AI-powered stock pickers are already operating in in the market and one of the most prominent is operated by EquBot, a San Francisco-based issuer of two Exchange-Traded Funds (ETFs). Equbot CEO Chida Khatua is all praises for his AI model which combines deep financial analysis wirh the pattern-finding strengths of machine learning.
Marie Beerens interview Khatua and she tells us how AI-powered ETFs operate in this article in Investors Business Daily:
Using IBM’s (IBM) AI Watson combined with a proprietary in-house AI model, EquBot is able to replicate the work of 3,000 research analysts, Khatua said. Its two funds, AI Powered Equity (AIEQ) and AI Powered International Equity (AIIQ), have gathered $150 million in combined assets since their launch in the past year and a half.
The platform is autonomous and able to rebalance every day if necessary, Khatua told IBD in a recent interview. It is also purely data driven and as such doesn’t introduce any human bias. It’s an adaptive model, which means it is not fixed in rules but based on market conditions and able to learn and improve every day.
[…]
The sheer amount of data now available to investors is staggering. Everything from social media posts to satellite photos of retail parking lots to CEO mannerisms and expressions during earnings calls and beyond can help an investor uncover opportunities and make key portfolio decisions. But the volume of data is impossible for a person, or even a team of people, to properly and accurately synthesize and act upon. That’s where AI comes in, and that is the thinking that underpins each of our funds.
The EquBot model, with the help of powerful computing platforms like IBM Watson, compiles and computes over a million data points daily, from quarterly reports, news articles, social media posts, financial statements and more. The model then builds predictive financial models for thousands of domestic and global companies to identify timing and positioning for optimal stock selection and potential capital appreciation. We’ve set the fund’s objectives to meet commensurate volatility measures and have so far delivered benchmark-beating returns in both ETFs.
Aside from the amount of data it can process daily and how quickly it can do so, the EquBot model also continues to learn as it goes, allowing it to improve its process over time. That allows the model to uncover unique insights and opportunities that a human alone may not have found or uncover events that may act as catalysts for unrecognized value. Additionally, this approach helps minimize human error and the inherent biases that can skew the results in more traditional investment approaches.
