Instrument Updated 2026-09-12 By Ivo Renner

Leveraged ETF perpetuals, and the double leverage nobody mentions

In our launch archive through 29 August 2026, nineteen of seventy-two Binance TradFi perpetuals track exchange-traded funds. Twelve of those funds are unleveraged; seven use daily-reset leverage. A perpetual on a leveraged fund adds another layer of exposure. This page works through the arithmetic; the counts and launch limits are historical, not a complete current roster.

The archive contains nineteen fund contracts

Read the contract archive by category and the shape of it is not what most people assume. Forty-seven contracts track a single company. Two track a metal. Four track a non-US listing. The remaining nineteen (more than a quarter of the board) track a fund: twelve ordinary exchange-traded funds and seven leveraged ones.

Binance does not separate those two groups in the trading interface. SPYUSDT and INTWUSDT appear in the same list, under the same [TradFi] tab, and share traits such as USDT margin, no expiry and 24/7 trading. Funding intervals and caps must be checked contract by contract: Binance changed nine TradFi contracts to four-hour funding with a ±1% cap on 10 September 2026. The instruments are not remotely comparable, and nothing on the order ticket tells you so. The categories below are ours, not Binance's.

Index funds, sector funds, and the leveraged ones

1. Plain index funds

Two contracts: SPYUSDT on the State Street SPDR S&P 500 ETF Trust and QQQUSDT on the Invesco QQQ Trust, both listed on 6 April 2026 at 10×. These are the most forgiving instruments on the roster. A single company's earnings miss moves the S&P 500 by a fraction of what it does to that company's own shares, so the overnight gap risk that defines every single-name contract is smaller here. There is no daily reset, no path dependence, no decay beyond the fund's ordinary running costs.

The catch is behavioural, not structural. Index moves are small, so people compensate with size.

2. Sector, country and thematic funds

Ten contracts, and the group where most of the interesting exposure lives: XLEUSDT (energy sector), XBIUSDT (biotech), EWZUSDT (Brazil), EWTUSDT (Taiwan), URNMUSDT (uranium miners), GDXUSDT (gold miners), KSTRUSDT (the SSE STAR Market 50 index), LYTEUSDT (photonics and optics), DRAMUSDT (memory) and BITOUSDT, which gets its own section below.

These are baskets, so they still diversify away individual company risk. But diversification within a sector is not diversification. Every holding in a uranium-miners fund reacts to the same handful of headlines. URNMUSDT at 20× is not safer than a large-cap stock at 20×; it is a concentrated macro bet with the appearance of a fund.

EWTUSDT and KSTRUSDT deserve a flag: they give Asian exposure through a US-listed wrapper, which means US trading hours and US session handling. That is a different pricing problem from the direct Korean and Chinese contracts; see the guide to Korean, Chinese and Hong Kong equity perps.

3. Daily-reset leveraged funds

Seven contracts, and the reason this page exists.

What "daily reset" actually means

A 2× long fund does not borrow once and hold. It rebalances its exposure at the end of every trading session so that it starts each new day with exactly twice the fund's current net asset value in underlying exposure. That is the reset, and it is the whole mechanism.

The consequence is mechanical. When the fund wins, the reset makes it buy more exposure at a higher price; when it loses, the reset makes it sell exposure at a lower price. In a market that goes up, then down, then up again (which is what markets normally do) the fund is buying high and selling low every evening, and the accumulated cost of that is volatility drag.

Note what is not the explanation. This is not a fee, not tracking error, not a sign the fund is badly run. A perfectly managed 2× daily fund with zero costs still bleeds in chop. It is arithmetic.

The arithmetic, with actual numbers

Start with the simplest case. An underlying at 100 rises 10% on Monday, then falls 9.0909% on Tuesday. Tuesday's fall exactly undoes Monday's rise: 110 × 0.909091 = 100. The underlying is flat over two days.

Now the 2× daily fund, also starting at 100. Monday it returns twice 10%, so +20%: it closes at 120. Tuesday it returns twice −9.0909%, so −18.1818%: 120 × 0.818182 = 98.18.

You were flat on the underlying and lost 1.82% on the fund. Nobody charged you anything. Now extend the sequence, because two days is not how anyone holds.

Repeat that pair for two full trading weeks — ten sessions, five cycles. Each cycle multiplies the fund by 1.20 × 0.818182 = 0.981818, and five of them compound to 0.9123. The underlying is at 100. The 2× fund is at 91.23. Run it for a trading month and the fund is at 83.24, down 16.8% on an underlying that has not moved at all.

A 3× fund on the same path is worse than proportionately worse. Each cycle multiplies it by 1.30 × 0.727273 = 0.945455, which over ten sessions is 0.7554 — down 24.5% on a flat underlying. TMFUSDT tracks a 3× long 20-year Treasury fund.

Ten per cent daily swings are extreme, so here is a realistic version. Take an underlying that alternates +3% and −2.9126% — again netting to exactly flat every two days. Each cycle multiplies a 2× fund by 1.06 × 0.941748 = 0.998252, a loss of 0.175% per cycle. That sounds trivial. Compounded over a trading month it is −1.7%. Compounded over a trading year it is roughly −20%, on an underlying that finished exactly where it started.

Rule of thumb: for a fund with daily multiple L, the drag over a period is approximately (L² − L) / 2 multiplied by the underlying's variance over that period. For L = 2 that is one times the variance; for L = 3 it is three times. A stock running 3% daily moves has an annualised variance near 0.23, which is where the 20% figure above comes from. It is an approximation, not a formula to price with, but it tells you the order of magnitude before you open the position.

These seven funds also carry their own management fees and internal financing costs, which sit on top of the drag. We have not verified the expense figures for any of them — read the issuer's factsheet if you intend to hold for more than a session.

Up ten, down ten, and the two points that vanish

Every sequence above was rigged so the underlying finished exactly flat. Do it the way you would actually test it instead: up 10% on Monday, down 10% on Tuesday. The underlying goes 100, 110, 99 — a two-day return of −1.00%.

Ask what a 2× daily fund did over those two days and almost everyone answers −2.00%. It goes 100, 120, 96: a two-day return of −4.00%. Two whole percentage points are missing, and no fee, spread or tracking error accounts for them.

The shortfall triples when the multiple goes from two to three: 50% more leverage, three times the cost, exactly what the (L² − L) / 2 rule above predicts.

Now wrap a perpetual around the 2× row. At a 10× setting that −4.00% costs 40% of your margin; the same 10× on the underlying would have cost 10%. Ten points of the difference is exposure you chose. Twenty points is the reset, on a path whose destination you called correctly.

In a trend, the same mechanism pays you

Drag gets written up as a tax. That is half the mechanism, and the half that only loses money cannot explain why anyone buys these funds.

Take five sessions of a clean +5% a day. The underlying compounds to 1.05⁵ = 1.2763; the 2× fund to 1.10⁵ = 1.6105. That is +27.63% against +61.05%, where twice the underlying's return would have been 55.26%. Now run the trend downwards at −5% a day and the surprise is the sign: 0.95⁵ = 0.7738 against 0.90⁵ = 0.5905, or −22.62% against −40.95%, where twice the loss would have been −45.24%.

Every sign in that table comes out of the same evening trade: a winning day forces the fund to add exposure into a market that has just moved its way, a losing day forces it to cut. That is trend-following, hard-coded, run once at the close, with no way to switch it off. Over a trend it is the right trade five days running, whichever way the trend points. Over a chop it is the wrong one five days running.

So the accurate sentence is not that leveraged ETFs decay. Holding one makes you short realised volatility and long persistence of direction, whether you had a view on either or not.

Stacking a perp on top of a leveraged fund

Now put a Binance perpetual around it.

INTWUSDT tracks the GraniteShares 2× Long INTC Daily ETF, and it launched with a maximum leverage of 25×. Open that contract at the cap and your notional exposure to Intel is roughly 2 × 25 = 50× your margin. Work through what that means:

• Intel moves 1% against you. The fund moves about 2%. Your position loses about 50% of its margin.
• Intel moves 2% against you. Your margin is gone. Intel moves 2% in a session routinely.
• Intel moves 2% in your favour and closes there. You have doubled, and you are now holding an instrument that will decay if the stock chops sideways from here.

Two things make this worse than the headline suggests. Liquidation is driven by the mark price, not the last trade, and outside US regular hours that mark comes from a smoothed index rather than live constituent prices. And the leveraged fund itself gaps harder at the open than its underlying does, because it opens with twice or three times the overnight move already applied. Run the numbers in the liquidation price calculator before sizing anything here.

The liquidation distance you think you set

Move the leverage selector to 10× and your instinct answers without being asked: ten per cent against me and I am out. On AAPLUSDT that is close enough. On TMFUSDT it is wrong by a factor of three, in the direction that ends the position early.

The correction is one division. The move in the ultimate underlying that erases your margin is one divided by your leverage setting times the fund's daily multiple.

The middle rows are the ones to sit with, because 10× is what a careful reader settles on after deciding the cap is reckless. On TMFUSDT that careful 10× is a 30× position in long-dated Treasuries, ended by a 3.33% move in a bond index.

Two corrections shorten every row. Liquidation fires at the maintenance margin, not at zero equity — the tiers, and where to read your own, are a page of their own. And the estimator linked above prices the fund, not the underlying. So convert before you calculate: take the underlying move you are willing to sit through, multiply by the fund's multiple, and hand the calculator that.

The seven leveraged-ETF contracts

Every one of these tracks a fund that resets daily. The final column applies the division above at each contract's launch-day maximum leverage.

Three observations about that table.

The inverse funds are not symmetrical. TBTUSDT and SKDDUSDT are 2× short daily funds. They suffer the same drag, plus a structural asymmetry: the thing they are short can rise without limit while their own value is bounded below at zero. Held through a sustained rally, such a fund decays toward nothing rather than going negative — which sounds protective until you notice it means the fund can be nearly worthless while your perp on it is still open.

Two of them stack on a foreign underlying. SKUUUSDT and SKDDUSDT are US-listed funds on SK Hynix, whose home market is shut for most of the Western day. The fund prices during US hours; SK Hynix prices during Korean hours; the perp prices constantly. Three clocks in one instrument.

The pairs invite a trade that does not work. Holding SKUUUSDT against SKDDUSDT, or TMFUSDT against TBTUSDT, is not a hedge. Both legs decay and both pay or receive funding on their respective schedules. Long both sides of a mirror pair is a bet on low realised volatility, structured in the most expensive way available.

BITOUSDT: the strangest contract on the board

BITOUSDT was listed on 27 July 2026 at 25×, alongside TMFUSDT and TBTUSDT. It is a perpetual contract on the ProShares Bitcoin ETF — a fund that takes its bitcoin exposure through futures rather than by holding coins. It is not a leveraged fund and the roster tags it as an ordinary ETF.

It is still an odd thing to own here, for a reason that has nothing to do with leverage: Binance already offers bitcoin perpetuals directly. If you want leveraged, USDT-margined, 24/7 bitcoin exposure, that product has existed on the venue for years and is far more liquid than anything in the TradFi tab.

So what does BITOUSDT add? A layer of intermediation. The ETF's price reflects the futures curve it holds, the fund's own mechanics and equity-market demand for the fund's shares, and it only exists during US trading hours. The perpetual on it trades continuously. When vendor quotes are unavailable during maintenance, weekends or holidays, Binance now derives the equity-perp index from its own order book's impact mid price and smooths it with an EWMA; it does not hold the ETF's last quote fixed. Bitcoin does not stop moving at 16:00 New York time.

The spread between BITOUSDT and a straight bitcoin perp is therefore driven mostly by session timing and by the ETF's own basis, not by bitcoin's price. That makes it a basis instrument rather than a bitcoin-exposure instrument. Traded as a way to be long bitcoin, it is the noisiest possible route to a position you could take directly. Traded deliberately against a bitcoin perp to express a view on the fund's premium or discount, it is coherent, but that is a relative-value trade with two funding legs, and it should be sized like one.

When a leveraged ETF perp is a defensible trade

None of the above is an argument that these contracts should not exist or that nobody should touch them. Volatility drag is a cost, and like any cost it can be worth paying. The conditions under which it is worth paying are narrow and specific.

1. Short holding period. Over a single session the drag is close to zero, because the fund delivers its stated multiple over exactly one day. Over a week it is noticeable. Over a month in a choppy tape it is the dominant term in your P&L. If your thesis needs a month, express it with an unleveraged contract.
2. Strong directional conviction, not a hedge. If your view is "this moves sharply and soon", the reset works with you, on the arithmetic above. If your view is "range-bound" or "I want protection", this is the wrong instrument in both directions.
3. Small size, sized off the stacked number. Size from the ~50× or ~75× figure above, never from the perp's leverage setting. The position size calculator takes a risk-per-trade figure; feed it the effective exposure.
4. Not held across a session boundary you cannot price. The fund resets at the US close; the perpetual does not stop. Every hour outside US regular hours is an hour in which your liquidation level is set by an estimated index. Read what happens outside market hours before assuming a stop will protect you.

What to check before opening a leveraged-ETF position

First, confirm what the fund actually is. The fund's name carries the multiple and the direction — "2x Long", "UltraShort", "Bull 3X". If the name contains a number, treat it as a leveraged fund.

Second, check the current leverage tier in the trading interface. The caps in our table are launch-day maximums read off the announcements. Binance revises margin tiers after listing, and the parameters panel inside the interface is what governs your position.

Third, look at the funding history and live terms. The settlement interval and cap are contract-specific and can change, so read both before sizing a position. The funding cost calculator turns the displayed rate and interval into a daily number.

If you have not traded a TradFi perpetual before, start with what these contracts actually are and work through the pre-trade checklist. These seven are the last thing on this venue you should learn on, not the first.