Pinned Tweet
People see the winners.
They don’t see the process.
Before buying a stock I ask:
Is the business getting stronger?
Is the market underestimating it?
Am I willing to hold through a 30-50% drawdown?
That framework led me to positions like:
• HOOD ~$20
• PLTR ~$34
• TSM ~$165
• GOOGL ~$192
• TSLA ~$189
One thing I’ve learned from studying great investors:
You don’t need to be right on every stock.
Even Warren Buffett’s fortune came from a relatively small number of massive winners.
A few great decisions can outweigh dozens of average ones.
The stock picks matter less than the process.
Find the edge. Keep your money.
1
2
161
Wrote a Substack article that goes further in depth in my thesis about the AI arms race
Check it out below
wiseguyroi.substack.com/p/th…
15
Wise Guy ROI retweeted
Everyone knows I've been bearish on specific parts of the AI trade. Just search my profile, I've been saying it for months.
To be clear, I'm not bearish on AI. I'm bearish on bad business models being priced like good AI companies.
If you break the trade down to first principles, there are two very different categories in the basket right now:
Infrastructure that has to exist regardless of who wins
$ASML - There is literally no alternative. If you want to make an advanced chip, you go through them. It's a monopoly plain and simple.
$SITM - Nobody talks about it, but timing is in every single piece of hardware. It's the lowest level, most defensible layer. It doesn’t matter what happens with data centers, technology is moving towards MEMS with robotics, space, military, communications, ect. Forward P/E in mid 30s now, i think it’s one of the best buys in the market despite throwing the baby out with the bath water.
$NBIS - I am slightly skeptical on the path to profitability as well as valuation but it’s hard to argue that they’re not the best infra play . Owns the data centers, owns the software layer, has a backlog of actual enterprise clients who can't get capacity elsewhere. They have pricing power because they built something hard to build.
Those companies work even if capex slows down 20%. They work if the cycle pauses.
Then you have $CRWV and some other names including the frontier models.
This is the problem with the entire basket. CoreWeave is a neocloud that rents depreciating assets with leverage. $2.5B+ in quarterly revenue, $104B in backlog, but $21-25B in debt today with up to $68B in potential future obligations, a $35-39B capex plan for 2026 alone, deeply negative free cash flow (-$4.7B last quarter on $7.7B capex), and interest expense ($640M in Q2, guided $860-940M in Q3) growing faster than operating income. Adjusted operating income was $128M in Q2 on double the revenue vs $200M a year ago.
And the structure is circular. Nvidia owns ∼7% of CoreWeave, invested $2B in equity in January 2026, CoreWeave uses that money to buy Nvidia chips, Nvidia books it as revenue, and Nvidia contractually backstops $6.3B of unsold CoreWeave capacity through 2032. It's demand pulled forward with debt.
When rates stay higher, or when one contract slips, that entire leverage loop unwinds. And because passive flows and retail traders own AI as a *basket* through ETFs and thematic funds, the unwind doesn't just hit CRWV. It hits SITM, NBIS, ASML, NVDA. All of it gets sold together.
That's the flaw in the current market. We're letting the weakest business model set the multiple for the strongest companies.
Outside of the direct AI trade: the software and distribution layers of $RDDT, $TEM and $LMND will continue to perform. $ODD is likely seen as the biggest hedge for me at least and it is why I have been buying heavily the past month.
I'm not short AI. In the next phase, the market will stop buying "AI" and start buying actual infrastructure. Have no clue how aggressive this will get but I think its exactly what needs to happen.
New write up: The AI Dilemma
This is not a bear case. I want people long AI to do well. I just think we are confusing making money with being right and I want to put the other side on the table.
On hyperscaler capex, circular financing, ect
isolatedalpha.substack.com/p…
Days like today are why I keep a watchlist.
A falling stock isn’t automatically a failing business. If the thesis still holds up and the valuation works, a lower price means a better entry.
Don’t buy because it’s red. Buy because the price fell more than the value.
15
$MU just disclosed a $35.16B buyback authorization. Repurchases can begin Dec. 9.
I own Micron. The next catalyst may be fewer shares—not just higher memory prices.
But authorization ≠ execution.
What matters more next: pricing or buybacks?
AI kills consulting?
Over 65% of $ACN bookings are fixed-price, including outcome-based work.
My buy zone: $190–$200
12-month base target: $248 (~26% upside from $197, plus dividends).
Oct. 14 Investor Day: who keeps the AI savings—Accenture or clients?
1
80
wiseguyroi.substack.com/p/ac…
Here’s the link to my Substack article that goes into further detail on why I like $ACN at the current prices
50
Everyone’s hunting for the next nuclear startup.
DOE announced up to $4.2B for $VST to add 433 MW from plants that already exist.
That’s the moat: licensed, interconnected power already in place.
Would you rather own the operating plant or the reactor waiting to be built?
54
Two AI-power stocks cleared my return hurdle: $NRG and $TLN.
I modeled four companies through 2031, looking for roughly 15% annualized under reasonable assumptions.
Not just growing electricity demand. Cash left for shareholders after paying for that growth.
My Sept. 28 model:
$NRG at $97 → 17.5% annualized
$TLN at $303.50 → 16.8%
$VST at $138.02 → 12.7%
$CEG at $260.43 → 8.4%
These are conditional base-case estimates—not guaranteed returns or updated stock quotes.
The moat I’m looking for:
Operating plants, grid connections, licenses and established sites competitors can’t quickly recreate.
But scarce infrastructure isn’t automatically a cheap stock. The price still has to leave room for shareholders to earn an attractive return.
My approximate entry prices for a 15% annualized base-case return:
$NRG: $108
$TLN: $329
$VST: $124
$CEG: $192
These are model-based entry levels, not future price targets. They change when earnings, construction costs, financing or valuation assumptions change.
NRG came out ahead on valuation. Talen also passed.
Neither is a free lunch. Debt, operating misses and a lower exit valuation can erase that 15% return.
But the research did produce candidates—not just a watchlist of great businesses at prices I can’t justify.
I went into my copper research expecting to find which major miner I wanted to own.
Instead, I found something more important:
The copper thesis may be right.
The stocks may still be wrong.
I modeled BHP, Freeport-McMoRan and Southern Copper using normalized long-term copper around $5.00–$5.25/lb — not today’s ~$6.80+ price.
My hurdle: roughly 15% annualized over 5 years.
None passed.
FCX had the best copper leverage.
BHP looked strongest on a risk-adjusted basis.
SCCO may have the best assets/economics.
But at current valuations, all three appear to require either:
• permanently extreme copper prices
• very generous terminal multiples
• or both
That’s exactly what I don’t want to underwrite.
The lesson:
You can be right about a commodity shortage and still overpay for the companies benefiting from it.
So I’m not abandoning copper.
I’m changing the question.
Instead of:
“Which major copper miner should I buy?”
I’m now looking for:
“Which company has NOT yet priced in $6–$7 copper?”
That’s where I think the opportunity may be.
$FCX $BHP $SCCO
75
Stanley Druckenmiller’s copper thesis caught my attention.
But I think the strongest version of it is slightly different than “there’s no new supply.”
There IS new supply coming.
BHP is targeting ~5% annual copper growth through FY2035.
Antofagasta has projects that could increase production ~30%.
Rio expects to reach ~1M tonnes/year by 2030.
The problem?
It still may not be enough.
The IEA’s latest 2026 outlook still projects roughly a 25% copper supply gap by 2035, even including announced projects.
Meanwhile BHP sees global copper demand rising from ~34M tonnes in 2026 to >50M tonnes by 2050.
Data centers alone could go from consuming roughly 0.5M tonnes today to ~3M tonnes annually.
And AI isn’t even the whole thesis.
Grids.
Electrification.
EVs.
Industrial demand.
Data centers.
All competing for a commodity that takes years to bring out of the ground.
The catch: this is not an undiscovered story.
Druckenmiller himself called copper a “big consensus trade.”
So now the question I care about isn’t:
“Will copper demand increase?”
I think the evidence for that is pretty strong.
It’s:
Which copper company can I buy today where the valuation still works without needing an insane copper-price forecast?
I’m starting with $FCX, $BHP and $SCCO.
That’s the next rabbit hole
1
542
I dug further into $DECK, and I think the real question is becoming clearer.
It’s not simply:
“Is HOKA slowing?”
It’s whether Deckers can turn HOKA + UGG into much larger global brands.
FY26 Deckers sales:
Domestic: +0.2%
International: +26.8% to $2.28B
International was already ~42% of total revenue.
But Q1 FY27 makes the story less obvious:
HOKA: +7.7%
Domestic: +3.2%
International: +8.4%
Operating income: -6%
So I can see both sides now.
Bull case: international penetration still has years of runway, DTC keeps expanding, and buybacks turn moderate business growth into stronger per-share growth.
Bear case: HOKA is maturing faster than expected and the huge international growth rates are already normalizing.
Management is still guiding HOKA to low-double-digit growth for FY27 and raised EPS guidance to $7.35–$7.50.
Next step for me:
Build out what $DECK could realistically earn around FY31 and work backward to the price where I can underwrite ~15% annual returns.
That number matters a lot more to me than how far the stock has fallen.
3
275
I’m trying to build an investing account where every opinion eventually has to face the math.
Not:
“$XYZ is down 50%, looks cheap.”
But:
What can this business realistically earn in 5 years?
What multiple might the market pay then?
What return does today’s price imply?
And most importantly:
What price would make me actually put my own money into it?
Then I want the receipts left up.
Bought it. Passed. Thesis changed. Math was wrong.
No hindsight.
That’s what I’m building with Wise Guy ROI
31
Intuit $INTU down almost 4% today.
still valued by a lot of people like it’s basically TurboTax + QuickBooks.
But look under the hood:
TurboTax grew 7% last year.
Credit Karma grew 20%.
QuickBooks Online Accounting grew 20% in Q4.
The part of Intuit I’m watching most closely isn’t tax software.
It’s whether Intuit can turn its massive financial data advantage into a platform that increasingly decides:
Who gets a loan.
Which credit card fits.
How a small business gets paid.
How it manages cash.
How it handles payroll.
If AI makes that recommendation layer dramatically better, Intuit starts looking a lot different than “the company that sells TurboTax.”
That’s the part of the thesis I think is most interesting.
I’ll be a buyer if it gets around $270. For now I’m still researching it a little deeper
1
318
I wrote down 3 questions before $INTU Investor Day.
Here’s what I got:
1. Are customer relationships getting stronger? → MIXED
Business retention is still 83%.
But management admitted new-customer growth missed expectations, QBO growth slowed, and price was the #1 reason quality DIY tax customers left.
2. Who captures the value from AI? → PROMISING, NOT PROVEN
Intuit spent about $110M on AI tokens last year.
One interesting datapoint: AI-powered invoicing through QuickBooks/LLMs drove 80%+ month-over-month payments-volume growth from June to July.
But I still didn’t get the clean answer I want:
How much incremental profit is AI actually creating?
3. Where does the margin expansion come from? → CLEAREST ANSWER
Intuit expects significant FY27 margin expansion from operating leverage, AI/model optimization and lower expense intensity — although part of the GAAP improvement is simply lapping restructuring.
My takeaway:
The thesis survived Investor Day, but I’m not raising my valuation just because management told a good AI story.
The customer-growth issue is real.
Now I want to see whether AI actually turns into stronger retention, more new customers and higher owner earnings.
1
443
$INTU just reaffirmed FY27 guidance ahead of Investor Day.
That’s not the part I care about most.
I own Intuit, and there are 3 things I want answered today:
Are customer relationships actually getting stronger?
Online Ecosystem revenue grew 19%, but year-end paying customers grew only 3%.
Who captures the value from AI?
Does AI make QuickBooks more valuable — or make accounting platforms easier to replace?
Where does the margin expansion actually come from?
Intuit is guiding to much faster EPS growth than revenue growth.
I wrote these questions down before the event.
Afterward I’ll compare what management actually said with what I needed to hear.
That’s the part I care about — not how many times they say “AI.”
1
812
Wise Guy ROI retweeted
The FULL Process Behind $20,000 -> $70+ Million in Verified Profits!
Market Wizard @ChrisCamillo Who Has Taken $20,000 to $70+ Million WITHOUT using a Chart, WITHOUT using Stop Losses and WITHOUT using Indicators.
One of the World’s BEST Investors Shares His Complete System For FREE.
Chris sits down with Chart Academy for a 20+ Module Step by Step Masterclass on the exact way he trades and how you can use information to formulate trade ideas.
LIVE Right Now 👇
Deckers ($DECK) is one of the most interesting setups I’ve found lately.
HOKA is still growing.
DECK has ~$1.6B in cash.
No debt.
Generates $1B+ in annual free cash flow.
Trades around 10x FY27 earnings.
So why is the stock sitting near a 52-week low?
Because growth is slowing — and the market seems to think HOKA’s best days may already be behind it.
That’s the part I’m digging into now
1
192



