The host believes the Anthropic IPO faces skepticism in the current environment; nobody believes in it right now, and it needs to be extremely successful to calm market sentiment.
The host criticizes CoreWeave, saying its backlog claim is offset by a horrible balance sheet that is $8 billion short, with bankruptcy risk.
The host does not expect an Iran deal until the 20th to the 27th, setting the market up negatively in the near term; however, if stocks fall further, Trump is more likely to sign a deal faster, after which stocks and bonds should rocket.
The host notes over 50% of CPI categories are above 3% year-over-year, pushing yields higher and making AI data center buildouts more expensive, especially for smaller neoclouds without massive cash flow.
The host believes the AI capex cycle—capital raising by SpaceX, Broadcom and Oracle—creates a liquidity suck, spawning hardware volatility and typically hardware downside, setting the market up negatively until an Iran deal.
The host believes the 10-year Treasury yield topped at 5.27; despite hot inflation data, fear around the AI trade is driving investors into risk-free 5% yields, pushing yields down.
The host believes the labor market is still doing fine for now, with job generation continuing, but early warning signs exist; consumer pressure may show up about a year out, and AI is affecting hiring.
The host says he is also bearish Sandisk, viewing it as a commoditized product facing memory cycle headwinds.
The host says the Beige Book makes the Fed cautious about hiking into a depression even as inflation expectations broaden, lowering rate hike expectations from the morning shock, but the market still sees a December hike as almost certain.
Despite hot inflation data such as PPI, the 10-year Treasury yield has topped at 5.27 and is falling as AI-trade fear drives rotation into risk-free government bonds.
The host maintains that 5.27 is the ceiling for the 10-year Treasury yield and expects it to snap back down, supporting bond prices.
The host is bearish on QQQ near term, expecting the AI capex suckening to keep it falling after losing 725, potentially toward 715 and 700, though he still believes it can reach 800 by year end.
PPI came in far above expectations and was miserably hot across the board, driven by fuel and gasoline; this should push yields back up, which is bad for bonds and creates problems for the Fed.
The host disagrees that gold will rally on an Iran deal, arguing bond strength reduces fiscal deficit worries and gives gold less momentum, bearish on gold.
Broadcom, along with SpaceX and Oracle, is raising large capital, creating an AI capex suck that drives hardware volatility and downside.
The host says after memory fears, AMD benefits from cheaper memory as it packages logic next to memory, bullish for AMD.
The labor market is still doing just fine for now and continues to generate jobs; AI is preventing hiring and destroying some jobs, but overall employment has not deteriorated yet.
The host sees his first red flag in the S&P 500 via SPY, his largest IRA holding, with a confirmed daily downtrend; he will not fully hedge unless SPY loses weekly EMA 12, at which point he would hedge up and be more open to bear trades.
The US dollar has seen a big breakout to new highs and remains strong; it needs to lose the daily EMA 12 to weaken, but is still holding firm.
The S&P 500 is around 7715; it will likely continue lower to test the 7610 support. If 7610 holds, the pullback is limited, with next major support at 7500. The broader market remains at high levels, so the downside is seen as a normal pullback.
Affected by factors such as Chinese memory makers entering the market, semiconductors have adjusted for several days. Despite growing compute investment and strong chip demand, these positives could not stabilize the Philadelphia Semiconductor Index; the index broke down again, the third breakdown since its April doubling, with almost all 30 components declining.