Matt

Tactical technicals + fundamentals (was all-in $TSLA 2015-22). $NVDA and $PLTR in 2023-25. Fib channel specialist. Former tech startup founder and pro gamer.

[Matt @mgreig]

Lots of concern around the AI trade recently - particularly memory $MU $SNDK $DRAM etc.

$SNDK formed a bearish diamond top recently and it's implied target was $1500. Today it hit that and bounced hard. Also, price has followed technical levels almost perfectly albeit more drastically than usual.

$SMH printed a head and shoulders pattern that basically confirmed today - suggesting another -10% move (which is ~15% or so in higher beta names like $SNDK).

Bottom line: recent moves are technically driven rotation unwind from 3+ month parabolic moves upward.

$SNDK $1200 and $1500 were my buy back targets (I sold half at $2100)
$DRAM $40 and $55
$SMH $525

I’m 50% cash with hedged downside (covered calls and long puts) short term. Watching $SKHY Friday.

Long term, I’m still 10/10 bullish on agentic AI (memory/CPU).

[Matt @mgreig]

https://t.co/vgIJJMh9DQ

[Kalshi @Kalshi]

JUST IN: Elon Musk says he thinks SpaceX stock is "insane" buying opportunity

[Matt @mgreig]

Added a long position via $SPCX leaps and cash-secured puts (weekly) today.

I think a bottom candidate is in - especially if the ER comes in remotely decent. Too much downside priced in - I doubt insiders sell as feared.

Also entered long $PLTR via <30 dte calls

[Matt @mgreig]

@elonmusk been doing 95%+ of my vibe coding on @grok Build using 4.5 and it's been very impressive.

But tonight, as of 15 mins ago, Grok 4.5 has been hallucinating hard. For ex. it literally repeated the same copy paste to me for different queries 5-times back to back. Completely lost the plot/context of everything - which hasn't happened for 3 weeks of using it.

Hope this is just a temporary issue - was getting serious work done but now it's unusable.

[Matt @mgreig]

Breaking news

$TSLA might sell China unit to make the $SPCX merger happen - I like the angle @elonmusk

[Matt @mgreig]

This comment could not have aged any better lol

https://t.co/z25QjlmtGr

[Matt @mgreig]

@MartinShkreli Leopold better not be trying to outdo Bill Hwang

[Matt @mgreig]

@MartinShkreli At least he did it in 60 days and not 2

[Matt @mgreig]

@MartinShkreli Leopold better not be trying to outdo Bill Hwang

[Matt @mgreig]

Situational Awareness being acquired by Citiadel explains a lot - and is a textbook example of why you should manage your risk and avoid leverage.

There's no way Citadel wasn't running a play against Jane Street <> SALP with these conveniently timed bluffs.

Citadel knew what they were doing - and add into today's cool CPI - you see a +20-30% day for many AI names.

My scripts/technicals were showing signs of bottom candidates past few days as well. I released my covered calls yesterday for gains.

https://t.co/6AdFrE4mni

[Matt @mgreig]

Seeing some glimmers of a reversal bottom-candidate tonight in overnight trading. $SKHY and its Korean-listed counterpart both hit key levels and bounced +10%.

If FOMC goes well (no hike) and hyperscalers earnings reaffirm capex, I think $115 might've been the bottom for $SKHY

[Matt @mgreig]

@TrendSpider Increase capex? It’s flat - within prior guidance.

[Matt @mgreig]

FOMC holds rates steady despite last minute 30% chance of a hike.

*IF* hyperscaler earnings capex guidance come in as expected ($MSFT will be annual and most significant), I think this is a great bottom candidate.

I was 70%+ cash for a few weeks now, with the 30% in AI related names mostly hedged with covered calls.

Some of that cash likely gets allocated today.

$ALAB ~250 $SNDK ~1000 $SPCX 100-110 $TSLA 292-300

[Matt @mgreig]

$DRAM at $43 and $MU at $780 also gap filled in overnight trading and $SNDK hit $1000. $ALAB at $250 $CRWV at $66

Many, many key levels being tested into FOMC/hyperscaler earnings

[Matt @mgreig]

Seeing some glimmers of a reversal bottom-candidate tonight in overnight trading. $SKHY and its Korean-listed counterpart both hit key levels and bounced +10%.

If FOMC goes well (no hike) and hyperscalers earnings reaffirm capex, I think $115 might've been the bottom for $SKHY

[Matt @mgreig]

Probably the most important chart in the entire market right now is $MU / $SPY

If you aren't using ratio charts (NVDA/SPY was a key leading chart for years), then you're flying blind.

Below are 2 fib channels plotted over 30+ years ago (yes, 3 decades) and still relevant today.

Another textbook head/shoulders setup with price testing the neckline right now.

Will be a massive week with hyperscaler earnings + FOMC. Expect massive directional movement and volatility.

[Matt @mgreig]

Here's a textbook head/shoulders setup $ALAB printed that targets ~251 which was just hit and bounced off of.

Does this mean the selling is over? No.

This means a near-term (possibly longer term) bottom candidate was just reached by many traders who were targets ~251.

This is not a unique example. I see similar setups across many AI-related names.

https://t.co/FWyPuF6SPv

[Matt @mgreig]

Seeing lots of people on X getting wiped out or destroyed.

It doesn’t matter what the forward PE is for MU SNDK etc. right now.

It doesn’t matter what you think is ridiculous and irrational price movement.

It doesn’t matter that there were recent AI deals worth hundreds of billions announced.

What matters right now: the Korea unwind, rates rising to 2007 levels, circular financing, earnings durability and major technicals flashing red.

When stocks rally parabolically 10-50x in a year, always presume the inverse can happen.

Do not let your own bias cloud your judgement. Selling into weakness (and even for losses) is fine. Sell a portion that lets you sleep at night. You can always rebuy that amount for a lower price as a form of hedging if your conviction remains.

[Matt @mgreig]

Seeing lots of people on X getting wiped out or destroyed.

It doesn’t matter what the forward PE is for MU SNDK etc. right now.

It doesn’t matter what you think is ridiculous and irrational price movement.

It doesn’t matter that there were recent AI deals worth hundreds of billions announced.

What matters right now: the Korea unwind, rates rising to 2007 levels, circular financing, earnings durability and major technicals flashing red.

When stocks rally parabolically 10-50x in a year, always presume the inverse can happen.

Do not let your own bias cloud your judgement. Selling into weakness (and even for losses) is fine. Sell a portion that lets you sleep at night. You can always rebuy that amount for a lower price as a form of hedging if your conviction remains.

[Matt @mgreig]

can’t believe I agree with @Jason on latest @theallinpod about @perplexity_ai having the best harness and routine to Grok 4.5, GLM 5.2 etc is much better than paying 4-6x to Claude for similar - or often better - results.

Perplexity confirmed I am one of their highest paying users, so reducing my monthly spend by 75-85% by using Grok 4.5 (Grok Build also good btw) and GLM 5.2 cannot be under appreciated.

When grok 4.6 @elonmusk?

[Matt @mgreig]

@codermatt Interesting. You use agent swarm on high or max?

[Matt @mgreig]

This sell off isn’t fundamental driven. It’s technical and momentum driven into capex hyperscaler earnings and quarterly rotation. Korean unwind, too much easy money, leverage, crowded trade etc.

Doesn’t matter what the forward PE multiple is. It’s a risk off trade right now. Same reason why a stock can crush earnings and sell off. Not everything is by the book multiple based.

Will $SKHY $DRAM $MU be higher in a year? Probably.

Will they be higher next few months? Less likely.

Either hold with conviction or hedge with covered calls/sell some and buy back dips.

Short term voting machine; long term weighing machine.

[Trade Whisperer @TradexWhisperer]

$SKHY 4.87x. You gotta be kidding me.
Even lower than $MU at 5.56x.

At these levels, they can buy back the entire company in just ~4 years. Shortage until 2030.

Pure madness.