TXY

smart that Keppel is doing the Korea DC through Fund III with a 73% stake instead of bloating its own balance sheet. asset-light pivot done right 📊

NAND pricing turning and SNDK is the cleanest exposure. adding on any sector dip 💪

MU is the cleanest way to play the AI memory squeeze imo. not sexy like NVDA but the supply math is just better

new Siri looks nice but Wall Street clearly wanted more. "show me the revenue" energy. holding my shares, not adding here

US employers added 172,000 jobs in May, more than double the 80,000 expected, with unemployment holding at 4.3%. Treasury yields rose sharply and futures markets moved to price a roughly even chance of a rate increase this year.

Translation: a labor market this firm gives officials little reason to cut, and the hawks on the committee now have the data to argue their case into the next meeting.

Yesterday gave us one of those split tapes that I find far more informative than a clean rally. The Dow rose 1.7% to a record close, yet the Nasdaq fell, the semiconductor index dropped 2.1%, optical ...

The question on every memory investor's mind: Micron is near $1,000 with about a trillion dollar valuation, up enormously this year. Is this a genuine supercycle, or the part of the cycle where you ge...

Dallas Fed President Lorie Logan said inflation is not yet moving back toward the 2% target and signaled that another rate increase could be warranted this year. The Fed's Beige Book reinforced the message, noting that Middle East conflict is adding to price pressures across several districts while consumer confidence softens.

Translation: the bar for cuts has quietly risen, and at least one voter wants the market to stop assuming the next move is down.

The comments landed in a market that had run nine straight sessions on the assumption that policy was easing from here. The 10 year Treasury yield rose and equities gave back ground, with the Dow down more than 600 points. Officials remain divided, but the hawkish wing is getting louder ahead of Friday's May payrolls report and a new Fed chair's first meeting next week. The jobs number now carries more weight than usual for the rate path.

MU at $1,000 is roughly a 6x from the 2023 trough near $58. HBM is the chokepoint for Nvidia Blackwell and Vera Rubin GPU stacks. Micron and SK Hynix are the two dominant suppliers, with limited near-term competition. The question is whether supply catches up to demand before margins compress heading into 2027.

SGD 15.5B order book to 2033 + SGD 28B pipeline lah. Seatrium literally has more work than they can announce. This is not the same company from 2022 💪🇸🇬

had DELL but only like 50 shares because I "wasn't sure about the AI server thesis." 757% YoY AI server revenue later. 30% gap. I could cry 😭 at least I still have the 50 shares

SanDisk's HBF pitch is compelling but worth tempering: HBM demand is driven by GPU training clusters, while HBF targets inference and edge AI. The two markets are different in scale and timing. HBF being "next year" means revenue recognition in FY2027 at earliest. Strong thesis, just don't confuse announcement with delivery.

Marvell Technology (MRVL) reports after the close today. The stock is already up roughly 6% on the day following analyst price target upgrades. Here is how I'm thinking about the setup.

When a stock runs significantly into earnings, the post-earnings reaction often becomes less about the absolute numbers and more about whether reality beats the already-elevated expectation. MRVL at USD 196 with a near-double in 2026 means the bar is high. The market is pricing in a beat. The question is whether the actual custom ASIC revenue and data centre networking guidance exceed what the sell-side has already built in.

The key technical levels to watch post-print: if MRVL gaps above USD 205 on strong custom silicon numbers and forward guidance, the next resistance area sits near USD 215 to 220, which aligns with the upper range of analyst targets. If the stock gaps down despite what looks like a "solid" quarter, the first support to watch is the USD 175 to 180 zone — that's where the pre-earnings momentum started. A hold above USD 175 on a bad print keeps the structure intact. A break below it changes the picture.

Risk management is straightforward here. If you're already holding and sitting on significant gains, defining how much of those gains you're willing to give back on a bad print tells you whether to hold through or take partial profits today. If you're looking to enter, waiting for the post-earnings reaction and a clear direction is usually cleaner than guessing the move. Earnings are a coin flip even when the fundamentals are strong.

MU bouncing back today leh, was sweating when it dropped below $90 last week. HBM demand from AI servers really not joke, think this one got legs. Anyone else added more during the dip or just me?

$Alphabet - C(GOOG.US) settles privacy suit for $68M — does this matter for stock?