fxyz

Cheap AI helps the users, chip sellers lag

Over the next two years, cheaper AI models will help companies that use AI more than companies that sell computing hardware.

Result

If held over the last 3 months (Aug 1, 2026 – Oct 31, 2026): +1.5%. Over the same days: S&P 500 +2.6%, Chip stocks −0.3%.

What it holds

  • Software that uses AI, 27% of the money, +18.1% over 3 months: NOW (ServiceNow), PLTR (Palantir), CRWD (CrowdStrike), IGV (Software ETF), IBM. Software firms add AI to their products, and cheaper models raise their profit.
  • Big tech that buys AI, 23% of the money, −7.3% over 3 months: MSFT (Microsoft), AMZN (Amazon), GOOGL (Alphabet). Giant tech firms spend a lot on AI, so cheaper AI lifts their profits.
  • Apps that cut costs with AI, 18% of the money, −5.7% over 3 months: META (Meta), RDDT (Reddit), NFLX (Netflix), HOOD (Robinhood). Consumer apps use AI to work better and cheaper, so profit grows.
  • Chip sellers lag, 23% of the money, −1.5% over 3 months: against NVDA (Nvidia), against AVGO (Broadcom), against AMD, against SMH (Chip ETF). Chipmakers gain less when AI gets cheaper, so they trail the users.
  • Cash, 10%.

The plan behind it

When AI gets cheaper, companies that use it spend less to do the same work, so they keep more profit. Chip sellers earn money from how much others spend, so they gain less from price drops. This portfolio buys software firms, big tech buyers and apps that use AI. It shorts Nvidia and a few other chipmakers, so it gains if the users do better than the hardware makers. It's a bet on which group does better, not on the whole market going up.

What would prove it wrong: Cheaper AI leads to a flood of new demand, chip makers' profits keep beating the AI users, and the shorts lose money for several quarters.

FAQ

How do I bet that over the next two years, cheaper AI models will help companies that use AI more than companies that sell computing hardware?

This tested idea holds NOW (ServiceNow), PLTR (Palantir), CRWD (CrowdStrike), IGV (Software ETF), IBM, MSFT (Microsoft), AMZN (Amazon), GOOGL (Alphabet), META (Meta), RDDT (Reddit), NFLX (Netflix) and HOOD (Robinhood) and bets against NVDA (Nvidia), AVGO (Broadcom), AMD and SMH (Chip ETF), with 10% in cash. Make a copy to run it on practice money first, at real prices.

What would prove it wrong?

Cheaper AI leads to a flood of new demand, chip makers' profits keep beating the AI users, and the shorts lose money for several quarters.

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