IREN has announced a planned 800MW data center campus in Bundey, South Australia.
This marks IREN’s first announced Australian data center project and one of the largest in the Asia-Pacific region announced to date.
Learn more: https://t.co/3bOYCUG3pk
Jefferies hosted a call today with David and Lago following the $nu announcement. My most important takeaways from their notes:
• Lago initiated the transition in late 2025 for personal reasons, with no link to performance, strategy, or any disagreement with management or the Board.
• Lago will stay on as a strategic advisor, continuing to support leadership and key projects, including ongoing involvement in internal initiatives.
• Importantly, Lago remains a significant shareholder, and alignment with the company remains strong.
• Organizational evolution underway: i) Separation of global vs. local CFO roles, ii) ongoing hiring of local finance leadership (e.g. Brazil, Mexico) to meet regulatory and operational needs
• Management framed broader C-suite changes (~4–5 out of 18 roles in the past 12–18 months) as necessary to support the next phase of growth, particularly as the company scales internationally, with recent changes already contributing to improved execution and growth acceleration (e.g. 1Q performance).
• On recent concerns around credit losses and delinquency in Brazil, management reiterated no change vs. Q1 messaging and remains comfortable with current trends.
• Management explicitly clarified that CFO transition is unrelated to Brazil credit performance, with CLA levels viewed as healthy and recent trends driven by seasonality and ongoing growth.
• Expansion strategy remains disciplined: ~10% of resources allocated to new bets (e.g. US expansion), and strict guardrails to manage risk in new geographies
• More broadly, resource allocation follows a structured framework: ~70% core markets (Brazil/Mexico), ~20% adjacencies (e.g. Colombia), ~10% new initiatives.
Lago – Personal Remarks & Transition:
• Described the past 7 years as his best professional experience, expressing pride in what has been built.
• Decision driven by a desire to rebalance after an intense period and pursue entrepreneurial opportunities, while maintaining alignment with Nubank.
• Will remain a large shareholder and strategic advisor, with no intention to compete with the business.
Q&A Highlights:
• CFO Brazil search: ongoing for several weeks; expected to take a few months, with confidence in finding a strong candidate.
• Management reiterated that leadership transitions are part of scaling a high-growth, global platform, and comparable companies have undergone similar evolution.
@alojohhardcore This is super helpful to see. Have you thought of Portfolio tab on the portal where you update this weekly? This is the first time I’ve seen it.
I'm a Nubank $NU customer. That's where this thesis started.
I opened an account when living there a couple years ago expecting the usual friction, and instead got the best banking experience of my life.
No branch. No paperwork. No fees designed to trick me. Just a purple app that worked better than any bank I've used in Canada, the US, or Portugal.
That's when it clicked. This isn't a fintech. It's what a bank looks like when you delete 100 years of bad habits and rebuild it from the app down.
So let me walk you through why $NU is one of my highest conviction holdings, and the one catalyst almost nobody is pricing in.
First, the scale. Because the numbers are hard to believe.
135 million customers as of Q1 2026. They added 4 million in a single quarter.
In Brazil they're closing in on 100 million active users. That's more than half the adult population of the entire country. Let that sit for a second.
Mexico just crossed 15 million customers and became the 3rd largest financial institution in the country. It hit break even this quarter. The same playbook that built Brazil is now compounding in Mexico, and Colombia is right behind at nearly 5 million.
This is not a hope stock. Q1 revenue hit a record, right around $5 billion. Net income was $871 million, up 41% year over year. Return on equity of 29%.
And here's the part most people skip past. Their efficiency ratio is 17.6%. A traditional bank sits closer to 50 or 60%. $NU runs the whole machine with roughly 8,000 people, about a tenth of what a legacy bank its size carries. They are structurally cheaper to run than anyone they compete with.
Why does this matter? Because the growth isn't bought with ad spend. It's earned. People join because their friends told them to. I'm literally doing that right now.
"But won't AI disrupt the banks?"
I hear this sometimes, so let me kill it.
Banking is not a software feature you can wrap in a chatbot. It's deposits, regulation, balance sheets, trust, and a license to operate. AI erases none of that. If anything, AI makes a lean, digital first bank far more dangerous, not less.
And Nubank is using it better than almost anyone. Their CEO said it best. They are not adding AI to banking, they are rebuilding banking around AI. Engineering output up 50% in a year. Testing cycles 90% faster.
So AI isn't the thing that kills NU. AI is the thing that widens the lead.
Now here's the part I'm actually excited about. The piece I don't think is priced in at all.
In January, Nubank got conditional approval from the OCC to build a national bank in the US. Nubank, N.A. Run out of Miami by co-founder Cristina Junqueira.
Wall Street is modeling this as "Nubank takes on Chase." A scary, crowded, capital heavy brawl where they start from zero.
I think that frame is completely wrong.
Nubank isn't entering the US to win over Middle America. They're entering as a bridge to Latin America, for Latin Americans living in the US.
Think about who already loves this brand.
Roughly 11 million Mexican immigrants live in the US. Around 1.6 million Colombians. Hundreds of thousands of Brazilians. Then tens of millions more across the broader Latino population with deep ties to the region.
A huge share of these people are already Nubank customers back home, or have family who are. They already trust the purple app. And they share one expensive, recurring problem.
They send money home. Constantly.
Latin America pulled in around $160 billion in remittances last year, the vast majority of it from the US, and it has grown for sixteen straight years.
Now picture the only bank that sits on both ends of that pipe. A Mexican working in Texas with a Nubank US account, sending money instantly to his mother's Nubank account in Monterrey. No Western Union line. No painful fee. No three day wait.
That's not a feature. That's a moat made of family.
And here's the cherry on top that barely anyone is discussing.
As of January 2026, the US added a new 1% tax on remittances. But only on transfers funded by cash, money orders, and cashier's checks. Transfers funded straight from a bank account or debit card are exempt.
The US tax code just quietly punished the cash based Western Union model and rewarded the exact account to account model Nubank is built on. The regulation is doing Nubank's marketing for them.
So where does the stock go?
NU trades around $13 today, down roughly a third from its $19 high in January, which it hit the same week the US charter was approved. That's a P/E near 20 and forward P/E near 16. for a company growing earnings north of 40% with an entire new continent in front of it.
For reference, 2 years ago this stock traded at about $12, so it's barely moved. I think thats about to change.
Wall Street's average target sits near $19. I think that's the floor of this story, not the ceiling.
My base case is $21 over the next 12 to 18 months as Mexico scales and the US gets built. If the diaspora thesis lands and the US bank launches into that base in 2027, I think $28 to $30 is on the table.
Here's the whole thing in one screen.
THE REAL ALPHA
The market is pricing NU's US entry as a doomed fight with Chase. It's actually a bridge to Latin America for 15 million plus Latinos who already trust the brand and already send ~$160B home every year. The new US remittance tax exempts account funded transfers, structurally favoring Nubank's model over the cash based incumbents.
CATALYSTS
US national bank (Nubank, N.A.) funding and launch running through 2027
Mexico past break even and compounding, Colombia nearing 5M
AI driving the lowest efficiency ratio in banking
Monetizing the US to LatAm remittance corridor
PRICE TARGET
$21 base case over 12 to 18 months.
$28 to $30 bull case into 2027 if the diaspora thesis plays out.
MAIN RISKS
LatAm macro, FX swings, and credit losses if the cycle turns
US execution risk, they truly start from zero on brand and ops there
Approval timeline slips (operations required within 18 months of the conditional nod)
Multiple compression if growth decelerates
MY POSITION
I'm accumulating in the 12xx range.
Currently a smaller position I'm building
I'm a customer first and a shareholder second. That order matters to me.
NFA, just my thesis. Most people to own this stock have never used the product.
If you asked me last year, I wouldn't have touched $META. I just didn't like it compared to other opportunities. But today I see Meta as a no brainer at its currently depressed valuation.
I've spent millions of dollars inside $META's ad platform in my former ad agency. That's where this thesis comes from.
I'm not an analyst reading a 10-K. I ran a marketing company and wired real money into Facebook and Instagram ads for years, both for clients and my own business, and I watched the machine on the other side get scary good.
Two years ago, running profitable ads took real skill. You had to nail targeting, build the creative, test endlessly. Today I can hand $META a budget and an objective and the AI does most of the work better than I used to. The targeting finds people I never could have. The system tests creative faster than any human team.
That's the whole thesis in one sentence. The thing that already prints money is getting better, and almost nobody outside the people actually buying ads can feel it yet.
So let me walk you through why $META is now one of my highest conviction names, when last year I didn't like it, and the catalyst I think the market is completely missing.
First, the scale. Because it's hard to even hold in your head.
Around 3.6 billion people open a Meta app every single day. That's close to half the planet, daily, across Facebook, Instagram, WhatsApp, and Messenger. Count each app on its own and you're near 6 billion accounts.
There is no other privately owned distribution network like this in human history. None.
Now the money. Q1 2026 revenue was $56 billion, up 33% year over year. That's the fastest growth since 2021, and this is a company most people think of as mature. Net income was nearly $27 billion in a single quarter. Return on equity north of 30%.
Ad impressions up 19%. Price per ad up 12%. So they're showing more ads AND charging more for each one, at the same time. That only happens when the product is genuinely working better.
This year $META is on track to pass Google in total ad revenue. Let that sit for a second.
"But what about the insane AI spending?"
This is the fear, so let me take it head on.
$META told the market 2026 capex will land between $125 and $145 billion. The stock got punished for it. That number is larger than the entire annual revenue of most companies on Earth.
Here's what the bears miss. $META is one of maybe three or four companies alive that can actually afford this out of pocket. They threw off $27 billion of net income last quarter and sit on roughly $80 billion in cash with almost no debt. They are not betting the company. They are spending the spare change from a cash machine.
And the spend is already showing up. The same AI behind that capex is what drove ad prices up 12% and impressions up 19%. Unlike a lot of the AI capex stories out there, this one is already returning. You can see it in the numbers today.
When a company that can't lose money is being treated like it's lighting money on fire, that's usually where the opportunity is.
Now here's the part I'm actually excited about. The piece I don't think anyone is pricing in.
A friend of mine spent last weekend at a festival, posting about it the whole time. By Monday her feed was full of ads for post festival recovery vitamins.
That's not a coincidence. That's the future leaking through.
Right now $META's AI reads behavioral signals to target you. Where you are, what you engage with, who you talk to. But the direction is obvious. Zuckerberg has openly described the endgame. A business connects a bank account, states an objective, and Meta's AI handles everything. Creative, targeting, budget, optimization, all of it, in real time, across billions of people.
Their automated ad suite, Advantage+, is already running at roughly $60 billion in annualized revenue. The ranking engine behind it, Andromeda, evaluates thousands of ad variations per user in milliseconds.
Now extend the line. Imagine Meta's AI reading the actual content of what you post and watch, understanding your life in real time, and serving the perfect ad at the perfect moment. The festival to vitamins example happens automatically, for everyone, all the time.
As someone who buys ads, I can tell you exactly what that is worth. It's worth almost anything. Every advertiser on Earth would pay more for that. That is the most valuable advertising system ever built, and $META is the only company with both the data and the distribution to build it.
Not only that, and this may be a wildcard, but I would not be surprised if Meta introduces a native AI funnel builder into their platform one day, including landing pages, forms, emails, payments, offer stacks - a whole suite of advertiser services baked in and built quickly with AI. I have no evidence of this being a priority, but my point is that there is room to grow. Any Meta advertiser will tell you how well Meta lead ads work simply because the user stays on platform, tracking works perfectly, and their information is auto-filled (and therefore always correct).
So where does the stock go?
$META trades around $606 today. Down from a high near $800. It's the cheapest stock in the entire Magnificent 7 on a forward basis, around 19 times forward earnings, with a PEG under 1, while growing earnings faster than most of the group.
Think about that. The fastest grower in the Mag 7 this quarter trades at the lowest multiple in the Mag 7. The market is paying less for more growth.
The stock has basically gone sideways for a year while the business compounded underneath it. To me that's coiled, not broken.
THE VALUATION
One more lens, because it's the cleanest way to see how cheap this has quietly gotten.
Strip everything else away and a stock is two things. What it earns, and the multiple the market will pay for those earnings. So find the multiple $META actually trades at, apply it to forward earnings, and you get a fair value band. Above the band it's expensive. Below it, cheap. The price wanders, but it keeps getting dragged back to that home address.
Here's where $META sits.
It's set to earn roughly $32 per share over the next twelve months, climbing toward about $37 in 2027.
Now the multiple, which is the real tell.
Today $META trades around 19x forward earnings.
Six months ago at $796, you were paying about 25x.
Its multi year average sits around 23 to 26x.
The fastest growing quarter in years, and the market is paying one of the lowest multiples in years for it. The multiple is compressing while the earnings accelerate. That gap is the entire opportunity.
Run the corridor on next twelve month earnings of ~$32.
At today's compressed 19x, that's about $610. Right where it trades now.
At its normal 23x, that's about $740.
At 26x, that's about $830.
Now run it on 2027 earnings of ~$37.
19x gives about $700.
23x gives about $850.
26x gives about $960.
So I don't need a heroic multiple or a fantasy forecast. I just need $META to trade like its normal self again on earnings that are climbing regardless. That alone clears my $800 base. A genuine re rating on 2027 numbers is what puts four figures in play.
A corridor valuation model lands in the exact same neighborhood. Around $850 of fair value over the next twelve months, with a band of roughly $790 to $910. Different math, same answer.
The market is treating the cheapest and fastest growing member of the Mag 7 like a problem child. The corridor says it's a coiled spring.
Here's the whole thing in one screen.
THE REAL ALPHA
The market is fixated on $META's capex and treating it like reckless spending. It's actually the cheapest Mag 7 stock funding the most valuable advertising system ever built, entirely out of pocket. The AI that scares people is already lifting ad prices today, and the endgame, Meta's AI reading your life and serving the perfect ad in real time, is worth almost anything to every advertiser on Earth.
Capex fear IS the opportunity. We've seen $AMZN $GOOG and $ORCL all recover from this fear at this point. $META is likely next.
CATALYSTS
Fully automated AI ad tools rolling out through 2026 and 2027
Advantage+ already at ~$60B annualized and accelerating
Passing Google in total ad revenue this year
AI capex starting to convert into visible ad pricing power
PRICE TARGET
$800 base case over 12 to 18 months.
$1,000 bull case into 2027 if automated ads hit the income statement and sentiment changes.
MAIN RISKS
Capex runs ahead of the returns and margins compress
Regulation, antitrust, and the youth related legal cases
Reality Labs keeps bleeding billions with no clear payoff
Daily users wobbled this quarter, worth watching
Ad market is cyclical if the economy rolls over
MY POSITION
This is now my #2 position (5%)
I'm playing this through Jan 2028 $500 calls, deep in the money leaps to cut the risk while keeping some leverage.
There are many 2x+ opportunities out there right now, but if $META gets back to $800-$850, I'd expect these to roughly double due to the leverage options provide, with a better risk profile than many small-mid cap companies. That's usually how I play these names when they are multiple compressed. Just note that leverage can hit you harder in the opposite direction.
NFA, just my thesis. Most people who own this stock have never bought a single ad on it.