PRIMER Updated 2026-09-12 By Ivo Renner

What a TradFi perpetual contract actually is, and what it is not

You can open a position called AAPLUSDT on Binance and watch it move almost exactly like Apple stock. It is not Apple stock. It is not a share, a token, a fund unit or a claim on anything. Getting precise about what it actually is takes about ten minutes and prevents most of the expensive surprises.

The one-sentence definition

A Binance TradFi perpetual contract is a USDT-margined, USDT-settled derivative with no expiry date, whose profit and loss is determined by a price index that Binance constructs from data on a traditional-market instrument, and which uses funding payments exchanged between longs and shorts to help tether the contract to that index. The funding interval and cap are contract-specific.

Read it again slowly, because every clause is doing work.

USDT-margined and USDT-settled means you post USDT, your profit arrives in USDT, and your loss leaves in USDT. Not USDC, not the share, not dollars. Binance repeats this identically in every launch announcement we read.

No expiry date means the position exists until you close it or it is liquidated. There is no roll and no delivery month; the position simply continues until something closes it.

A price index that Binance constructs is the clause people skip. You are not trading the last print on Nasdaq. You are trading a number Binance computes — a weighted average of vendor prices during US regular hours, an exponentially weighted moving average in lower-liquidity sessions, and, when external quotes are unavailable during maintenance, holidays or weekends, a smoothed impact mid price derived from Binance's own order book. That machinery has its own guide, and it is where most of the surprises live.

Funding is a periodic cash transfer between the two sides that helps pull the contract's price toward the index. Its interval and cap must be read from the live contract specification: Binance changed nine TradFi contracts to four-hour funding with a ±1% cap on 10 September 2026. There is no arbitrageur delivering shares to close the gap, because no shares can be delivered.

That Binance treats this as its own instrument class, and not as a crypto perpetual with an unusual ticker, shows up in a place most readers never look: the futures API documentation carries a dedicated endpoint for signing a TradFi-Perps agreement, listed in both the USDⓈ-M and the Portfolio Margin references. Crypto perpetuals have no equivalent step. Whether signing it is a hard precondition for placing an order we did not verify and will not guess — only that a separate consent gate was built for this product and for nothing else on the venue.

The four things you do not get

Binance's own description of the product is that it lets users “trade on the price movements of these assets without needing to own them directly”. Read that clause closely and it gives away four separate omissions (no ownership, no dividend, no vote and no delivery) each missing for a different reason.

1. Ownership. A share is a slice of a company. A perpetual contract is an agreement about a number. If Apple were acquired tomorrow at a premium, a shareholder would receive the consideration; a perpetual holder would be at the mercy of whatever market notice Binance publishes about how the contract is handled. Binance states that it "will announce separate market notices on how to manage the events" for corporate actions such as splits and mergers — which tells you the handling is discretionary and case-by-case, not contractual and predictable. We cover what is known in dividends, splits and corporate actions.

2. Dividends. You do not receive them. This does not mean dividends are irrelevant: Binance says it maintains a separate dividend methodology for equity perpetual futures, so an ex-dividend date can still move the index your position is priced against. What you lose is the cash. A dividend-paying stock held as a perpetual gives you the price effect of the payout without the payout.

3. Votes. Irrelevant to most short-term traders, but the cleanest proof that this is not a share: there is no register with your name on it.

4. Delivery. This one Binance does not need to say, because it follows from the design: the contract has no expiry, so there is no settlement event at which anything could be delivered. You cannot convert a position into shares, ever, at any price. The exit is always a closing trade in USDT.

Against the real thing: buying the share

Start with the honest baseline. If you have access to a broker, that route gives you an asset rather than a position: it settles into an account in your name, it pays dividends, it has no funding cost, it cannot be liquidated unless you borrowed to buy it, and it trades only when the exchange is open.

The perpetual gives all of that up in exchange for four things: leverage up to 25× on some contracts, a 5 USDT minimum notional, settlement in a stablecoin you may already hold, and a market that never closes. If your holding period is months, you are paying funding at every contract-specific settlement and carrying liquidation risk to replicate something a share does for free. If it is hours, the share is the clumsier instrument.

Against a CFD

This is the comparison most people reach for, and it is the closest of the four, but "close" is not "the same", and the differences are structural rather than cosmetic.

A contract for difference is, in its ordinary form, a bilateral contract between you and one broker. Your counterparty is that firm. The price you get is the price that firm quotes. There is usually no shared central order book that other clients of other firms trade into, so there is no single public price for "the CFD" — there is your provider's price. Financing is typically charged daily on the notional, quoted as a benchmark rate plus or minus a spread, and it is a cost the provider sets.

A TradFi perpetual is different in three ways we can state with confidence:

• There is a shared order book. Every Binance user trading NVDAUSDT is trading into the same book, at the same displayed price, with the same tick size of 0.01 and the same 5 USDT minimum. The price is public and the depth is visible.
• The carry is symmetric and market-determined, not a fee. Funding on these contracts is paid between traders, not to the venue, and the interest component is fixed at 0%, so the entire funding rate is a measure of positioning imbalance rather than a financing charge. When longs pay 0.04%, that money goes to shorts. See funding on equity perps.
• The reference price is documented. Binance publishes how the index and the mark price are computed, session by session, including which smoothing mode applies when. That documentation is imperfect and, as we show elsewhere, internally inconsistent on at least one number, but it exists and you can read it.

What we will not tell you is how the two compare in regulatory terms. CFDs are treated very differently across jurisdictions, and the regulatory characterisation of Binance's TradFi perpetuals is not something we can source. Every Binance announcement carries the line "Products and services referred to here may not be available in your region", and the FAQ separately notes EEA restrictions relating to unauthorised stablecoins under MiCA. That is the extent of what we know. Your own eligibility is a question for your account, not for this page.

Against a tokenised equity

Several venues list products commonly described as tokenised stocks (the xStocks-style instruments are the best-known example) and people routinely lump them together with TradFi perps because both are "stocks on a crypto exchange". Structurally they are almost opposites.

A tokenised equity is, in form, a spot-like token: a transferable unit that you can typically hold without leverage, that exists as a balance rather than as an open position, and that has an issuer standing behind whatever claim it represents. The critical questions about it are all issuer questions — who issued it, what backs it, what happens on redemption, what happens if the issuer fails.

A TradFi perpetual has no issuer in that sense and no token. It is an open derivative position inside a Binance futures account, with margin, a liquidation price, and a funding obligation. There is nothing to transfer, nothing to hold on-chain, nothing to redeem.

Two consequences follow. A token can be held indefinitely at no carrying cost and cannot be liquidated by an adverse move; a perpetual accrues funding on its published schedule and can be closed out by the exchange. And a token's risk sits in the issuer and the custody arrangement, while a perpetual's sits in the pricing machinery and the margin engine. They fail in different ways.

We are deliberately not making claims about how any particular tokenised product is backed or regulated. Those differ by issuer and by jurisdiction.

Against a traditional dated future

The perpetual borrows its name and most of its mechanics from crypto, but its ancestor is the exchange-traded future. The single structural difference is the one in the name.

A dated future has an expiry, and that expiry does the tethering work: as delivery approaches, the futures price and the cash price must converge, because the contract settles against the cash market. Basis is a decaying quantity with a known deadline. Traders wanting continuous exposure roll from one contract month to the next and pay the roll cost explicitly.

A perpetual has no such deadline, so convergence has to be manufactured. Funding is one mechanism: when the contract trades above its index, longs pay shorts. On crypto perpetuals this works reasonably well because arbitrageurs can hold spot against the perp. On a TradFi perpetual that arbitrage is much harder — the underlying market may be shut, the trader may not have access to it, and no delivery is possible in any case. Because the funding schedule and cap vary by contract, traders must check the live terms. The tether can be weak, which is why persistent premiums and discounts can occur.

The five instruments, side by side

The right-hand four columns are general descriptions of instrument categories, not statements about any specific product. Only the perpetual column is sourced to Binance's own documentation.

Three of the five can be closed out for you

The comparisons above are about what each instrument is. The question the table cannot hold is how each one ends, and that is the part you do not control.

A share ends when you sell it; a tokenised equity ends when you sell or redeem it. Both are balances. A company can collapse underneath a share, but no third party reaches into the account and closes the position — not unless you borrowed to buy it. Those two are the outliers here, and the line between them and the other three is the most underrated one in the comparison.

The other three can be ended without your consent, and each uses a different price to do it.

• A CFD closes at your provider's price. The close-out is executed by the firm on the other side of your trade, under the margin terms of the contract you signed with it, at the level that firm was quoting. There is no external number to check it against, because there is no external market in that contract.
• A dated future settles whether or not you act. Expiry cannot surprise you (it is on the calendar) but it is unconditional: on the day the exchange publishes a final settlement price and closes the position against it under published rules. Before that, the clearing house can close you on margin like anyone else.
• A perpetual closes at a mark price you can compute but did not set. Liquidation runs on the mark (a median of three inputs, capped in how far it may deviate from the index) against a maintenance requirement that steps up in tiers as the position grows. The ladder is in leverage tiers and liquidation.

On transparency the perpetual wins: its close-out number is a published statistic, computed identically for every account, from inputs you can read before you open. The real difference is elsewhere. The perpetual's close-out mechanism runs continuously; the cash market behind it does not. A liquidation on NVDAUSDT at 04:00 UTC on a Sunday reads a mark whose standard median includes an Orderbook EWMA index derived from Binance's own contract book, with Nasdaq shut. Nothing else here can be closed out while its own market is dark.

The endgame nobody has published a rule for

Every instrument on this page has a defined terminal state except one.

A share ends in an event with a legal shape: an acquirer pays cash or stock and the register closes, a bankruptcy leaves a residual claim with a place in a queue, a delisting moves the line to an over-the-counter market and the security keeps existing. A dated future ends at expiry against a settlement price defined before anybody traded it. A tokenised equity ends at redemption, on whatever terms its issuer wrote.

A perpetual has no expiry (that is the feature) so it has no settlement event either: no moment written into the specification at which the contract is valued and closed. If the instrument the index is built from permanently stops producing prices, the index has lost its usual external input and nothing in the contract says what happens next. Binance documents temporary market closures: equity-perp indices move to Orderbook EWMA mode, using a smoothed impact mid price from the contract book until external pricing resumes. That is a rule for a state expected to end. A takeover that closes, a company taken private, or a halt that runs for weeks pending an investigation is a different state.

What Binance commits to is a procedure, not an outcome. The FAQ names four event types by code — forward splits (SPLT), reverse splits (RSPLT), mergers (MERGER) and spinoffs (SPINOFF), and says that for corporate events which may materially affect the underlying's price it “will announce separate market notices on how to manage the events”. Mergers being named at all is worth noticing. But naming an event class and publishing a rule for it are different things, and what is promised is a notice written at the time, not a term you can read today.

The nearest existing mechanism is not the same thing. Binance has a standing procedure for removing a USDⓈ-margined perpetual and has run it repeatedly on crypto contracts: an announcement, a cut-off after which no new positions may be opened, then a stated time at which Binance Futures will “close all positions and conduct an automatic settlement”, then delisting. What that procedure does not settle is the number that decides your money. On a crypto contract the settlement price comes from a market still trading at that moment. On an equity perpetual whose underlying has gone dark, the last observable price could be a stale halted quote, a deal price the market never traded around, or nothing at all, and which of those becomes the settlement number is exactly what nobody has published.

Two things follow, and neither needs the answer. Do not plan to hold through a deal close: an announced cash acquisition pins the underlying near the offer price, so the volatility you were paying leverage to reach evaporates and you carry funding on a position that cannot move, waiting for an event whose contract treatment is undefined. The trade that expresses that view is a share trade. And treat the announcement feed as part of the position — a notice written at the time is the only channel Binance has committed to, and it arrives on their schedule, not yours.

Who this instrument is plausibly for

There is a real use case here, and it is narrower than the marketing implies.

Someone whose capital is already in USDT and who wants short-horizon equity exposure. If your balance sits on Binance, the alternative is moving money into a brokerage, which takes days and may not be available where you live. A perpetual converts that friction into a single order.

Someone trading an event that lands outside market hours. Earnings print after the US close. A macro release hits at 08:30 New York time. A Korean chipmaker guides overnight. The perp is open; the cash market is not. This is the clearest genuine advantage of the product, and it is also where the pricing machinery is at its least reliable — read trading a US stock at 3am on a Sunday before you rely on it.

Someone sizing positions below the price of one share. A 5 USDT minimum and a 0.01 tick let you size from your stop distance rather than from what a whole share costs — see the position size planner.

Someone hedging crypto-denominated exposure with an index. Shorting SPYUSDT or QQQUSDT against a long book without leaving the stablecoin is a coherent thing to want.

Who it is not for

Anyone building a long-term equity position. Recurring funding, no dividends, discretionary handling of corporate actions and a permanent liquidation risk make this a poor container for a multi-year holding. If your thesis is "this company will be worth more in five years", buy the company.

Anyone who wants the dividend. You do not get it. There is no version of this product where you get it.

Anyone who does not yet understand mark price and liquidation. These contracts liquidate on the mark price, not the last traded price, and the mark price is built from a smoothed index whose behaviour changes several times a day. If that sentence is unfamiliar, work through leverage tiers and liquidation and the liquidation price calculator first.

Anyone reaching for the maximum leverage on a leveraged ETF contract. Seven contracts on the roster track funds that are themselves 2× or 3× daily-reset vehicles. Stacking 25× on a 2× fund is roughly 50× notional exposure to the underlying, in a wrapper that decays in choppy markets no matter which way you are positioned. The ETF guide spells this out.

Where this sits: 72 contracts, 17 batches

This is not a pilot. Between 5 January 2026 and 25 August 2026, Binance listed 72 TradFi perpetual contracts across 17 separate launch batches — 47 single-name equities, 12 ETFs, 7 leveraged ETFs, 4 non-US equities and 2 commodities. The first two were gold and silver. The first single stock was TSLAUSDT on 28 January, capped at 5×, still the lowest leverage on the roster. By July the list included STRCUSDT, a perpetual on a preferred stock, at 25×.

Stock and ETF contracts share traits such as USDT margin, no expiry and 24/7 trading, but their specifications are not identical. Funding intervals and caps can vary by contract, as the September 2026 move of nine contracts to four-hour funding and a ±1% cap shows. Leverage tiers also vary. Read all live symbol specifications together; no single field is a complete risk statement.

The complete roster, with launch dates, venues and leverage caps read off the official announcements, is on the contract list.