1Microsoft's 25% dip has no trend under it, and the day's news flow went to the supply chain instead
Felix, on Felix & Friends, put three famous beaten-down names through the same three questions — is money flowing into the sector, which side of the 150-day line is the stock on, and is this a rest or a crash — and the biggest name came back with the worst answer. , down 25%, is the one he calls a falling knife: weakest of the Mag7 on relative strength, below every major trend line, in a private bear market of its own, and not a dip he would buy despite the enthusiasm for it on social media and an aside that Michael Burry is reportedly bullish. , down 45%, gets "great company, wrong moment" — the crowd has not left, but the trend has rolled over and the stock sits below its 50-day line. , down 40%, is the only one improving: institutional money is coming back and it has clawed above its short-term line, but the 150-day line has not turned up, so on his read it stays a watchlist name rather than a buy.
What makes that more than a bearish opinion is that every one of the three carries the same stated reversal condition — reclaim the 150-day line and turn it up — so it is a wait with a trigger rather than a verdict, and it comes with an expiry, since he also warns of a market shift by the end of July. The weakness of the segment is that it is mostly a teaching framework wrapped around a free-training funnel, and light on fresh data; the drawdown figures and the trend reads are the part a reader can check.
The tape underneath is not uniformly weak. Reuters had the Dow closing at a record on Monday, one session after the tech-led selloff Benzinga covered on the Friday, and the movers in the day's news feed sat almost entirely in the AI supply chain rather than in the platforms buying from it. joined the $1trn club alongside an upgrade to Strong Buy, moved on the optical side of the same build-out, and a combined $2.4trn investment cycle from Samsung and SK Hynix sits behind both. rose 10.8%, 28.7%, 2.65% and 1.27%, while and moved on an $8bn deal between them. The steepest falls were outside technology altogether: down 15.5%, down 14.5%, down 12.5%. So the checklist's first question — is money flowing into the sector — got a yes for the suppliers and a no for the largest buyer of what they make. Read together, the megacap drawdowns look more like a rotation than a valuation reset, and a rotation ends when those trend lines turn, which is the condition he named.
2The humanoid-robot case moved from software to factory capacity
Andrew Kang, chief executive of Robo Strategy, interviewed on The Compound's Live From The Compound, argues the binding constraint on humanoids is manufacturing rather than intelligence: he expects the intelligence problem solved in one to two years and scale in three to four. The numbers he cites from Goldman Sachs are the substance of it — humanoid and quadruped shipments up 250% year on year to roughly 53,000 units, more than 250,000 humanoids a year shipping by 2030, and a 2035 addressable market of $38bn that has been revised up sixfold. His framing is that humanoids productise physical labour the way language models productised knowledge work, against a global physical-labour market of about $50trn. is named the manufacturing leader — "pole position", reportedly building a facility for 10 million robots a year, against rivals in the thousands to tens of thousands.
The vehicle and the argument need separating. Kang's own fund is a listed closed-end fund holding private robotics companies, it trades at a premium to net asset value, and he was on the show promoting it; the show disclaimed any recommendation, and no bear case was put to him, which is why neither nor is logged as a call below. He was at least straight about the labour-displacement problem and the universal-basic-income argument that follows it. What survives the conflict is the shape of the claim: if the bottleneck is factory lines rather than models, the winners are decided by who can build at volume, and that is settled by published shipment counts rather than by demos.
3Float size, not short interest, separated this week's failed squeeze from the next candidate
Ross Givens spent Thursday's failure explaining Monday's setup. had roughly 30% of its shares sold short and the squeeze still failed, because there are 190 million shares out — institutions sold into the rally and the stock was halted on the way down to $7. His next candidate, , carries much the same short interest at 28–30% against a float about a tenth the size, near 19 million shares, worth around $250m — a fifth of GameStop's float value before the 2021 squeeze is the template for — on roughly 500,000 shares of daily volume. Same crowding, different plumbing.
There is a turnaround underneath the mechanics, which is what he says makes the downside survivable: a new chief executive in Mark King, who ran the Taco Bell turnaround, Del Taco sold to narrow the focus, about $99m of debt paid down, the business still profitable, the stock at 0.2 times sales against at 20 times, and an average analyst target near $20 versus about $13. He says he bought around $13 and would add above $14, starting small and adding only on volume confirmation — "worst case I bought a cheap burger stock." Hold that next to the packaging: this is a channel selling a $5-a-year trading service under a video titled "The Next 1,000% Stock?", and a squeeze thesis is momentum-dependent by construction. The mechanical part — that a 19-million-share float behaves differently from a 190-million-share one at the same short interest — is the part that survives the promotion, and the breakout above $14 is what settles it either way.
What the sources recommended
Three single-name calls came out of the day's three videos — one buy and two avoid-the-dips — and each of the three channels carries a promotional flag of its own, so none of them arrives clean.
The buy. from Ross Givens, bought around $13 with adds above $14 and deliberately small sizing: a squeeze bet with a turnaround underneath it, not a valuation call, made on the channel that titled the video "The Next 1,000% Stock?" and sells a paid trading service inside it.
The sells. and from Felix, and both are avoid-the-dip calls rather than shorts — do not buy the 25% and the 45% down, wait for the 150-day line to be reclaimed and turned up. gets the same treatment one notch friendlier: improving, but "too early", so watchlist rather than buy.
Two things that read like calls and are not. The robotics interview is bullish on and on , but it is a single guest promoting his own fund with the show disclaiming any recommendation, so neither is logged. And appears only as the worked example of a squeeze that failed, not as a position.
The risk-on read in equities rests on a truce that gets tested this week
Oil finished the quarter around $70 after its steepest quarterly fall since 2020, and Reuters, CNBC and the Finnhub feed each tie the level to a US–Iran truce they describe as fragile, with talks continuing in Qatar and the administration pressing fuel retailers lower. Cheaper energy is the input behind the risk-on reading in equities, so that truce is carrying more weight than its headline suggests. Payrolls land this week as the week's rate-path read, and quarter-end brings Q3-start pension rebalancing, a flow Reuters' Morning Bid and TradingKey both describe as two-sided rather than directional.
The earnings calendar is thin: (estimate 3.27) and (1.52) after today's close; (4.50), and before the bell on 1 July with and after it; on 2 July. The Quantinuum listing is the first real test of the quantum theme, alongside the SpaceX-listing rotation into space names — the same theme that moved and . One caution about the calendar itself: was flagged as an earnings catalyst in the news feed but does not appear in the next three days of the exchange calendar, so its timing is unconfirmed.
