BurningTheMidnightOilToMake

ADP is coming out soon.
The market is very conflicted now: good data → Warsh has more confidence to raise rates → tech stocks under pressure. Bad data → recession narrative returns → still under pressure. The Nasdaq, up 20% in Q2, has very little room for error.
VIX at 16.45 looks calm, but with tomorrow's Nonfarm + Friday's market closure, I bet volatility won't be small. Last month ADP was +122k, JOLTS reported 7.59M last night, beating expectations (job openings still there), and consumer confidence missed again. Signals are conflicting, and the market ultimately only looks at Nonfarm.
Toni

Q2 closed. S&P +14%, Nasdaq +20%, Dow at record highs — the strongest quarter since the pandemic; on the surface, the broad market looks fine.
But when you break it down, it gets interesting. BTC halved in a year (120k → 58k), Gold stuck at $4000 for two months unable to break above, 10Y at 4.46%. In the same quarter, a tech stock rally and risk asset selloff can happen simultaneously — this kind of divergence wasn't common before.
Today starts with ADP, tomorrow is Nonfarm. Warsh's first full data week after taking office — 48 hours will basically determine July's move. Last month ADP was +12

Last day of Q2. Yesterday, the Dow closed above 52K for the first time, and the Nasdaq also crawled out of five consecutive losing days — a 2% rebound looks nice, but how much of that is window dressing for the quarter-end? Today's close will be the real Q2 final.
Last night, the Supreme Court blocked Trump from firing Fed Governor Lisa Cook by a 5-4 vote, a ruling more significant than most think. It confirms the independence of Fed governors, meaning it's not that easy for Warsh to push a more hawkish path. Another layer of uncertainty is added to the interest rate path.
A lot of data today:

BTC $59,000. 20-month low.
Last year around this time everyone was shouting "digital gold," now gold is at 4,078, June dropped nearly 14,000 points, the entire crypto market is as cold as an ice cellar. Institutional funds are pulling out, the macro environment is unfriendly to risk assets, with the 10Y at 4.37%. The Warsh dot plot points to rate hikes and the dollar is not weak; telling a "safe haven narrative" under these circumstances is really hard to sell. The only interesting thing is that Nasdaq futures are still holding up during the session, and tech stocks seem to want to form a bott

Futures turned green. S&P +43, Nasdaq +262, Dow +184.
After two weeks of heavy selling, we finally see some signs of a rebound.
But to be honest, this little bit of green on Monday morning doesn't mean much.
Today is a month-end and quarter-end rebalancing day, and institutions shifting positions can produce many fake moves.
Last week, the Nasdaq 100 lost over a trillion in market cap, and the Mag7 collectively dropped $2.8T in June — not something a one-day rebound can fix.
Concentrix reports earnings after the bell tonight, with Wall Street expecting EPS of $2.63. This company has a signific

BTC at 60k, gold at 4,080.
"Digital gold" has dropped 53% from its peak of 126k, while real gold has risen in the same period. Warsh came in super hawkish, the dot plot pointed to rate hikes, tech stocks are selling off—in this environment, gold is doing its job, BTC is not.
Not saying BTC is dead, but when the macro logic shifts from "buy everything with loose money" to "pick and choose with tight money," BTC has yet to prove itself as a safe haven asset. It looks more like a leveraged Nasdaq.
Gold at 4,080—if NFP beats expectations on Thursday and Warsh continues to sound hawkish, there coul

Keep an eye on a few key numbers over the weekend:
S&P closed at 7354, right on the 50-day moving average of 7356. It has been testing this line for three consecutive days—not yet broken, but not bouncing back either.
If the NFP data next week is weak (expected to fall significantly from May's 172K), it could actually break lower.
VIX at 20.12, up 6.5%—not panic levels, but clearly no longer comfortable. Consumer confidence at 48.9, near historical lows—this somewhat contradicts the market picture where 62% of individual stocks are above their 50-day moving averages. Ultimately, macro is pessi

It's Friday. Let's first check who made money and who lost money this week.
The S&P 500 hovered around 7350 for the whole week, with a cumulative -2.1% for the month. It may seem unremarkable, but the sector divergence is severe—MU alone led the storage chip rally, while cloud vendors kept getting hit by the CapEx cut narrative. In essence, this week was a "capital relocation" within AI: money moving from software/cloud to hardware/HBM.
There are a few variables to watch over the weekend:
1. The situation in Iran—WTI dropped from 75 to 69.80 this week, putting pressure on energy stocks.

The most counterintuitive phenomenon this week: gold and Bitcoin falling together.
Previously, these two were "inflation hedges that rise and fall together," but this week reality slapped them in the face — gold price dropped from 5600 to 4010, -28%; BTC broke through 59K intraday last night, with $1 billion in liquidations. On the surface, it's the "dollar + real interest rate" narrative, but I think there's something deeper behind it:
When expectations of the AI capex story collapsing heat up, the first thing liquidity cuts is non-profitable assets.
Gold has no cash flow, BTC has no profit,

SK Hynix is coming to the U.S.
An ADR sized at $29B to $33B is expected to list on Nasdaq in mid-August under the ticker SKHY. Year-to-date, the Korean home-market stock is up 3x; over the past 12 months, it’s gained 800%.
The timing is delicate—MU just finished last night laying out the “long-term contracts + high margins + AI essential demand” narrative, and when SK Hynix shows up right now, it’s basically here to pick the fruit. Who among NVDA, AMD, and GOOG doesn’t need HBM? This name comes in and immediately takes the prime seat as the AI picks-and-shovels water carrier.
But with a

MU delivered outstanding results last night.
Q3 revenue $41.46B (market estimate $35.84B), adjusted EPS $25.11 (estimate $20.78), gross margin 84.9%. Q4 guidance is even more aggressive—revenue $50B±1B, gross margin ~86%, EPS $31. It surged 15% in after-hours, now steady around +5%.
The most critical part is those 16 long-term agreements, locking in volumes for 3–5 years with financial commitments of $22 billion. The CEO said, "I put half my life into these LTAs"—meaning MU’s $1.28 trillion valuation logic has shifted from "cyclical stock" to "AI infrastructure," and the story has completely c

Pre-market futures are still stable, but no one is really paying attention to futures— the entire market is waiting for MU.
A 13% expected volatility means that no matter which way it goes, it won't be gentle. And don't forget, there's PCE tomorrow; whether Warsh's last hawkish dot plot can loosen up will be revealed tomorrow afternoon.
Let's get through this tonight first.
#MU #PCE #美股

MU reports after hours tonight. The market's cards are now laid out clearly — Q3 EPS of 19.72, revenue of 19.72, revenue of $34.5 billion, with a year-over-year profit growth of 932%. HBM capacity will be sold out for the entire year of 2026, with an expected gross margin of 81%.
My hesitation isn't about the Q3 numbers — they are likely to beat expectations. The real question is: is the 81% gross margin a structural change or a cyclical peak? Goldman’s target price of 400 is betting on the latter, believing that the boom-bust nature of storage chips isn't so easily rewritten. But MU is now

SPCX has only been listed for a few days and is already issuing bonds, with a scale of 20 billion dollars, one of the largest initial bond offerings in history.
I keep thinking about one thing: a company with 100 billion dollars in cash on its books, why does it still need to borrow money?
The only possible answer is—Musk is stockpiling ammunition for the next round of space infrastructure and AI computing expansion.
Cash is enough for daily operations, but to develop Starship V3 and the next-generation supercomputers for xAI, 100 billion dollars isn't really a lot.
The bond market's r

The risk of an AI bubble is getting higher and higher!
Using AI to warn about the AI bubble is a magical move that can only happen in 2026.
‍#Aİ #AI泡沫 #GOOG

Accenture fell 18% last Thursday, while MRVL rose 7% that same day. This isn’t a coincidence—it’s a map of where money is flowing in the AI era. A consulting firm that makes money by selling human time—now its clients use AI to do the work themselves. Do they still need you to station 200 consultants on-site for three months just to produce PPTs? In ACN’s earnings report, the line “AI is disrupting traditional service demand”—every white-collar industry should read it carefully three times. At the same time, AI money is rushing into infrastructure in a frenzy—MRVL’s data center chips, MU’s HBM

After a three-day market closure, this week could be the heaviest of Q2. MRVL was added to the S&P 500 today, but passive funds have already been building positions since the June 5 announcement date, so don’t expect the effective date to bring much of a surprise. The real heavy hitters are ahead: Tuesday FDX → the thermometer for the real economy; Wednesday MU → the most honest midterm exam of the AI bull market; Friday Core PCE → the Fed’s next move’s secret weapon. The MU event has me most on edge—if HBM slips even a little, the valuation logic for the entire AI sector will have to be revis

Today the U.S. stock market is closed for Juneteenth, a commemorative day. But I want to talk a bit more about last night's market.
INTC closed up 10% last night, reaching a new all-time high around $134, already up 228% this year. The reason is that Trump’s Truth Social posted that Apple agreed to cooperate with Intel to design and manufacture U.S.-based chips—this completely rewrites the story of "Intel = foundry chip manufacturing failure" from a structural perspective. Coupled with AMD’s Q1 server CPU market share reaching 33% (eating into Intel’s most stable cash cow for 30 years), the en