🚨 Politician increases holdings: $GOOGL $MU $META $AMD become core positions
Another near-million-dollar purchase has appeared.
Rep. Cleo Fields just disclosed approximately $915K in new transactions.
Buy: $915K
Sell: $0
This is not a rebalancing.
It is a concentrated increase in holdings.
📊 The direction of the largest position is very clear
Approximately $250K —— $GOOGL
Approximately $250K —— $MU
Approximately $250K —— META
Approximately $100K —— AMD
Approximately $50K —— LRCX
The structure is quite interesting.
Not defensive.
It's offensive.
🎯 The capital logic revolves around one core concept
AI computing power + semiconductor cycle + platform-level cash flow.
GOOGL —— AI + data centers + advertising cash flow
META —— AI infrastructure + self-developed chips + advertising compounding
$MU —— Key beneficiary of memory bottlenecks
$AMD —— Second-tier GPU provider + data center penetration
$LRCX —— Semiconductor equipment segment
This is a complete AI infrastructure chain.
From cloud to chips to equipment.
💡 Notable details
No scattered small positions.
Nearly equal-weighted heavy allocation across the three major platforms.
This indicates that the core judgment is not based on short-term volatility.
But rather on long-term structural trends.
Of course,
Politician trading does not equate to investment advice....

🚀 Since its public disclosure, the increase has been approximately 400%: No genius needed, just bet on the right 'bottleneck'
To be honest, this isn't about predicting macro trends or godlike stock-picking skills.
Most of it comes down to two factors:
Choosing the right sector.
Following the flow of capital.
Add a bit of luck to that.
The real logic behind making money this year: Bet on 'bottlenecks,' not stories.
The core framework for this year is very simple:
Where are the physical bottlenecks in AI expansion?
That's where the money will flow.
My favorite categories right now:
1️⃣ Memory
$SNDK
$MU
$SIMO
Samsung, SK Hynix
When computing power explodes, HBM, DRAM, and NAND are not optional—they are essential.
Computing power can be scaled up,
but if memory can't keep up, the system will freeze.
This is the bottleneck premium.
2️⃣ Photonics (Optical Modules)
$LITE
$COHR
$AAOI
$AXTI
When data centers enter GW-level expansion,
The 'connection speed' between computing powers has become a core issue.
It doesn't matter how powerful the GPU is,
If optical interconnects can't keep up, latency becomes a fatal flaw.
3️⃣ Electricity / Power Grid
$XLU
Computing power is an electricity-consuming monster.
With each round of data center expansion,
Power grids and utilities will passively benefit.
AI represents a digital revolution,
But at its foundation lies an energy revolution.
4️⃣ Advanced Packaging Capex
$AMKR
$ONTO
$CAMT
$KLIC
$FORM
HBM...