Professional rock-turner. Finding the companies the market hasn't found yet, but can't ignore for long. 10+ years in the markets. Deep dives in the newsletter ↓
I spend all my time in a corner of the market most investors are trained to avoid: small & micro caps.
Here's how I find great companies that the market can't ignore for long. Hidden in the dirt.
THREAD
"Back then, everyone thought they were a genius.
Every -5% and half panicked while the other half talked about a generational opportunity.
Every +5% and everyone screamed bubble. Eventually one of them turned out to be right, by sheer persistence. What a wild time."
One winner can offset all your losses, and those of your descendants.
Studying your losses should be a hobby for when you have too much free time. Your missed gains deserve the real work.
One of my biggest investing regrets.
I discovered Games Workshop $GAW in 2021. Never bought it. Too expensive, no cash, better opportunities elsewhere.
Or so I told myself.
My wife bought it in 2022. Hasn't touched a single share since. It's now 50% of her portfolio. She slept just fine.
I paid the price of my own stupidity. Learned a few lessons along the way.
She makes sure I don't forget them. Rightly so.
GOLDMAN SACHS PROJECTS SPACEX'S AI REVENUE WILL GROW 100X BY 2030
The bank's projections are the foundation of SpaceX's $1.78 trillion IPO valuation pitch, per the Financial Times.
Here's what Goldman is telling potential investors:
The AI growth model:
- xAI revenue projected to grow from $3.2B in 2025 to $322B by 2030
- That's a 100x increase in five years
- xAI lost $6.4B in 2025
- xAI TAM claim: $26.5 trillion vs roughly $2 trillion for Starlink and space operations combined
The full SpaceX revenue picture:
- Total revenue: $18.7B in 2025 to $474B by 2030
- Starlink: $144B by 2030
- Rocket division: $4.1B in 2024 to $8.3B by 2030
- Adjusted EBITDA: $6.6B in 2025 to $352B by 2030
- Free cash flow: negative $13.8B in 2025 to positive $72B by 2031
What has to happen for the AI projections to hold:
- Musk's Grok models must catch up and surpass Anthropic, Google $GOOGL, and OpenAI in coding, cybersecurity, agents, and chatbots
- xAI has been hamstrung by upheaval, with Musk pushing out all 10 co-founders within two years
- Musk has been renting the Colossus 1 data center in Memphis to Anthropic as Grok failed to gain traction
The IPO mechanics:
- Could raise up to $86B
- Goldman Sachs $GS leads
- Morgan Stanley $MS, JPMorgan $JPM, Citigroup $C, Bank of America $BAC, and UBS $UBS also involved
@801010athlete I simulated millions of portfolios:
- Selling winners is the worst strategy
- Selling at random beats selling winners
- Selling losers is the best strategy
- Selling losers is the only one that outperforms buy and hold
@Mr_Derivatives Private valuations are a price set by a handful of investors, buying preferred shares with downside protection built in.
Seller knows more. Seller picks the moment. Buyer's only move is to walk.
So walk.