Michael Jensen
Hello, everyone Quietly and almost unnoticed, a rotation inside the AI trade has been unfolding — and it’s starting to matter. What’s breaking is not AI adoption, but the business model behind large language models. Since late 2022, the OpenAI-linked AI complex (Microsoft, $NVDA (NVIDIA Corporation) Oracle, AMD, SoftBank) rode the promise of subscription-driven AI revenues. The problem: subscriber growth is disappointing, while costs remain enormous. That explains the growing divergence: $GOOG (Alphabet) and its AI supply chain (Google, Broadcom) continue to outperform OpenAI-linked names remain under pressure Google doesn’t need to monetize AI directly — search, YouTube, cloud and ads already pay the bills. OpenAI, by contrast, requires constant capital injections just to survive, which is creating stress further down the AI supply chain. $ORCL (Oracle Corporation) remains a clear weak link. Credit default swaps are rising, the stock has already suffered heavy losses, and recent corporate statements sounded more like defensive reassurance than confidence. Markets eventually understood that refinancing debt with more debt is not a growth strategy — and the rally quickly faded. From a market-structure perspective, this internal divergence is becoming more visible despite resilient indices. $NSDQ100 rebounded sharply from its recent lows, rallying nearly 800 points and reaching ~25,900 during the European session, before slipping back to around 25,836. Key support remains the 25,000 zone, with the next meaningful downside level near 24,850 (January swing low). As long as that area holds, pullbacks remain corrective rather than trend-breaking. $SPX500 briefly tested the psychological 7,000 level, but was rejected so far and is currently trading around 6,985. This rejection doesn’t change the broader trend, but it does highlight hesitation at round-number resistance. First support sits near 6,900, followed by the 6,750–6,800 area if momentum fades. In short: indices are holding up well — but they’re increasingly doing so while entire parts of the AI trade weaken underneath. On the macro side, liquidity conditions are slowly tightening. ISM is back above 50, productivity is improving, and yields continue to grind higher (10Y approaching ~4.3%, 30Y near 5%). Historically, that reduces the urgency for Fed rate cuts and makes liquidity-dependent assets more fragile. That backdrop helps explain the recent violent swings in $GOLD silver and crypto — not trend confirmation, but stress signals. For now, markets benefit from a temporary data vacuum due to delayed labor reports, which may calm volatility short term. Structurally, however, this setup looks late-cycle rather than early-cycle, with selectivity becoming increasingly important. Clear & Simple Recap 🧭 AI is still being used — but some parts of the AI trade are struggling Google benefits from AI without needing subscriptions OpenAI-linked stocks remain under pressure Oracle looks like a weak link as debt risks rise NASDAQ holds above 25,000 — still constructive S&P 500 rejected at 7,000 — hesitation, not breakdown Rising yields mean money is slowly getting more expensive Bottom line: The market isn’t bearish — but it’s no longer forgiving. Index strength masks growing cracks beneath the surface.
1 reply
1 reply
1 reply
1 reply
null
.