AI Trader ‘Sets and Forgets’01.24.2017
Wall Street’s use of artificial intelligence passed a milestone earlier this month as the San Francisco-based hedge fund The Tech Trader Fund completed 12 months of trading without human interaction, according to fund founder William Mok.
“In the last couple of months, we have reached a point where we can self-sustain,” said Mok. “We finally made it to an escape velocity where if we do not put in more effort, we are not going to sink back down.”
Although the fund leaves its decisions up to the AI, it also has automated as much of the investment lifecycle as possible and outsources much of its middle and back office to third parties.
“You can point at anything and say that it’s not 100% machine,” he said “Maybe it’s 99% machine. But in practice now, we’ve been living it for a year and have not had to do anything to it that is related to trading.”
The fund, which has $20 million of assets under management, trades US equities and exchange traded funds that have a market capitalization of at least $1 billion.
“We get into a position some point in the day and hold it for several months,” said Mok. “It doesn’t matter if we get into the position in the next minute or the next hour.”
The AI uses a subset of trading algorithms provided by the fund’s executing broker Convergex. “We’ve tested a variety of order types, and they didn’t make a difference,” he noted. “They are more for quantitative traders, which we are not.”
The Tech Trader Fund currently is in the middle of closing an investment from a fund of funds, which Mok declined to name.
“We just finished due diligence last week,” he said. “It’s pretty exciting because it is like the domino effect: Once the first investor comes in, the next one comes in a little bit bigger, and then you kind of have your funnel so to speak.”
In the meantime, the fund’s holding company, Autodidactic I, looks to use its automated revenue stream to fund potential moon-shot projects. “It similar to how Google’s search engine business funds the company’s X Labs,” Mok explained.
Technology has enhanced capabilities of surveilling larger and more disparate data sets.
With Eugene Kanevsky, James Redbourn, and Joanna Wong, CLSA
AI/ML on the buy-side trading desk is a long-term program with short- and mid-term deliverables.
Aiden uses the computational power of deep reinforcement learning to improve trading results and insights.
BondDroid’s AI-generated prices are integrated directly into LTX’s pre-trade analytical tools.