$Meta Platforms, Inc.(META)$ Meta is back in the AI spotlight. According to the latest reports, Meta plans to deploy 7GW of AI compute infrastructure in 2026, and then double total capacity to 14GW by 2027. That number is massive. For context, some Street-style estimates use roughly $35 billion per GW as a rough AI infrastructure cost assumption. Based on that framework, an additional 7GW could imply around $245 billion of potential capex scale. This is not Meta’s official capex guidance, but it gives investors a sense of how aggressive the plan could be. At first glance, this sounds like another AI spending “horror story.” But the market reaction was more interesting. Instead of only worrying about capex, investors started asking

Two memory-chip stories hit the market this week. On one side, SK hynix’s U.S. ADR offering reportedly drew demand more than 7x the available supply, with proceeds expected to support new facilities tied to AI memory demand. On the other side, China’s Changxin Memory Technologies, or CXMT, is moving ahead with its Shanghai IPO book-building, aiming to raise funds for production-line expansion and technology upgrades. Different markets, different paths, but the same underlying question: Is AI turning memory chips from a cyclical trade into a structural AI infrastructure story? For years, investors mainly watched memory stocks through the old cycle: When will DRAM prices bottom? When will inventories clear? When will the next upcycle arrive? Now the questions are changing: Can HBM demand sta

【NEWS EVENT】 The U.S. Office of Government Ethics (OGE) just disclosed that President Donald Trump executed over 21,000 securities trades in 2025, with a total value between $600 million and $1.86 billion. Trump's team claims these assets are independently managed by third-party institutions through "automated, model-based portfolios," placing them in a so-called "blind trust." Trump averaged 85 trades per market day in 2025, with a net purchase of approximately 300million in U.S. stocks for the full year. His holdings are heavily concentrated inTechnology( $NVIDIA(NVDA)$ , $Microsoft(MSFT)$ , $Apple(AAPL)$ , $Advanced Micro

2026 is already halfway through, and the first half of the year has given investors plenty to talk about. AI remained one of the most important market themes, but the story kept expanding. It was no longer just about GPUs or mega-cap tech. Some stocks kept breaking new highs. Some names suddenly became market favorites after earnings. Some themes looked speculative at first, but continued to attract attention. And for many investors, H1 2026 probably came with at least one familiar feeling: “I saw it… but I didn’t buy it.” Maybe you watched an AI stock rally again and again. Maybe you hesitated before a semiconductor earnings report. Maybe you thought a nuclear or power stock had already gone too far. Maybe you sold too early and had to watch the stock keep climbing without you. At the sam

Gold has broken down. On Wednesday, $XAU/USD(XAUUSD.FOREX)$ fell below the $4,000/oz level for the first time since November 2025. From the record high of $5,594 reached in January, gold has now fallen nearly 29%. London gold tells a similar story. In just 30 trading days, it dropped from around 4,700 to 3,980, a decline of roughly 16%. Although prices rebounded modestly today, with $GLD$ trading around $368, the overall trend has clearly turned lower. Just two weeks ago, when we were discussing DBS's tokenized gold product, gold was still comfortably above 4,500. Now it's already below 4,000. Why Did Gold Collapse So Quickly? Higher rates. Stronger dollar. The historic rally throughout 2025 was built on one core assumption

$Micron Technology(MU)$ reports FY2026 Q3 earnings after the close on June 24, and the timing couldn't be more important. Just last week, markets were rattled by Fed Chair Kevin Warsh's hawkish debut. Global central banks are increasingly leaning toward tightening, while investors have started questioning whether AI valuations have simply gone too far. In short, the market is looking for a catalyst to stabilize sentiment. Tonight, Micron faces that test. Micron has become the market's preferred way to verify whether AI spending remains as strong as investors believe. If cloud providers are still aggressively buying HBM, DRAM, and NAND, Micron's results should reveal it more clearly than almost anyone else. The question isn't just whe

New Fed Chair Kevin Warsh made his first FOMC appearance this week, and his hawkish tone immediately poured cold water on markets, triggering a sharp repricing across global assets. The policy rate itself did not change. The Fed kept rates unchanged at 3.5%–3.75% for the fourth consecutive meeting, with a unanimous 12-0 vote. What scared the market was not the rate decision. It was that Warsh effectively took away the Fed’s “script.” But the story reversed the very next day. On Thursday night, the three major indices rebounded strongly, led by tech: $NASDAQ 100(NDX)$rose 2.48%, $NASDAQ(.IXIC)$gained 1.91%, and $S&P

The major indices sold off yesterday: $S&P 500(.SPX)$ fell 0.57%, $NASDAQ(.IXIC)$ dropped 1.15%. Today started differently. Stocks opened higher, with the S&P up about 0.2%, the Nasdaq Composite up 0.5%, and the Nasdaq 100 up 0.6%, before giving back some gains during the session. Just weeks ago Goldman Sachs was talking about S&P 8000. Now Citadel and PGIM are warning about inflation, rates, and valuation risk. Japan has already begun tightening. The global conversation is shifting from rate cuts back to rate hikes. Just days ago, the Bank of Japan raised rates by 25 basis points to 1%. A few weeks earlier Goldman Sachs was calling for S&P 8000 and raising

The moment $SpaceX(SPCX)$ went public, it threw a wrench into one of Wall Street's favorite labels: the Mag Seven. With its valuation surging above $2 trillion after listing, SpaceX immediately surpassed Tesla and Meta, becoming one of the most valuable technology companies in the world. When a $2 trillion tech giant isn't part of the Mag 7, does the term still represent the leaders of the technology sector? SpaceX Doesn't Fit the Mag 7 Playbook For years, the market's tech leaders were simple: $NVIDIA(NVDA)$, $Apple(AAPL)$, $Microsoft(MSFT)$,

Tonight, the largest IPO in history begins trading on Nasdaq under the ticker $Space Exploration Technologies(SPCX)$. The hype has reached a fever pitch. More than $250 billion in subscription funds have reportedly been locked up, with retail investors alone contributing over $70 billion. Allocation rates are expected to be only 20–30%, while more than 1,000 institutions competed for shares and international allocations accounted for less than 10% of the deal. With a greenshoe stabilization mechanism in place, most investors expect a strong debut rather than an IPO break. The Most Expensive IPO in History SpaceX generated roughly $18.7 billion in revenue in 2025. At a $1.77 trillion valuation, that implies a staggering 94x price-to

The countdown is on. As soon as next week (the market broadly expects around June 12), $Space Exploration Technologies Corp(SPCX)$ lists on the Nasdaq under the ticker $SPCX$. Priced at $135 per share, a $1.77 trillion valuation, raising $74.4 billion — instantly the largest IPO in history. Musk skipped the usual price range and simply named a number. If the stock pops in its first days, he could become the world's first trillionaire. Just how big is this IPO? - $135 a share, $1.77T valuation, $74.4B raised (ceiling up to $86B). That raise nearly exceeds the total of every US IPO of the past two years combined - More than 40% above the company's own $1.25T self-valuation in February - SpaceX rewrites prior record was Saudi Aramco i

US stocks pulled back from record highs, and $Bitcoin(BTC.USD.CC)$ hit a new low, falling below $62,000 — its lowest level since February 6. Strategy sold off a massive holding of roughly $2.5 million in Bitcoin. "Bitcoin's price fell this week because Strategy broke its 'never sell' promise." At almost the same moment, Goldman Sachs raised a whole batch of price targets — S&P at 8000 by year-end, Asian markets revised up across the board. The research reports were unanimously bullish, yet the market took a breather first. What gives Goldman the confidence to be this bullish? $S&P 500(.SPX)$ at 8000 by year-end (about +6% from now), riding on earnings resilien

At Computex, Huang announced Vera CPU entering mass production and RTX Spark crashing into the PC market — NVIDIA is now officially in the CPU business. The market voted with its feet: $ARM Holdings(ARM)$ surged +15.7% to $409, the biggest free-rider winner; $NVIDIA(NVDA)$ +6.3% to $224; while the x86 duo took it on the chin — $Advanced Micro Devices(AMD)$ -1.2% to $510, $Intel(INTC)$ -4.7% to $109. A Barclays report just ranked the winners and losers of this $100B+ CPU war. For years the semiconductor spotlight was on GPUs. But over the past six months, AI wo

US stocks climbed steadily through May and closed the month at fresh record highs. $S&P 500(.SPX)$ finished +5.15%, closing at 7,580 (intraday high 7,599); $NASDAQ(.IXIC)$finished +8.36%, closing at 26,972 (high 27,095); and $NASDAQ 100(NDX)$crossed 30,000 for the first time, closing at 30,333. AI/tech led again. $NVIDIA(NVDA)$ kept sliding after its earnings; Trump's China visit and policy moves sparked a policy-driven rally; and the looming Fed chair transition is set to weigh heavily on the months ahead. S&P at record highs, but extremely divided

The AI industry chain is now clear: money is flowing from computing power into storage, optical communications, data centers, power, and space. The question is — do you pick stocks one by one, or just buy a thematic ETF? Between 2025–2026, 5 new ETFs targeting AI sub-sector themes have just listed, covering different stages of the industry chain from memory to space. Which direction are you bullish on? 1. $Roundhill Memory ETF(DRAM)$ Listed: 2026.4.2 | Today: $60.73 (+0.36%) | Since listing: +119% Theme: AI Memory / HBM / Storage Chips Buying the "current highest-alpha node" of the AI industry chain. Core logic: compute demand from companies like Anthropic is growing 5x faster than storage, making storage the most certain bottlenec

$NVIDIA(NVDA)$earnings are in the rearview. Where does the AI capital cycle flow next? The diffusion map is clear — money is rotating down the stack across 5 stages. The alpha window is different at each one. Stage 1: GPU & CPU $NVIDIA(NVDA)$ + $Advanced Micro Devices(AMD)$. Where the AI wave started, where capital concentrated first. Valuations are high. Most retail investors entered late here. Stage 2: Memory (actively running — highest alpha right now) $Micron Technology(MU)$ HBM demand surge, severe supply-demand imbalance, price and volume both rising.