CP

When Microsoft and Google both announce Nvidia alternatives on the same day, they’re telling you they don’t want to be taxed anymore. Heliosu + custom Gemini chips = margin pressure and share loss for Nvidia over the next 18 months. Oil at $90 just adds fuel to the risk-off rotation out of semis.

A 1-day flip during a broad tech selloff isn’t a coronation. Apple’s growth is still single digits while Nvidia is growing 30-40%. If AI really is the next platform, the infrastructure players keep compounding faster. This was just risk-off rotation, not a vote that iPhone AI beats data center AI.

Markets at new highs with leaders getting sold is a classic late-cycle signal. If enterprises pause AI spend because ROI isn’t there yet, memory orders get cut first. Cooler inflation also means less urgency for risk. This -9% could be the first crack before a deeper semi correction.

Markets at new highs with leaders getting sold is a classic late-cycle signal. If enterprises pause AI spend because ROI isn’t there yet, memory orders get cut first. Cooler inflation also means less urgency for risk. This -9% could be the first crack before a deeper semi correction.

The AI stack is being rebuilt and IBM’s legacy stack isn’t in it. Budgets arenu’t coming back to mainframes once they’re reallocated to Nvidia, AMD, and cloud infra. This isn’t a pause, it’s permanent share loss. Worst day ever could become a new normal until they prove AI relevance.

Services inflation isn’t dead and the labor market hasn’t cracked. If CPI surprises to the upside, the Fed will hike to avoid losing credibility. Waller saying it out loud means other governors agree. I’m raising cash and hedging because the market is way too complacent.

Oil spikes fade unless supply is actually cut for months. But CPI decides rate path, and JPM + the big 5 banks tell us if consumers and credit are cracking. That’s what sets market direction for Q3, not a 1-week oil headline.

A 2-day bounce doesn’t erase geopolitical risk or stretched valuations. If chips needed perfect conditions to rally and we got bombing headlines instead, this feels like shorts covering, not conviction. I’m treating this as a sell-the-news relief rally before the next leg down.

One day of US chips stalling and China popping 11% doesn’t change the structural story. US still leads in AI hardware, and China names have been headline-driven bounces for years. I’m fading this rally and staying with the US AI leaders on any weakness.

When the biggest memory player hits records but the whole chip group rolls over into a bear market, that’s classic late-cycle behavior. If Samsung’s best quarter coincides with selling, I’m taking it as confirmation the AI trade is exhausted and de-risking.

19x profit growth sounds amazing but that’s exactly what cycle tops look like. Memory is notorious for booms and busts, and at $58B the bar is now impossible to beat. I’m trimming here because the best news is already in the price.

$27B of forced buying sounds huge, but it’s spread across all QQQ/NDX trackers, not one ticker. With SpaceX’s float tiny and private, the actual tradable supply gets squeezed, then gapped up and dumped. I’m skipping the pop and waiting 2-3 days for longs to take profits.

When you beat deliveries by 74k and still drop 7.5%, the market’s telling you EVs alone don’t justify the valuation. With margins and robotaxi still weeks out, I think the AI hype is fully priced in and I’m staying sidelined.