Unlocking Alpha: how AI is helping investors fine-tune their strategies
Artificial intelligence (AI) is driving some of the brightest and boldest innovations in finance and is quickly becoming a proven mainstay in the pursuit of personalisation, security, compliance and financial accuracy. But perhaps a lesser known arena where AI is making a splash is in stock investing.
Stock investing has long been a game of predictions, of recognising the risks and volatility that stand between the initial investment and some form of return. Assessing the viability of achieving this outcome has historically been left to the expertise of wealth managers, who through industry experience and insight attempt to determine which risks could be outweighed by the potential returns.
This assessment describes the difference between what’s known as Alpha and Beta investing. Alpha is a measurement that defines a stock’s performance in relation to a specific market index. In essence, the more Alpha a stock is able to obtain, the better it performs, and stocks that incur maximum Alpha are highly regarded within the portfolios of investors.
On the flipside, Beta defines how volatile a stock is/could be compared to the general volatility of a market, and as an indication, enables investors to weigh the risk of their investment against the potential returns.
Pinpointing the ideal equilibrium between these two measurements isn’t a task for the faint of heart, namely because calculations demand the critical analysis of hundreds of thousands of constantly changing data points within often short amounts of time.
In the shadow of such a momentous undertaking, AI has thrown wealth managers seeking to improve both the accuracy and speed of their decision making an invaluable lifeline with its increasingly proven ability to assess and make sense of huge amounts of data.
UK-based machine learning company 3AI, founded by Hassan Salamony, Jacob Ayres-Thomson and Simon Judd, is a start-up seeking to help investment professionals leverage the power of AI to improve the analysis of big data in their pursuit of achieving Alpha.
The first of its two main product lines, its Smart Alpha Insights function, seeks to serve investors with stock picking and live predictions, which the company claims have enabled its 1,000 highest-rated stocks to surpass the lowest 1,000 by 28% over predicted 12-month periods.
This function is backed by 3AI’s Deep Factor AI, which is used to assess 326 different data factors per stock and spans global market capitalisation exceeding $100 trillion in equities across all levels of institutional investing.
3AI’s Smart Alpha Insights product is joined by its flagship US Smart Alpha Index, which is also powered by Deep Factor AI to produce investment indices for equity covering over 20,000 stocks.
The company says that its indices function not only offers “a deeper research investment strategy”, but also seeks to mitigate “look-ahead bias” – a concept where analysis, often associated with quantitative index construction, relies upon information that was not yet available at that specific time.
It claims that its Smart Alpha Index solution can be “tailored to any investing mandate and thematic”, and that the function has outperformed the S&P 500 by 1.6 times since going live in 2020.
Salamony tells FinTech Futures that the company’s application of Deep Factor AI intends to “empower Tier 1 hedge funds, global banks and national security organisations with actionable insights about companies and the equity market”.
He describes this process as a “bottom-up company analysis approach” that’s able to craft investment strategies with lower volatility while also mitigating the risk of hindsight bias.
By combining machine learning, quantitative investments and deep factor analysis, Salamony says the company “aims to provide investors with reliable Alpha solutions”.
“When we’re assessing stock, we’re looking at something like 30 million different data points per stock. It’s a vast amount of data. And all of that data is out there, it exists and is available for anyone to go and subscribe to.”
Yet despite this wide accessibility, Salamony still recognises the challenges in applying large volumes of data, and describes the ability to process such volumes and apply it within decision making as a “key problem” for investors working on their own.
‘A fascinating glimpse into the future of AI in finance’
While 3AI has been eager to demonstrate the benefits of AI in cultivating investment strategies, it also understands the multiple risks associated with leveraging the technology in such a context.
Investment managers must at all times understand what’s pushing their investments towards a positive outcome, but this requirement meets multiple barriers when faced with AI’s lack of transparency. Managers may understand the potential for Alpha, but may simultaneously struggle to interpret why a solution like 3AI’s Deep Factor AI has reached the decision it has.
Ayres-Thomson explains how the company’s application of Explainable AI (XAI), an innovative form of AI that’s able to divulge its thought processes to the user, works to overcome this barrier.
“Layering explainability onto our Deep Factor AI has made the system’s thinking transparent, which is crucial for trust and validation,” he says.
Ayres-Thomson suggests that this level of scrutiny involves a massive computational effort, but that the effort is rewarded with “invaluable insights” into the AI’s reasoning while also revealing new AI discovered knowledge.
Although he describes the company’s use of XAI in stock analysis as “a significant advancement” for the sector, Ayres-Thomson predicts that the need for explainability will diminish going forwards as “societal comfort with AI grows, and the focus moves towards validation of performance”.
“I think culturally in a decade or two from now, we’ll move beyond the requirement for XAI, because when you get to XAI, you realise that there’s so much going on in the explanation, and what you’ll really want is validatable AI. You want to validate that the AI is doing an equal or better job than what you were doing before.”
3AI has placed itself in the heartland of stock investing’s biggest barriers, and has carved out an avenue that not only enables institutional investors to think smarter and act quicker, but also to apply innovative technologies like AI within their decision making while benefitting from full insight into how exactly a decision has been delivered.
While the technology arguably has more evolutions to come, both Salamony and Ayres-Thomson agree that right now, “it’s a fascinating glimpse into the future of AI in finance”.