@tengyanAI on X

Chip crash is panic, memory wins, debt is the risk

AI chip demand through 2027 is still solid, the crash was mostly crowded traders and credit fear, and only a funding freeze for debt-heavy AI cloud firms could really hurt.

Result

If held over the last 3 months (Aug 8, 2026 – Nov 7, 2026): −7.6%. Over the same days: Chip stocks −21.7%, S&P 500 −5.0%.

What it holds

  • Memory is sold out, 34% of the money, −18.9% over 3 months: SKHY (SK hynix), MU (Micron), SMSN (Samsung), SNDK (Sandisk). Long contracts and sold-out supply mean memory demand is not going away by 2027.
  • Chip leaders and toolmakers, 34% of the money, −25.0% over 3 months: NVDA (Nvidia), TSM (TSMC), AVGO (Broadcom), AMAT (Applied Materials), LRCX (Lam Research). If demand holds, the chip designers and the firms that build chips keep getting orders.
  • Debt is the weak spot, 12% of the money, +51.7% over 3 months: against IREN, against CRWV (CoreWeave). If credit shuts, heavily indebted AI cloud firms can't pay for deliveries, so they fall first.
  • Cash, 20%.

The plan behind it

The portfolio buys the chip names that get paid first if 2027 demand holds. Memory is the biggest bucket, because SK hynix and Samsung say their 2027 supply is mostly sold. Chip leaders and the firms that make chips come next. Because financing is the real risk, a small short bet on debt-heavy AI cloud firms pays off if credit shuts. A good chunk of cash is kept in case the panic isn't over.

SOX fell 12.2% yesterday, its worst day since march 2020 (covid crash) ppl are fearful because the whole AI trade is being repriced at once. weak qualcomm and arm prints, data center curbs after the midterms + a financing scare. hedge funds were already deleveraging so every bearish headline hit a crowded trade. my humble take is that most of this is positioning plus credit fear running ahead of the demand evidence. demand into 2027 still looks intact to me reading into the tea leaves.. SK hynix has LTAs covering over half of its 2027-28 production capacity. samsung says its 2027 HBM supply is sold out at an agreed target price. contracts aren't cancellation-proof but that is not a market where demand has suddenly disappeared. think about the timelines ahead. the fresh political risk is mostly about future projects. bloomberg reported micron has all but halted work on a taoyuan fab due in 2028. the week's hard fab news points to less future supply, not more. where i think the bears have a real case is financing. $IREN says $6B of its remaining FY27 funding is "subject to market conditions". if credit shuts, deliveries can slip even when supplier books are full. that can become a demand problem. probably similar issues with other neoclouds, and even worse if they're heavy leveraged so my base case is 2027 demand intact.. and most of the damage from positioning could be over (nfa). financing is the channel that can still bite and i'll be watching closely. and 2028 is where the demand debate belongs. will be watching funded delivery schedules and whether those memory contracts actually convert into shipments.

What would prove it wrong: Credit markets shut and AI cloud firms start slipping or canceling chip deliveries, or memory contracts stop turning into real shipments.

FAQ

How do I bet that AI chip demand through 2027 is still solid, the crash was mostly crowded traders and credit fear, and only a funding freeze for debt-heavy AI cloud firms could really hurt?

This tested idea holds SKHY (SK hynix), MU (Micron), SMSN (Samsung), SNDK (Sandisk), NVDA (Nvidia), TSM (TSMC), AVGO (Broadcom), AMAT (Applied Materials) and LRCX (Lam Research) and bets against IREN and CRWV (CoreWeave), with 20% in cash. Make a copy to run it on practice money first, at real prices.

What would prove it wrong?

Credit markets shut and AI cloud firms start slipping or canceling chip deliveries, or memory contracts stop turning into real shipments.

Can I make one from my own idea?

Yes. Write what you believe in plain words at https://f.xyz/create. fxyz asks a couple of questions, builds the portfolio and shows how it would have done over the last 3 months.

Can I turn a post on X into a theory?

Yes. Reply to the post and tag @fxyzai. It tests the view in the post and replies with the 3-month result and a link.

What can it trade?

Stocks (Tesla, Apple, Nvidia), crypto (Bitcoin, Ethereum, Solana), pre-IPO companies (OpenAI, Anthropic), indices (Nasdaq 100, S&P 500), commodities (gold, silver, oil, copper) and currencies (EUR/USD, USD/JPY). One theory can mix them, and bet for or against any of them.

How does it react to news?

On each check it reads the last two days of news on its biggest holdings and says whether the idea still holds. It can make one small change at a time: trim or drop one holding by up to 5 points. On autopilot the change waits 10 minutes so the owner can skip it; otherwise it asks first.

How do I run it with real money?

For real money, tap Make it mine, open your copy and choose Use real money, with an exchange connected: Hyperliquid, Lighter, Lighter (Robinhood), Backpack, Pacifica or Binance. The money stays in your own account there. fxyz can place trades, but it can never withdraw.

Can the AI get it wrong?

Yes. fxyz's AI turns the idea into holdings, and it can get that wrong: add a coin or stock that doesn't fit, or miss one that does. Check the holdings before you put money in. In your own copy, ask its chat to add or drop any of them.

What do the 3-month numbers mean?

What $100 would have done holding this mix over the last 3 months, at real prices. It is a what-if, not a record: past or simulated results are no promise of future ones, and nothing here is investment advice.

Test your own idea on fxyz · More theories