‘Autonomous’ Artificial Intelligence Advances in Buy-Side Trading
Algorithms work great…until they don’t.
This old adage in the trade was primarily based on the days when algos were essentially a “set and forget” proposition, created in a set of circumstances that, when materially changed, became less effective. Likewise, cruise control in cars works great for maintaining a steady speed, until it encounters icy conditions or obstacles. Then came autonomous driving, capable of making real-time, real-world driving decisions that adapt to material changes in driving conditions. For institutional investment managers, AI in algorithmic trading is that speed governor towards a moment of total control in algorithms. And some sales companies are leading the way.
Trading algorithms route buy and sell orders to the exchange or dark pool with the most liquidity and the least friction – no small feat in a complex, high-speed US stock market which spans dozens of lit and unlit trading venues. AI can optimize these algorithms to better recognize and adapt to real-time market conditions, as well as improve trade-to-trade performance.
“We offer buyers much more advanced strategies that can recognize patterns, spread forward or backward, and learn on their own,” said Sam Clapp, executive director of Mizuho in New York. “It’s a lot more sophisticated and a lot smarter than before.”
Large institutional firms on the buy side are notorious for their risk aversion and caution, so it’s no surprise that their adoption of artificial intelligence has been deliberate. In Q3 2021, only 31% of investment managers were using algorithms incorporating artificial intelligence and machine learning, according to a Greenwich Coalition survey of 36 buy-side firms. And the skepticism remains: In a separate Greenwich Coalition survey of 234 buy-side respondents, 41% said the promise of artificial intelligence is overstated, compared to 27% who said it was understated.
Money is flowing into AI because companies don’t want to be left behind if and when the technology realizes its potential. A 2021 Broadridge survey showed that buy-side and sell-side companies plan to increase their AI, blockchain and cloud spending by 33% within two years.
BlackRock, the world’s largest investment manager with $10 trillion under management, is exploring how AI can improve trading. “We are constantly researching how data science and artificial intelligence can help us augment human intelligence through computing, and do it at scale,” said Supurna VedBrat, global head. trading at BlackRock, at Markets Media Europe last year. “I think going forward this will change the trading strategies used by the buy side significantly, but today from a risk perspective we are seeing better results.”
While AI in buy-side trading is in its infancy, there is a wide green field for sell-side banks to establish themselves as providers of choice. Mizuho, which won Best in AI at Markets Media Group’s 2022 Markets Choice Awards, does so with its Compass trading algorithm, which goes beyond machine learning and into deep learning. This involves computers learning to think using structures modeled on the human brain.
By way of background, Clapp explained that ten years ago, trading algorithms were static and calendar-based. “There were no gaps based on peripheral events, or even just more standard events or nuances in the market, like volume is up or gaps are wide,” he said. “The strategies of 10 years ago just didn’t take that into account.”
Sell-side brokers entered the AI space several years ago, typically starting with a price prediction feature that predicted the price of a security at a specific time in minutes or hours in the future. .
Compass comes into play when a trader receives an order from the portfolio manager. “The trader knows how many shares to buy or sell, and he has a specific benchmark to beat a certain price,” said Don Hundley, head of e-commerce at Mizuho in Tokyo. “What Compass aims to do is trade those stocks in a way that will achieve the best possible outcome.”
Hundley noted that Compass goes beyond price prediction to a concept called clustering, which groups data points based on certain similarities and allows the AI to know from past experience when trading d a group of actions in a certain style is more effective.
“It could be very aggressive; this could include not posting in enlightened markets and focusing only on dark flow; it could take more cash now than later, or do the opposite,” Hundley said. “So where static algorithms trade publicly traded stocks the same way, clustering allows us to fine-tune the style of execution.”
A buy-side institutional trader said, “Compass has allowed our desk to really leverage some of the deep learning available by not only using what Mizuho has done around AI, but also giving us the ability to customize our own version of Compass with contributions.”
Mizuho, who has a partnership in place with a renowned technology university in China to collaborate on algo AI trading technology and has published research on the subject, further enhances Compass with volume prediction. This feature takes into account various factors such as news feed, earnings releases and recent volumes, to predict the volume that will be traded in a given future period.
“The purpose of AI for a buy-side trader is to augment their own trading, whether that means automating some of the vanilla, non-difficult trades, or using an AI-powered broker algorithm to get a better execution result for a tough market trade,” Hundley said. “Our goal is to achieve better trade results against any given benchmark using the AI tools to our disposal.”