Three hundred and fifty-six. That is how many stock perpetual contracts Bitget switched on โ 20x leverage, USDT-settled, trading around the clock. The announcement reads like routine product expansion. It is not. It is a synthetic equity book with no closing bell, no circuit breakers, and no named price source. I have spent enough time reconciling team wallets against on-chain distribution charts to recognize when a headline is hiding the only variable that matters. Here it is: the oracle. Bitget does not name it. In a cash-settled perpetual that prints through weekends and Tokyo holidays, the oracle is the product. The leverage, the 356 tickers, the marketing โ decoration bolted onto a single unverified input.
Start with what a stock perpetual actually is. You do not own a share. You own a price exposure, settled in Tether, with no dividend, no vote, no claim on anything. The mechanism is lifted wholesale from crypto: a funding rate tethers the contract to a reference index, and the exchange clears the difference in cash. This is not innovation. Binance ran tokenized stocks until 2021, when BaFin and the FCA leaned in and the product quietly died. Bybit and Gate.io run variations today. Bitget is not first. It is scaling a category that already failed once under regulatory pressure.
Understand the sequence. Crypto exchanges watched spot and derivatives margins compress through two years of sideways chop. Volume is rangebound, spreads are thin, and the fight for the same leveraged trader is a zero-sum knife fight. The escape hatch is new asset classes โ equities, FX, commodities โ sold to a crypto-native book that already trusts the interface. Stock perpetuals are not a response to user demand. They are a response to revenue anxiety. When an exchange cannot grow its core, it grows its menu.
Now the structural problem, and it is not small.
A 24/7 contract on a market that closes five days a week must invent its own price for roughly 70% of the calendar. When the NYSE is dark and the TSE is shuttered, there is no last trade to reference. There is only the exchange's index โ a number Bitget constructs, weights, and can revise. That number is the settlement basis for every liquidation, every funding payment, every stop. No third-party attestation. No disclosed constituent list. Just a centralized figure dressed as a market price.
I have traded through a liquidity freeze โ a flash loan drained a pool my bot was integrated with, and I pulled $30,000 out in minutes because I could watch the reserves bleed in real time. On-chain, I had that visibility. Here, a user holding a leveraged stock perpetual over a weekend has none. They are trusting one counterparty's arithmetic while the underlying market sits frozen and cannot correct it. Impermanence is the only permanent yield; the corollary is that an unverifiable price is the only permanent risk.
Run the leverage mismatch. Traditional equity margin under Reg T starts at 2x. Bitget offers 20x on the same asset class. Equities are lower-volatility than crypto โ that is the entire premise of holding them. Apply 20x to a large-cap that drifts 1% a day and you have manufactured a liquidation engine. A 5% weekend gap โ routine in single names โ wipes a 20x position before the user can click. The product pairs the volatility of a blue chip with the liquidation velocity of a memecoin, and calls it access.
Then the funding mechanism. In crypto, funding arbitrages a deep, continuous spot market. In stocks, funding would arbitrage... what, exactly? During the close, the perpetual and the underlying cannot converge, because one of them is frozen. So the funding rate becomes a free parameter the exchange sets against a price the exchange sets. Arbitrage is just patience wearing a math mask โ but here there is no underlying leg to arbitrage against. The mask is the whole face.
Who is actually on the other side? Not long-term investors. The buyers are Asian-timezone retail hunting US and Japanese equity exposure outside market hours, plus high-frequency desks farming fee tiers. The sellers are hedgers and market makers the exchange either sources internally or does not source at all. If Bitget carries net exposure without a disclosed hedge, it is the counterparty to every position. Cash-settled, centrally cleared, no transparency on the book. That is not a market. It is a house.
And the revenue tells you the intent. A perp on a low-vol equity does not generate carry. It generates churn โ fees from traders who cannot hold overnight without gap risk, so they flip. The product's economics depend on turnover, not conviction, which means the exchange profits most when its users are most anxious. That is a business model, and it is not one that pays you to sit still.
There is a second-order effect the exchange will not advertise. If Bitget sources its hedge in the cash market, it needs prime-broker access, borrow, and settlement โ none of which an offshore Seychelles entity holds cleanly. If it does not hedge, it warehouses the risk and prices it into spreads. Either way, the user pays. A hidden spread is still a fee. The difference is that one appears on the confirmation screen and the other does not.
The indirect read-through lands on BGB, Bitget's platform token. Fee revenue from 356 equity contracts, in theory, feeds the same treasury that funds buybacks and burns. In practice, a single product launch does not move a token's supply curve. The signal is directional, not material: Bitget is diversifying revenue away from crypto beta, which marginally de-risks the platform's earnings base. But the distribution to holders is opaque, discretionary, and centralized โ the same terms as every exchange token. Do not confuse platform growth with holder capture. They are not the same ledger.
Here is the contrarian read, the one the timeline will miss.
Retail will frame 24/7 stock trading as democratization โ access for the trader in Seoul who cannot touch the S&P after the bell. That framing is the trap. The "24/7" label is not a feature; it is the exact vector of the risk. A market that never closes cannot protect you. There is no circuit breaker to halt a cascade, no closing auction to reset a dislocated price, no weekend for the exchange to reconcile its index against reality. You have imported the worst property of crypto โ continuous liquidation โ onto an asset class that spent a century building halts precisely to prevent it.
Liquidity doesn't announce itself before it leaves. On a thin single-name equity, the manipulation cost is trivial: move the reference index in a dead hour, trigger a cascade of stops, harvest the liquidations. The 356-contract roster makes this worse, not better โ most of those tickers are not megacaps. They are mid-caps with shallow underlying books and wide real-world spreads, repackaged as a 24/7 casino with a 20x lever. The depth you see on Bitget is manufactured; the depth that matters sits in a cash market that is currently closed.
Watch the funding rate print in the first hour after a US close. If it dislocates sharply while the cash market is shut, you have your proof: the perpetual is trading a number the exchange alone controls, and the funding is taxing whoever guessed wrong about it. That is the observable test any reader can run.
The smart-money tell is the precedent, not the product. Binance's stock tokens did not fail because demand evaporated. They failed because regulators forced them out. Bitget runs from Seychelles, sells to a global book, and almost certainly geo-blocks the US, UK, EU, and Japan โ the standard offshore posture. That is regulatory arbitrage wearing a product label: operate where the license is cheap, distribute where the demand is real, and pray enforcement stays regional. When I watched Terra unwind in 2022, the lesson was identical โ yield that isn't backed by collateral or genuine revenue is a claim on someone else's solvency. A stock perpetual priced by its own issuer is a claim on the exchange's index, and nothing more.
So what do you do with this.
Treat the weekend as the danger window. Do not carry leveraged stock-perp positions into a market close โ the gap risk is unhedgeable and the liquidation is automatic. If you must trade the product, treat it as an hours-only instrument: enter after the underlying opens, exit before it closes, and never let the oracle price you alone. Track three signals that move first: any FSA, ESMA, or SEC statement naming the category; Bybit and Gate following with 24/7 stock books of their own, which turns a product into a regulatory target; and any wobble in USDT, the settlement rail every one of these contracts depends on. Verify the underlying, too โ if FASTRETAILING and its 355 siblings are mid-caps, the manipulation math is not theoretical.
If you are building a position, size it as if the oracle will gap against you, because it will โ the only question is which weekend. And if you are watching from the sidelines, treat this as a leading indicator for the whole tokenized-equity complex: the product that scales fastest will attract the enforcement that ends it first.
The broader question is not whether Bitget's stock perps work. It is whether any offshore venue can scale synthetic equities without inviting the exact enforcement that killed the last attempt. Volatility is the tax on imagination. Regulation is the tax on pretending a market never sleeps. Strategy is the art of surviving your own leverage โ and the traders who survive this product will be the ones who never let it trade them at 3 a.m.