Yes: Grasshopper, a Singapore-based proprietary trading firm, was named after the moniker of the lead character in the 1970s T.V. show “Kung Fu”. Master Po would caution his young charge, Caine, by using his nickname. “Patience, young grasshopper” became a tagline for martial-arts fans in the West.
For John Lin, founder and CEO, Grasshopper was that young, thrusting player in the big world of high-frequency trading. He founded it in 2006 to trade Nikkei futures as well as stocks and exchange-traded funds. As markets went electronic, he participated in the race for speed: co-location, cable lengths, whatever it took to drive latency to ten millionths of a second.
He brings a similar playfulness – and competitiveness – to the office conference room: upon a shelf sits a black box with “algos” inscribed on its lid. Trader humor, ha ha.
Grasshopper now trades over $500 billion of notional value every month for its own account. But for Lin and his crew, the quest for speed became a race to the bottom, and Grasshopper has since shifted to using machine learning to generate alpha. And now he is bringing that tech to trading crypto-currencies over the counter.
From latency to learning
Lin’s theme is that high-frequency trading went through its naughty “Flashboys” phase – and then latency became commoditized. Speed alone no longer provided an edge. Competition moved to exploiting market “microstructure”: market rules, regulations, and tech infrastructure. Then it became about the systematic trading of algorithms.
“High-frequency now has a lower place along the risk-return curve,” Lin said. “And we matured with it.”
Today, advantage is found in artificial intelligence. Lin says machine learning has taken over as the driver behind new algos and tactics.
“We talked about regression analysis and neural networks for a long time, but there was no traction,” he said, until about 10 years ago. “Then it became an edge. Today it’s clearly an edge.”
Machine learning changes the nature of trading, by emphasizing the ability to sift and interpret data. It creates the need for data scientists to access and clean data, and to train machines and seed them with raw information.
Machine learning no longer defaults to that elusive race to zero
“But the trump card is that machine learning no longer defaults to that elusive race to zero,” Lin said. Unlike latency, which at some point becomes a commodity, the permutations of machine learning are not going to run dry.
“The simplest of algos are just if/then statements,” he said, “but with machine learning we’re talking about multiple layers of logic based on different interpretations of the past.”
But what about explainability – what about the black box?
“There are so many blind spots with algos, especially when people pile into a trade,” he said. “But instead of spending money on a race for speed, let’s spend it on creating a real playing field – let’s repopulate the ecosystem that’s been devastated by high-frequency trading and its toxic flows and lack of fairness. That’s your blind spot.”
Now his firm is looking to apply similar techniques to trading crypto-currencies. He has set up an affiliated company, Tilde, as an over-the-counter broker serving clients.
“I like crypto because it’s the confluence of tech, economics and philosophy,” Lin said, sounding a little like Master Po.
He also likes it because it lets a prop shop be itself. Lin doesn’t care about headcounts or volumes or how many exchanges the company touches – because those metrics don’t matter in an industry awash with hype and fakery.
We see OTC as a bridge
He also likes it because it is challenging. Machine learning presents a steep innovation curve to high-frequency trading, but doesn’t fundamentally alter it. “Crypto,” on the other hand, he says, “is disruptive,” as it creates asset classes that don’t correlate, and because it does away with the need for a central clearing house or other post-trade infrastructure. It creates new risks in the form of counterparties; knowing who’s on the other side of a trade.
“We bring knowhow and experience to the token world,” Lin said, noting that Tilde trades can account for as much as 25% of a given coin’s market share.
For market makers, the challenge is liquidity, and Lin doesn’t claim to have found a magic bullet. He’s seen other market players resort to illegal measures to fluff volumes, but reckons some will land themselves in lawsuits. “We need a war chest, a trading strategy, and to use third-party market makers in order to remove conflicts of interest,” he said. “We just have to nurture the market among centralized exchanges” until the industry scales to the point that decentralized exchanges – the true peer-to-peer networks – become viable.
“OTC broking is contrary to the philosophy of crypto, so we see it as a bridge,” he said.
But he’s not too swept up in the industry’s anarcho-libertarian roots, either. He says he’s raised the idea with peers of the industry developing a code of conduct but has been rebuffed – despite the demand to institutionalize the space if real-money accounts are to embrace crypto.
“All those libertarians want widespread adoption, but they’ll pay a price for that,” he said. “Let the mainstream adoption happen; the real upside of crypto isn’t about competing against fiat currencies. It’s how you tie in the world.” Meaning adding the transfer of value to the internet, and joining that to properties such as the commerce of the World Wide Web, the power of social media, and the physicality of the internet of things.
“It’s not a foregone conclusion,” Lin says of this all-encompassing vision. “But it’s possible; it’s a good challenge.”
The new world of crypto makes Grasshopper (or Tilde) once more an apprentice. And, for that matter, Master Po sounds like an advocate for decentralization. From an episode from series 3 of the show, counseling Grasshopper:
“To those who would destroy us, in the past, when we have relied on our great teachers to lead us, our enemies in high and low places could deal us a mortal blow by simply lopping off our heads. Now in our oneness we are not like a great beast, which may be destroyed by a single well-planned stroke to the brain; rather we are like an ocean of many waves, or a field of flowers: though one or more may be uprooted, the others still live with a life of their own.”