Key Takeaways
- Perpetual futures ('perps') are derivatives with no expiration date and leverage that can exceed 100x.
- A 'funding rate' paid between longs and shorts, usually every 8 hours, keeps perp prices near spot.
- Robinhood plans U.S. crypto perps in the coming months, with up to 10x leverage on BTC and ETH.
- A September 21 squeeze liquidated $648 million in shorts, yet total futures open interest rose.
Perpetual futures — “perps” in market shorthand — are the fastest-growing trading instrument in global finance right now. They started in crypto, stayed in crypto for years, and are now crossing into mainstream U.S. brokerage accounts.
The CFTC formally approved Bitcoin perps from regulated U.S. platforms in May 2026, and Coinbase has them. On September 29, 2026, Robinhood announced that eligible U.S. customers will be able to trade crypto perps in its app in the coming months, with up to 10x leverage on Bitcoin and Ether.
And they scare me a little. Not because they’re inherently evil — they serve real purposes. But the combination of no expiration, extreme leverage, and 24/7 trading creates a system where things can unwind very fast and very badly, in either direction.
A record short squeeze in August made that concrete, and two more liquidation waves in September showed the same mechanism running both ways. Let me explain what they are, why traders love them, and why regulators and risk managers are watching nervously.
What Is a Perpetual Future?
A regular futures contract is an agreement to buy or sell an asset at a set price on a set future date. When that date comes, the contract expires — either the trade settles or you roll into a new contract.
A perpetual future has no expiration date. You can hold your position indefinitely: no settlement date, no rollover, no forced close. You buy a BTC perp, take a leveraged long position, and hold it as long as you have enough margin to keep the position open.
This sounds simple, but it introduces a problem: how do you keep the perp price aligned with the actual Bitcoin spot price if there’s no expiration mechanism to force convergence?
The answer is the funding rate.
The Funding Rate: The Mechanism That Makes Perps Work
Every 8 hours (on most platforms), traders on one side of the market pay traders on the other side. If perp prices are trading above spot — more demand for longs — long holders pay shorts. If perps trade below spot, shorts pay longs.
The effect: the funding rate creates a financial incentive to trade against the crowd and pull the perp price back toward spot. When lots of people are long, holding a long gets expensive, which discourages new longs and encourages new shorts.
It’s a clever mechanism. It mostly works. But it breaks down when markets move violently — in either direction, as August and September 2026 both showed.
| Market condition | What happens |
|---|---|
| Perps above spot | Longs pay shorts |
| Perps below spot | Shorts pay longs |
| Funding near 0 | Market roughly balanced |
| Funding above 15% APR | Crowded longs, a historical warning sign |
That last row matters. A funding rate above 15% annualized means so many traders are long that they’re paying a steep premium to stay in. In every major crash since 2020, extremely high funding rates preceded the unwind by days or weeks.
As of late September 2026, the picture is warmer but not overheated. In a September 21 snapshot, funding was positive on 24 of the 25 largest Bitcoin perps, with a median near 0.006% per 8-hour period — roughly 6.5% annualized.
That’s up from about 2.35% in mid-September, but still well below the 15% danger zone. A moderate funding rate doesn’t mean risk is gone. It just means positioning isn’t as lopsided as it was before the January crash.
Why Traders Love Them
For traders, perps solve real problems that traditional futures don’t.
No rollover friction. With a traditional futures contract, you have to close and reopen a position as expiry approaches. That creates transaction costs and sometimes unfavorable spreads. Perps eliminate that entirely.
Leverage. Platforms commonly offer 10x, 20x, 50x — and some offshore exchanges offer 100x or more. A trader with $10,000 can control a $500,000 position at 50x. The gains (and losses) scale accordingly.
24/7 trading. Crypto never closes. Unlike stock futures with overnight gaps and market hours, perps trade continuously, which suits global crypto markets and the algo traders running around-the-clock strategies.
Access to short positions. It’s much easier to short an asset via perps than through borrowing mechanisms in spot markets.
How Liquidation Cascades Work
This is where it gets dangerous.
When you hold a leveraged perp position, your exchange requires a minimum margin level — a buffer of capital relative to your position size. If prices move against you enough to erode that buffer, you get liquidated: the exchange forcibly closes your position at a loss to protect itself.
Here’s the cascade, using a long-side example:
- Bitcoin drops 5%.
- Traders holding 20x long positions have only a 5% buffer — those positions get liquidated.
- The exchange sells their Bitcoin to close the positions, adding more selling pressure.
- Bitcoin drops another 3%.
- Now traders at 10x are getting margin calls. More forced selling.
- The cycle repeats until leverage is purged from the system.
This is exactly what happened on January 30, 2026 — over $2.56 billion in leveraged long positions liquidated in a single trading day.
The same mechanism runs in reverse when shorts get crowded. On August 19-20, 2026, Bitcoin broke out of a six-week range and short sellers were forced to buy back. CoinGlass data reported by CoinDesk put short liquidations at about $2.7 billion in 24 hours, a record for the short side.
More than $1 billion of Bitcoin shorts were closed in roughly an hour. Bitcoin jumped almost 8% to around $69,100 and later topped $71,000 as the forced buying fed the rally.
On September 2, the cascade ran the other way: roughly $368 million in liquidations hit within 24 hours, about 82% of it longs. The trigger was macro — rising oil prices pushed up the odds of a September Fed rate hike, and Bitcoin fell to roughly $76,500.
Then on September 21, it flipped again. A short squeeze wiped out about $648 million in bearish bets and pushed Bitcoin to around $85,000, its highest level in eight months.
The part that caught my eye: total crypto futures open interest rose about 7.6% to roughly $156 billion that day, per CoinDesk. The squeeze didn’t flush leverage out of the system. New leveraged positions replaced the ones that got wiped out.
These aren’t rare events. With open interest in Bitcoin perps alone around $81 billion, even moderate price moves can trigger nine or ten figures of forced liquidations. And macro headlines, like a Fed meeting or an oil shock, are now just as capable of triggering one as crypto-specific news.
Get perps and crypto-risk updates by email
When Robinhood’s perps go live or the next big liquidation hits, I’ll post what it means for regular investors. Get it free in your inbox.
Free. You’ll get my new posts, including these updates. Unsubscribe anytime, and check your spam folder for the confirmation email.
The Systemic Risk Case
Here’s why I think this warrants serious attention, beyond just individual trader risk.
Perps were once isolated to crypto, and the losses stayed within crypto. But as perp trading moves onto regulated U.S. platforms — the CFTC’s May 2026 actions set a framework for listing crypto perpetual contracts, and Coinbase and Kalshi both offer them — the links to traditional finance are growing.
Consider the chain:
- Institutional traders on perp platforms use leverage provided by prime brokers
- Prime brokers fund those positions through repo markets and bank credit lines
- A large enough liquidation event can create margin calls that ripple back into bank balance sheets
- Banks managing collateral exposure to crypto positions may need to sell other assets to cover
We saw a version of this in 2022, when the collapse of Terra/Luna and FTX hit crypto-focused lenders and then touched Silvergate and Signature Bank. Those were small banks. As perp trading brings in larger institutions — and now mainstream retail platforms like Robinhood — the transmission paths get stronger.
The other piece that concerns me is concentration. A handful of large algorithmic trading firms account for a disproportionate share of perp volume. If they run similar strategies, their simultaneous unwinding in a stress scenario could be destabilizing in ways that are hard to model.
August’s record short squeeze and September’s back-to-back liquidation waves are small-scale examples of exactly that dynamic: crowded, one-sided positions unwinding all at once.
Real-World Examples
Example 1 — How leverage amplifies gains and losses: Sarah puts $5,000 into a Bitcoin perp position with 20x leverage, controlling $100,000 of BTC exposure. If Bitcoin rises 5%, she makes $5,000 — a 100% return on her capital. If it falls 5%, she’s liquidated and loses it all.
The same 5% move either doubles her money or erases it. This isn’t investing — it’s closer to options trading in risk profile, except there’s no natural theta decay limiting the loss.
Example 2 — Funding rate as a warning sign: Mark noticed BTC perp funding rates hitting 18% APR in January 2026 — a historically elevated level signaling extreme long crowding. He reduced his long exposure and added a small short.
Two weeks later, the January 30 crash liquidated $2.56 billion in longs. Mark’s short profited while leveraged longs around him were wiped out. Reading the funding rate didn’t require predicting the crash — it just told him the market was fragile.
Example 3 — Getting caught on the wrong side of a squeeze: Priya built a short position in mid-August 2026 after watching Bitcoin stall below its prior highs. When Bitcoin instead broke out on August 19-20, her leveraged short was liquidated as part of the roughly $2.7 billion short wipeout.
Her mistake wasn’t the thesis — plenty of traders were short at the time. It was using enough leverage that a single day’s move against her could wipe out the whole position.
Where This Goes From Here
Perp trading is moving into mainstream finance whether regulators are ready or not. The CFTC’s May 2026 actions were a first step toward proper margin rules, reporting, and customer protection. The agency has since issued a no-action letter on how futures brokers can post customer crypto as margin with foreign brokers.
The biggest change since my last update is Robinhood. Per its September 29 announcement, eligible U.S. customers will be able to trade perps on eight coins in the coming months:
- Leverage: up to 10x on BTC and ETH, and 3x on SOL, XRP, DOGE, ADA, LINK, and HYPE.
- Fees: one basis point (0.01%) per trade through the end of 2026.
- Risk tools: stop-loss and take-profit orders, a real-time liquidation price, and at-risk alerts.
A 10x cap is tame next to 100x offshore platforms, and the risk tools are the right idea. But it puts perps one tap away from millions of retail investors. At 10x, a 10% move against you is a total loss, and Bitcoin moved about 8% in a single day in August.
I’ll also be watching whether perps end up inside Robinhood’s new agentic accounts, where AI agents can trade around the clock. Robinhood hasn’t said. I cover the agent side in my Robinhood agentic trading experiment.
Offshore platforms still handle a majority of global perp volume, with leverage ratios that remain extreme. And coordinating oversight of an instrument that trades 24/7 across dozens of jurisdictions is genuinely hard.
I’ll keep watching open interest, funding rates, and regulatory developments. This is one of those risks that isn’t obvious until it becomes obvious — and September showed it doesn’t take a crypto-specific shock to set one off. I’ll update this page as things develop.
Where Perp Traders Get Burned
Misunderstanding that leverage amplifies losses just as much as gains. Platforms make leverage look like a free upgrade — more exposure for the same capital. It’s not. A 10x position in a normally volatile asset like Bitcoin can hit liquidation on a routine 10% swing, and many retail traders discover this the hard way.
Holding through funding payments without counting the cost. If you hold a long in a crowded market, you may be paying 0.05% every 8 hours in funding — roughly 54% APR. On a position held for months, funding alone can eat much of your gain even if the price moves your way.
Assuming betting against the crowd is automatically “safe.” August’s record short squeeze, and the long flush on September 2, show that a contrarian position isn’t immune to a liquidation cascade. It just moves the risk to the other side of the trade.
Assuming liquidation is the worst case. On some platforms during extreme volatility, auto-deleveraging (ADL) can partially close your winning position to cover someone else’s losing one. It’s disclosed in the fine print but surprises many traders the first time.
Conflating perp trading with investing. Holding a Bitcoin perp long-term is not the same as holding Bitcoin — you pay funding, you can be liquidated, and you have no claim on the underlying asset. For long-term exposure, spot holdings or regulated ETFs carry a different risk profile.
Letting a macro headline catch you over-leveraged. Two of September’s liquidation waves were driven by Fed and oil news, not crypto news. If you follow the broader market, my post on 2026 stock market volatility covers the rate and oil backdrop.
If you’re weighing whether to hold crypto directly at all, see my breakdown of the factors worth considering before investing for the case for and against spot exposure.
