaverse hasn’t interested investors enough to overcome the media negativity relating to algos, whistleblowers, & Apple’s privacy policies.

This thread shares some thoughts about Meta & the metaverse. 2/ During the month prior to Facebook Connect on 28th Oct & subsequent Nov rally, the core enabler of m'verse remained flat whilst key players were hyped.

Jan market meltdown leveled things but we see the Oct/Nov action reveals the underappreciation of Meta’s role/potential.

$FTNT investment case. Not investment advice

1) Secular tailwinds. 2) Significant relative mispricing. 3) Moderate DCF undervaluation. 4) Share buyback.

$FTNT's been our best performer in the past 13 months though we're still looking to add on dips. This thread will explain why 1) Secular tailwinds

There are several catalysts + tailwinds that funnel straight into FTNT. And these trends are still very young thus providing a long runway of growth.

This diagram is the basis for the thesis.