Perpetual Futures, Explained

What Are Perpetual Futures?

Most financial innovation exists, at least on paper, to serve the real economy. Crypto skips that step entirely. What it does produce, once in a while, is something intellectually interesting - a genuinely clever mechanism that has zero connection to productive activity. Perps (perpetual futures, if you're being formal) fit that description perfectly. They are fascinating. They're worth understanding. And they have absolutely nothing to do with real economic output.

Stablecoins get all the press for payment use cases. But here's what most people miss: a huge chunk of stablecoins are actually sitting there as collateral for perp positions. Of the roughly $300 billion in stablecoins outstanding right now, about a quarter lives on exchanges. And most of that quarter? Backing perp trades. Not payments. Not remittances. Gambling collateral, plain and simple.

Perps dominate crypto trading by volume - typically 6-8 times larger than spot. That ratio shouldn't shock anyone who's looked at traditional markets. Derivative volume swamps spot volume basically everywhere. The degree varies - market to market, Schelling points, user culture. In India, most retail equity investing happens through derivatives. The U.S. is the opposite for individual stocks, where most retail exposure comes through spot markets (directly holding shares, or indirectly via ETFs and mutual funds). But stock index trading volume? Overwhelmingly derivatives too.

The Core Problem

I want to be precise about what large crypto exchanges actually are. Casinos. That is not metaphor - it is a functional description. Crypto markets supply the numbers the same way roulette wheels and dice do at a physical casino. Patrons enter with money and, statistically speaking, leave with less. Now, physical casinos need enormous capital investments - real estate, infrastructure, ongoing operating costs, and returns on all that speculative capital. They'd gladly choose less capital intensity if they could. Market forces and regulation partially stop them.

The Trust Requirement

A crypto exchange is capital intensive too, but for completely different reasons. It is not the websites or APIs - those are relatively cheap by financial software standards. Not physical plants either. It's trust. And trust is expensive. Bettors need to believe they'll actually get paid when they win. So do the sophisticated market makers providing liquidity. That means exchanges have to keep assets on hand - mostly crypto, plus some cash if they're well-regarded - exceeding customer account balances.

Those assets just sit there. Doing nothing. There is an implicit cost of capital associated with them, whether nominal (borne by the gamblers) or material (borne by sophisticated market making firms, exchanges, or exchange-affiliated traders). Dead money.

So why do perps exist? Because they give gamblers the risk they want while cutting the total capital that exchanges and market makers need to keep the whole enterprise profitable. That is it. The entire value proposition, right there.

Historical Origins

In commodities futures markets, you can contract to buy or sell standardized valuable things at defined future dates. Straightforward enough in principle. What's less obvious to outsiders is that the overwhelming majority of these contracts never result in anyone actually taking delivery. Nobody wants the pork bellies. They get cancelled by offsetting contracts well before the specified dates.

Since speculation and hedging are the core use cases anyway, the financial industry came up with a clever refinement: cash-settled futures. No physical delivery of pork bellies or oil barrels required. Reference prices get established (with considerable intellectual effort toward making them robust and fair), and on delivery day people net short just pay people net long. Clean. No warehouses.

Shiller's Proposal

In the early 1990s, economist Robert Shiller asked a question that sounds obvious in hindsight: if you don't actually want pork bellies or oil barrels for consumption in April, and we accept that almost nobody trading futures does, then why close contracts in April at all? Why fragment liquidity across April, May, June contracts when you could just... keep the market going? Forever?

The idea sat around for years before finding its natural habitat. Crypto. Bitmex is generally credited as the popularizer. What I'm describing below is the standard crypto implementation, though variations exist across exchanges.

How Perps Actually Work

Continuous Settlements

Rather than all futures settling on the same day, perps settle multiple times daily for particular markets on particular exchanges. The mechanism is called the funding rate. At a high level: winners get paid by losers every few hours (often every 4 hours), and then the game keeps going. Unless you've been blown out because you overleveraged. Or for other reasons nobody bothered to mention on the onboarding screen.

Here's a toy example. A retail user buys 0.1 Bitcoin via a perp. Screen shows $86,000 each, so they put up $8,600 cash. Price rises to $90,000 before next settlement? Roughly $400 in winnings gets credited to their account, and their position continues reflecting exposure to 0.1 Bitcoin units through the perp. At this point they could sell (or at any point, really), paying one commission and spread to buy, one of each to sell, and maybe walk away from the casino with winnings. Or they play another round.

Where did that $400 come from? Someone on the other side who was symmetrically short Bitcoin exposure via a perp. And with some important caveats I'll get to shortly, this is a closed system. No good or service gets produced here except speculation itself. Your winnings are someone else's losses. Zero sum, full stop.

Funding Rate Mechanics

One wrinkle worth noting. Some exchanges cap how much the funding rate can move in a single settlement period. The intent resembles circuit breakers in traditional markets - automatically blunting feedback loops before they spiral. But the resemblance is superficial at best. Funding rate caps can't actually break circuits. They can delay when losses get realized, sure. But they cannot prevent them, because they don't prevent the symmetrical gain realization on the other side of the trade.

Perp funding rates also embed interest rate components. You might see it quoted as 3 basis points daily, or 1 basis point every eight hours. Sounds tiny. But leverage changes everything. At 10X leverage, that 3 basis points becomes 30 basis points daily. There is a reason consumer finance legislation standardizes borrowing costs as APR rather than basis points per day - it's so unscrupulous lenders can't bury what amounts to 200% APR in the fine print. Crypto perp exchanges, unsurprisingly, do not follow that convention.

Price Convergence Mechanisms

Perp prices don't, as a matter of physics, exactly match the underlying asset's spot price. Some users consider this a feature. Not a bug.

The Basis Trade

When markets are exuberant - and crypto markets are exuberant more often than not - perps tend to trade above spot. That gap creates an opportunity sophisticated participants exploit through what's called the basis trade: short the perp, buy spot in equal size, offsetting positions. Because funding rates are set against reference prices for the underlyings, longs end up paying shorts more (as percentages of the perp's current market price). For the degenerate longs, this is just the cost of doing business. Gambling prices went up - whatever. But for other participants, it's a market incentive to close long positions. Closing means selling. Selling pushes price down. Toward spot. Which is the whole point of the mechanism.

Market makers running this trade can afford to wait. If prices converge, they close at a profit while having collected funding payments the entire time. If perps keep trading rich? They just keep collecting the elevated funding rate. To the extent that rate exceeds their own cost of capital, this is spectacularly lucrative. The kind of lucrative where you start asking yourself why you'd bother doing anything else with your money.

Flip everything to understand the other direction. Same mechanics, reversed signs.

The basis trade, when executed properly, is delta neutral. You are not exposed to the underlying at all. You don't need beliefs about Bitcoin's adoption story, or its fundamentals, or market sentiment, or halvings. None of it matters. You're getting paid to provide the gambling environment - and a critically important piece of it, because perp prices must stay reasonably close to spot to keep attracting participants. You are also renting out access to your capital so others can lever up. Remarkably clean business model. Right up until the moment it isn't.

Exchange Risk

Because here is the part the basis trade pitch decks leave out. You're underwriting the exchange itself. If the exchange blows up, having your collateral become a bankruptcy estate claim is the happy scenario. Galois Capital learned this the hard way - a crypto hedge fund running basis trades that had roughly 40% of assets sitting on FTX when it imploded. They wound down the entire fund. Sold the bankruptcy claim for 16 cents on the dollar. Sixteen cents.

Remember the trust problem from earlier? Markets can't function without some system that says "this counterparty is good for it if you win." For market makers, trust comes from collateral kept on the exchange. Your money. Sitting on someone else's servers. In someone else's jurisdiction.

Most market makers operate across many crypto exchanges simultaneously, and this is exactly why capital efficiency obsesses them. Fully collateralizing every potential position across every venue they trade on would require absurd amounts of capital. But if they don't pre-deploy it, they miss profitable opportunities that evaporate in milliseconds. It is a permanent tension. There's no clean solution and there probably never will be.

Leverage and Liquidations

Gamblers like risk. That is the whole point - risk amplifies the fun. And since crypto offers plenty of casinos to choose from, any exchange providing only "regular" Bitcoin exposure via spot or unlevered perps is offering a less exciting product than the competition down the street. So how much leverage should you offer? More. The answer is always more. Right up until the entirely predictable consequences start happening.

Regulation T Comparison

In standard U.S. brokerage accounts, Regulation T has set maximum leverage limits for almost 100 years now (by setting margin minimums). You get 2X when you open a position, and 4X "maintenance" before forced closure. If volatility pushes you past those limits, your brokerage is obligated to forcibly close your position. Not optional. Not a suggestion. Required.

Quick example. You have $50k cash. Reg T lets you buy $100k of stock - $50k equity, $50k loan. That is 2x leverage. Now the stock drops to about $67k. You still owe the $50k loan, leaving you with only $17k of equity. You're teetering at 4X leverage, and a margin call is imminent - if your broker hasn't already blown you out of the trade without asking.

What's the relevance to crypto? Just hold onto that number for now. 4X.

Extreme Crypto Leverage

Perps are offered at 1X for people who want unlevered exposure. But they're routinely offered at 20X. At 50X. At 100X. SBF, during his press tour about being a responsible financial magnate, voluntarily self-limited FTX to 20X and acted like this was some major concession to prudence. Think about that for a second. Twenty times leverage as the responsible option.

One reason perps are structurally better for exchanges and market makers: they make it simpler to blow out leveraged traders. The exact mechanics depend on exchange, position size, various parameters. But generally you either force customers into closing trades or assign their positions to someone willing to take on risk in return for a discount. Efficient. If nothing else.

Liquidation Fees

Blowing out losing traders is lucrative for exchanges. Except when it catastrophically isn't. It is a priced service in many places - Binance describes it as "a nominal fee of 0.5%," which sounds harmless until you realize it's calculated on the amount at risk, not the amount remaining. So that "nominal" fee can eat a huge portion of whatever is left. If the account's negative balance is less than the liquidation fee? Wonderful. Thanks for playing. The exchange - or "the insurance fund," same thing really - pockets the rest as a tip.

When accounts go negative by more than the fees cover, "insurance funds" can choose to pay winners on behalf of liquidated users. This is at management's discretion - an important detail that few participants think about until it matters. Management usually does pay up. A casino that develops a reputation for stiffing winners doesn't stay a casino for very long.

But tail risk is real, and this is where things get genuinely scary. The capital efficiency comes at a price: there physically is not enough money in the system to pay all winners given sufficiently dramatic price moves. Not nearly enough. When forced liquidations start cascading, sophisticated participants yank their liquidity or the exchange gets overwhelmed - technically, operationally, or both. The liquidations chew through whatever diminished book liquidity remains, and each successive move keeps growing in magnitude. A feedback loop with no natural brake.

And then crypto gets reacquainted with automatic deleveraging. ADL. One of those perp contract details that almost nobody reads about until it is happening to them, personally, at 3 AM.

Automatic Deleveraging

Risk in perps has to be symmetric. If there are 100,000 Somecoin exposure units long (accounting for leverage), then there are 100,000 units short. Simple accounting identity. But - and this is the part that matters - symmetry of exposure says nothing about whether either side is sufficiently capitalized to actually pay up in all instances. That's a very different question.

The ADL Mechanism

When management decides that paying winners from the insurance fund would be too costly - or straight-up impossible - they automatically deleverage some of the winning positions. In theory, there is a published, transparent process for selecting who gets ADL'd. Because it would obviously destroy confidence if the exchange ADL'd unaffiliated accounts while paying out its friends, its affiliated trading desks, or its particularly important counterparties. In theory.

In theory, the exchange ADLs highly levered accounts before less levered ones, and accounts with high profits before those with lower profits. In theory. I keep saying "in theory" because, well.

So let's say you thought you were 4X leveraged going into a 20% move. You should have earned 80%, right? Except it turns out you were only 2X leveraged, so you earned 40%. Why were you retroactively only 2X? That's what automatic deleveraging means in practice. And why couldn't you collect the other 40% you feel entitled to? Because the collective group of losers on the other side simply doesn't have enough money to pay you, and the insurance fund was either insufficient or deemed insufficient by whoever's running the show.

Impact on Sophisticated Traders

ADL is particularly painful for sophisticated market participants doing e.g. basis trades, because they thought they were 100 units short via perps and 100 units long somewhere else via spot. If it turns out they were actually 50 units short via perps, but 100 units long, their net exposure is +50 units, and they've very possibly just gotten absolutely shellacked.

In theory, this can happen to the upside or downside. In practice in crypto, this seems to usually happen after sharp price decreases, not sharp increases. For example, October 2025 saw widespread ADLing as more than $19 billion of liquidations happened across various assets. Alameda's CEO Caroline Ellison testified they lost over $100 million during Terra's stablecoin collapse in 2024, but since FTX's insurance fund was made up, when leveraged traders lost money, their positions were frequently taken up by Alameda. That was quite lucrative much of the time, but catastrophically expensive during e.g. the Terra blowup. Alameda was a good loser and paid the winners, though: with other customers' assets they "borrowed."

Traditional Market Comparison

In traditional markets, if one's brokerage deems one's assets are unlikely to cover margin loans from the brokerage, one's brokerage issues a margin call. Historically that gave relatively short periods (typically a few days) to post additional collateral, either by moving in cash, transferring assets from another brokerage, or experiencing asset value appreciation. Brokerages have options, and sometimes requirements, to manage risk after or during margin calls by forcing trades on behalf of customers to close positions.

Negative Balance Treatment

It sometimes surprises crypto natives that, in cases where one's brokerage account goes negative and all assets are sold with negative remaining balance, traditional markets largely still expect you to pay that balance. This contrasts with crypto, where market expectation for many years was that customers were pseudonymous entities, and dunning them was time-wasting. Crypto exchanges have mostly either stepped up their KYC game or pretended to do so, but market expectation remains that defaulting users will basically never successfully recover.

Loss Allocation

Who bears losses when customers don't, can't, or won't? The waterfall depends on market, product type, and geography, but as a sketch: brokerages bear losses first, out of their own capital. They're generally required to keep reserves for this purpose.

Brokerages, in ordinary business courses, have obligations to other parties that would be endangered if catastrophically mismanaged and could not successfully manage risk during downturns. In this case, most counterparties are partially insulated by structures designed to insure peer groups. These include clearing pools, guaranty funds capitalized by clearinghouse member firms, clearinghouse own capital, and perhaps mutualized insurance pools. That is the rough ordering of the waterfall, varying depending on geography/product/market.

One can imagine true catastrophes burning through each protection layer, and in that case, clearinghouses might be forced to assess members or allocate losses across survivors. That would be a very bad day, but contracts exist to be followed on very bad days.

One commonality with crypto though: this system is also not fully capitalized against all possible events at all times. Unlike crypto, which for contingent reasons pays lip service to being averse to credit even as it embraces leveraged trading, the traditional industry relies extensively on underwriting various participants' risk.

Will Traditional Finance Adopt Perps?

Many crypto advocates believe they have something traditional finance desperately needs. Perps are crypto's most popular and lucrative product, but they probably won't be materially adopted in traditional markets.

Existing derivative products already work reasonably well at solving cost of capital issues. Liquidations are not traditional brokerages' business model. And learning, on a day when markets are 20% down, that you might be hedged or you might be bankrupt, is not a prospect that fills traditional finance professionals with warm fuzzies.

Conclusion

Perpetual futures represent a fascinating financial innovation optimized for one thing: enabling maximum speculation with minimum capital requirements. They solve real problems for crypto exchanges and market makers by reducing capital intensity while maintaining the gambling environment that attracts users.

However, perps create systemic risks that become apparent during extreme market moves. Automatic deleveraging, opaque insurance funds, and extreme leverage create conditions where winning traders may not get fully paid and hedged traders may discover they're catastrophically exposed.

For traditional finance, perps offer lessons but unlikely adoption. The existing derivative ecosystem already addresses capital efficiency needs without introducing the tail risks that make perps problematic for participants seeking actual risk management rather than maximum speculation.

Understanding perps provides insight into crypto market structure and why crypto trading volume looks so different from traditional finance. It's not just that crypto participants are more speculative (though they are) - it's that the infrastructure itself is optimized for speculation rather than productive capital allocation.