The inside story of how a hike in Hong Kong changed crypto trading forever

It was sometime in 2015, on a hiking trail in Hong Kong, that the perpetual swap — also called a perpetual future or a “perp” for short — was born. Ben Delo, the mathematician and co-founder of BitMEX, was walking with a friend called Bavik, a derivatives trader, wrestling with a problem that had been nagging at him for months.

BitMEX had been trying everything. Quarterly futures, monthly futures, weekly futures, 48-hour futures, even a contract that lasted just 24 hours before resetting. Nothing was working. Customers kept complaining that their positions were closing without warning. They wanted something that looked like spot, traded like spot, but gave them the leverage that only a derivatives exchange could offer.

“What if a future never expired?” Delo asked.

Bavik’s answer was immediate. “Mathematically, it would be worth infinity,” Delo recalls him saying.

Technically, he was right. A futures contract’s value is partly derived from the time remaining until expiry and the cost of carrying the position. Remove the expiry date and that carrying cost compounds indefinitely, making the theoretical value infinite. But then Bavik offered a solution: just charge traders the bitcoin overnight rate, the way you might charge LIBOR (London Interbank Offered Rate) in traditional finance.

There was one problem. “I said, what’s that?” Delo recalls. “He said, ‘Hm, just charge them the overnight bitcoin interest rate’. I said, ‘I don’t think that exists.’”

So Delo built it. And in doing so, he invented one of the most consequential financial products of the 21st century.

Building BitMEX

To understand why the perpetual swap mattered, you have to understand what BitMEX was trying to be before it became the most liquid bitcoin market in the world.

When Delo and Arthur Hayes founded the exchange in 2014, they were not thinking about retail traders chasing 100x leverage. They were thinking about institutional hedgers. Hayes had worked at Deutsche Bank, while Delo spent years building high-frequency trading systems at JP Morgan. Their thesis was that bitcoin miners and payment companies needed a way to hedge their exposure, and BitMEX would provide the professional infrastructure to do it.

“We built it basically to look like a Bloomberg terminal,” Delo said. “We used Reuters instruction codes. Z14 meant expiring December 2014.”

The institutions never came. What came instead were traders, sophisticated but retail, people who had financial experience but were playing with their own money. And what they wanted was not guaranteed settlement or low leverage. They wanted to speculate, and they wanted to do it with as much size and leverage as possible.

BitMEX listened. By Halloween 2015, the exchange was offering 100x leverage, made possible by a real-time margining system that Delo had built from scratch. “I built the order matching engine, the position keeping system, the margining system, the PnL system, the settlement system,” he says. “Everything on that was me.”

The issue with futures, even short-dated ones, is basis, the premium at which a futures contract trades above the spot price of the underlying asset. A futures contract trades at a premium to the underlying asset, and that premium reflects an implied interest rate. In traditional finance, this is well understood. In crypto, in 2015, it confused almost everyone.

“Our customers would be like, why is bitcoin so expensive on your exchange?” Delo recalled “And we would say, ‘Well, if it is expensive, why don’t you short it?’ And that would blow some of their minds. You could short something rather than just long it.”

BitMEX kept shortening the expiry of its listed futures contracts. Weekly futures. Then 48-hour. Then a contract that relisted every single day.

“Every 24 hours, it would expire or settle. And people would say, ‘Why did you liquidate me?’ And we’d say, ‘We didn’t liquidate you. Your position closed at the index price. You got exactly the spot price,’” Delo said. “They’re like, ‘We don’t understand.’”

The customers knew what they wanted, even if they could not articulate it. They wanted a leveraged product that never went away. Delo’s hiking trail conversation gave him the framework to build one.

Inventing the funding rate

The perpetual swap launched in May 2016 with little ceremony. The core mechanic was straightforward: a futures contract with no expiry date, anchored to the spot price through a daily funding rate. Longs paid shorts, or vice versa, depending on whether the swap was trading above or below spot. BitMEX took no cut. The rate was purely a balancing mechanism.

The early funding rate was derived from third-party lending markets, primarily Bitfinex, where traders could borrow dollars or lend out Bitcoin. Take the dollar borrow rate, subtract the Bitcoin borrow rate, and you had something approximating the cost of holding a long position.

It worked, until it did not. As Bitcoin began its rise through 2016 and into 2017, demand for long exposure on BitMEX overwhelmed the funding mechanism. The swap started trading at a persistent premium to spot, causing the contract price to drift away from the actual price of bitcoin and undermining the mechanism designed to keep them aligned. The interest rate being imported from Bitfinex was simply not high enough to reflect what was happening on BitMEX itself.

“We had to dynamically adjust how we calculated that funding rate,” Delo said.

The original funding rate had been imported from external lending markets, a fixed reference point that could not respond to conditions on BitMEX itself. The new approach replaced that with a dynamic one, looking inward at how the swap was actually trading rather than outward at what Bitfinex was charging to borrow dollars.

The solution was elegant. Rather than looking outward to other markets, BitMEX would look inward. The exchange began measuring how far the swap was trading above or below spot over an eight-hour window, treating that gap as an implied basis, and back-calculating the annualized rate from it. That rate would then be charged at the end of the next eight-hour window.

“This was very important because you gave market makers notice of how you were calculating it, what it would be, and then when you would charge it,” Delo said. “Because it was paid from longs to shorts, if the swap was trading at a 1% premium, you would charge longs 1% but give 1% to shorts. And then immediately the market makers, knowing that, would come in, short the swap, and anchor it back down to the spot price. It was a dynamic equilibrium.”

This is, in essence, the funding rate mechanism that every major derivatives exchange in the world now uses.

The product that took over

By 2017, BitMEX was the most liquid bitcoin market on the planet. The exchange was processing $3-4 billion a day, and the perpetual swap was at the center of it all. Price discovery for bitcoin was happening not on Coinbase or Bitstamp but on the BitMEX order book.

The concentration of liquidity was itself a product of the swap’s design. Before it launched, BitMEX had been running quarterly, monthly, weekly, 48-hour and 24-hour contracts simultaneously, spreading market maker capital thin across six different tenors. The swap collapsed all of that into one instrument.

“By offering one product, they [traders] were able to consolidate their liquidity, which meant a more liquid market, tighter spreads,” Delo said.

Competitors noticed. Another competitor exchange copied so literally that it reproduced portions of the BitMEX FAQ without understanding how the product worked, Delo told CoinDesk. Others took the concept more seriously. Eventually, every major exchange in crypto offered its own perpetual swap, each one built on the funding rate architecture that Delo had begun assembling on that Hong Kong hiking trail.

“The fact that every other exchange has copied the swap just proves what a financial innovation it is,” he said. “I think it now does $40, 50 trillion dollars a year of turnover. It’s one of the most successful products in the history of capitalism.”

What comes next

BitMEX chose not to patent the perpetual swap. Delo says they considered it and decided the startup’s time was better spent building.

“We were a scrappy startup,” he says. “We thought, just get it out there. If it was any good, the market would show us.”

And show them the market did. Now, a decade on, the product is starting to attract the attention of traditional finance regulators. The CFTC is reportedly making room for perpetual swaps under its framework, and there is speculation that the CME could eventually list them on equities.

For Delo, that prospect is the final validation of something that started as a question on a hillside above Hong Kong, asked by someone who had grown tired of watching his customers complain about positions that kept disappearing.

“I think once traditional finance sees the benefits of this financial product,” he said, “it’ll be impressive.”