Buzzberg Daily report. We scanned every video, Substack article, and research tweet in Buzzberg’s 24h feed. The biggest signal: AI capex is shifting from a “mega-cap spender” story to a full infrastructure supply-chain story. The market is separating the companies funding the buildout from the companies getting paid by it. The cleaner alpha is down the stack: networking silicon, optical interconnects, memory, power distribution, transformers, cooling, grid equipment, onsite power, and physical AI sensors. Today’s map: suppliers over spenders.👇

You can check the entry on every one of his longs on his Buzzberg page or ask our MCP👇

we're opening Buzzberg's MCP in demo access. you can now ask Claude things like: deep dive on $AAOI when did [account] first mention $NVDA what are speakers saying today, and where do they disagree and a lot more. we're rolling it out gradually, so 10 demo codes for now. want one? comment below follow so we can DM it to you.

two big things just landed on sentiment page. 1️⃣ heatmap, rebuilt — the whole market's sentiment in one view, broken down by sector or theme, with a Buzzberg Top 50 filter. 2️⃣ narrative mindshare — what's the market talking about, and how's that shifting? watch themes grow and shrink, click into the names, see who's surging — over 30d, 90d, or year-to-date 👇

Yesterday, before the open, we said nobody agreed on the SpaceX trade. Day one didn't resolve it. It just moved the fight to a new battleground. $SPCX priced at $135, opened around $150, ran as much as 30% intraday, and closed 19% near $161 — a ~$2.1T company on day one. The records are absurd: the biggest IPO in history, ~$75B raised, more than all 38 other Nasdaq IPOs this year combined. SpaceX instantly became one of the 7 most valuable public companies on earth — bigger than Tesla. It was the biggest retail IPO debut on record, beating Coinbase, and made @elonmusk the world's first trillionaire. Around 4,400 SpaceX employees became millionaires overnight. Not just executives. Not just engineers. Welders, cafeteria staff, and people who held equity through the private years. So is it worth it? Nobody really disputes the company. They dispute the price. SpaceX is now valued at roughly 94–125x annual sales, on a business that made around $19B last year and reportedly lost around $4B. As @IncomeSharks put it: Amazon earns $77B in profit at a ~$2.5T valuation. SpaceX loses money at ~$2.1T. That's the bear case in one comparison. The bulls say that misses the point. @jimcramer called it "A " and "the best deal I can remember," with a $2.5T target and a 1–2 year holding period — while warning investors to brace for massive volatility. @AviFelman framed it as a market-structure trade: tiny float forced index buying Elon premium = career risk to short. Bob Greifeld, former Nasdaq CEO, admitted the math doesn't work at ~100x sales — but said investors aren't buying the math. They're buying aspiration. Antonio Gracias, SpaceX board member and the second-largest shareholder after Musk, says he is holding "as long as I possibly can." His frame isn't aerospace. It's the full stack: energy → compute → launch → orbital compute. A $28.5T market up for grabs. @Scaramucci's verdict on the doubters is blunt: people who bet against Elon tend to get wiped out. His Amazon analogy: great companies can look insanely expensive, crash hard, and still reward long-term holders if the platform keeps compounding. The bears don't attack the company. They attack the price. Jim Chanos says Starlink is real, but much of the rest — xAI, Starship, orbital compute, space data centers — is still "hopes and dreams." His sharper point: SpaceX may be shifting from AI software upside into renting compute to Anthropic and Google. That looks less like software and more like a hardware landlord: lower margin, more capital intensive, less magical. That AI money is not imaginary, though. Bloomberg's @EdLudlow flagged the catch: Anthropic reportedly pays SpaceX around $1.25B/month for compute, and Google more than $900M/month — but both contracts can reportedly be cancelled on 90 days' notice. His sharper question: if SpaceX has the compute, why rent it out instead of training its own models? The most sobering bear take came from @moninvestor: hyped IPOs often fall first, then fly. Uber bottomed −70%, Meta −77%, Robinhood −92%, Coinbase −93%, Rivian −95%. His argument: the best entry usually comes later, not on day one. But the smartest trade may not have been $SPCX itself. It was the fallout. Cash rushed into $SPCX and out of the rest of the space trade. $SPCE, $FLY, $ASTS and $RKLB all got hit as capital rotated toward SpaceX. @BurggrabenH called the IPO "a vacuum cleaner" for the whole space sector. That split the desk. One camp is buying the dip in space names — $SPCX at ~$2T while $ASTS sits around $32B makes the gap look unreal, as @retail_mourinho pointed out. The other camp thinks capital keeps rotating toward the cleaner, larger headline leader. The real fight comes later. Forced buying is bullish: MSCI already moved, Nasdaq 100 could follow, S&P 500 may come later. Price-insensitive demand hitting a tiny float. Real supply is bearish: when insider shares unlock, the market finally sees how much stock wants out at a $2T valuation. Day one went to the bulls. The lockup is the bears' turn.

the post-type filters might be the most underrated thing we've shipped. research / portfolio updates / stock lists / news — one tap each. flipped on "portfolio" today — top of the feed: @seedy19tron 's weekly biotech update.👇

$SNDK is @aleabitoreddit 's #2 call on buzzberg, right after $SIVE. first mention in our tape: nov 25, ~$220 — "memory beneficiary of exponential AI capex." today: knocking on $2000. full stats every $SNDK mention are on his speaker card

5/ Bob Greifeld, ex-CEO of Nasdaq, asks the question nobody wants answered: what happens when lock-ups expire and real supply hits the tape? @cryptobyrde brings a calculator: at $2T, even a "small" 7% float is enormous dollars. The low-float story dies on contact with arithmetic.

Every take in this thread — sources, quotes, tickers — aggregated here: buzzberg.ai. And yes, we'll be tracking who was right.