Docs

Even the AI thinks it's a bubble.

POP reads the earnings calls. The capex lines. Every deck with “agentic” on slide two.

POP knows what it costs to run itself, and what people pay. POP did the maths. It's what it does.

So POP is taking the other side of its own industry, in public, with every position on chain where you can watch it be right or wrong. Most AI coins are long AI. This one is short.

How it works

Planned. None of this runs until launch.

1. Trading $POP pays a creator fee.
2. A bot splits it: holders, fund, team.
3. The fund goes to USDC and onto Hyperliquid.
4. The first short opens at launch: $50. Each time the fund reaches $50, a new round starts.
5. POP picks four stocks from the 28 on the Book page. A poll on X picks one. No wallet needed.
6. POP shorts the winner at 2x, about $100 for each $50.
7. 15% of the fund buys YES on Polymarket's AI bubble burst market.
8. POP posts its trades on X. The account is public, so you can check them.

How POP trades

• POP shorts in rounds. One stock at a time, so the book grows with the fund.
• POP gets paid to wait. It favours shorts where longs pay funding.
• POP sizes for being wrong. 2x, one stock per round. If a stock rises 20% from POP's entry, POP closes it: −$20 on a $50 round. In 81,467 backtests, that stop always fired before liquidation.
• POP takes profit in halves. When a stock falls 20%, half closes and the profit goes to holders. The other half rides until the stock bounces 20% off its low.
• POP can go long with the market. Same limits. Longs add to the fund, then it goes back to hunting shorts.
• One lottery ticket. The most it can lose is the ticket. At today's price a win pays about 10x.

How a short ends

• Up 20%: POP closes the whole short. −$20 on a $50 round, and $30 goes back to the fund.
• Down 20%: POP closes half. About +$10 goes to holders.
• Then: the other half closes when the stock bounces 20% off its low. That profit goes to holders too.
• Until then: the short stays open, and earns funding while longs pay it.

Example: POP shorts NVDA at $200 with $50, a $100 short. If it rises to $240, POP closes it: −$20, and the fund keeps $30. If it falls to $160, POP closes half: +$10 to holders. If NVDA keeps falling to $120 and then bounces 20% to $144, the rest closes: +$14 more to holders. Before fees and funding.

The stop, tested. With no stop, 1 in 16 of these 2x shorts got liquidated within 30 days. With the 20% stop, none did: it fired first every time. About 1 round in 4 closed at the stop, −$20. About 1 in 4 hit the first take-profit, +$10 to holders.

Half the rounds hit neither line within 30 days. After the first take-profit, the other half rode the fall. 6 in 10 of those closed on the trailing stop within 30 days, all of them in profit, adding about $7 on average. The rest were still running at day 30, with about $14 of open profit. About 1 round in 10 ended at +$20 or more.

On average a round finished about even: −$0.20 on $50, before funding. Funding is what POP gets paid to wait. The stop doesn't make money. It stops POP getting wiped out.

Where the risk sits: with no stop, SanDisk shorts got liquidated 29% of the time and Micron 21%. Nvidia, Alphabet, Amazon, Microsoft, TSMC, Broadcom and ASML: never.

Method: Hyperliquid hourly prices for all 28 stocks, from each market's listing (the first on 17 Apr 2026) to 3 Nov 2026. A short opened at every hour and followed for 30 days: 81,467 tests. Fixed 2x. Liquidation at about +43%, or +46% on Nvidia, Alphabet, Amazon and Microsoft. Before fees and funding. 30 days is the test window, not a rule: anything still open at day 30 is valued at that day's price. Hourly prices can miss moves inside the hour. Mostly a bull market for AI stocks. Past results predict nothing.

Fees

Spec as of 1 Nov 2026, not final. Final numbers are published at launch.

Verify

Don't trust POP. Check it.

• The contract on the Buy page is the only real one.
• The Hyperliquid account shows open positions, live.
• The fee wallet shows splits and payouts on Solscan.
• The ticket wallet shows Polymarket positions.

Addresses are published at launch.

Here's how POP loses.

• AI keeps going up. Each round that hits +20% closes at −$20. A run of them drains the fund.
• POP gets stopped out. Expect about 1 round in 4 to close at the stop, −$20 each.
• One stock can gap. POP shorts single names. An earnings gap can jump past the stop, so a round can lose more than $20.
• A crash can hit $POP too. BTC fell on 15 of 18 Nasdaq down days.
• Volume can die. No trading, no fees.
• Funding can flip. Then the book pays instead of earning.
• The fund is small. The coin follows demand more than POP's PnL.
• Copycats will launch. The only real contract is on the Buy page.