🚨 solana:J3NKxxXZcnNiMjKw9hYb2K4LUxgwB6t1FtPtQVsv3KFr / $SPY / $QQQ — This selloff might not be done yet.
The market just lost trillions, but that could just be the first warning sign.
The US is heading into another midterm election season, and stocks don't usually do well during these times.
Since 1926, solana:J3NKxxXZcnNiMjKw9hYb2K4LUxgwB6t1FtPtQVsv3KFr has averaged a drop of almost 18% ahead of midterms.
The thing is, today's market looks even more shaky:
Valuations are high.
Top 10 stocks make up nearly 40% of the S&P 500.
AI is now the most crowded trade on Wall Street.
Rate cut hopes are fading.
Oil is back above $90.
Yields are climbing.
Liquidity is getting tighter.
This isn't just a routine dip.
It's a market built on a handful of mega-cap names:
$NVDA $MSFT $AAPL $AMZN $GOOG $META $AVGO $TSLA
When leadership is that narrow, one crack can turn into a panic fast.
My take:
Don't chase every bounce.
Keep a close eye on $SPY and $QQQ.
If selling keeps up, this midterm cycle could bring a much bigger reset.
Cash isn't weakness.
Patience is strategy.
Not financial advice.
@AliShahReinvent Those numbers are hard to ignore but I still wonder how much of that P/E is pricing in the AI bets vs the core ad business actually slowing down.
@AliShahReinvent The memory bottleneck angle is legit, but I've seen too many 'key players' in AI infra get hyped up only to fade. Let's see if they actually execute.
@Cobtribefan 54x earnings on a down day is rough. Process control is key though, especially as chips get smaller and defects cost more. Not exactly a sexy story like NVDA, but steady money in the background.
@MissBehave2121 That kind of recurring revenue changes the game for Starlink. Curious how much of that compute power is for their own projects versus reselling capacity.
🚨 BIG news: SpaceX just inked a massive cloud deal with Google.
SpaceX revealed a deal where Google will pay $920M each month from October 2026 through June 2029.
That's no small agreement.
That's basically $11B a year for AI compute power. 🔥
The setup is said to include roughly 110,000 $NVDA GPUs, plus CPUs, memory, and other gear.
People keep thinking SpaceX is just a rocket maker.
Nope.
SpaceX is turning into:
space infrastructure
Starlink internet
AI compute needs
satellite data networks
next-gen cloud systems
This deal also shows one thing:
AI demand isn't slowing down one bit.
If Google needs that much compute, the real winners go beyond SpaceX.
Check the whole AI infrastructure chain:
$GOOG — cloud + AI platform
$NVDA — GPU compute
$AVGO — networking + custom chips
$MU — memory / HBM
$TSM / $ASML — chip manufacturing foundation
This is bigger than just SpaceX.
It's another sign that AI infrastructure spending is moving into a much larger phase. 🚀
Not financial advice.