Replicating the strategies of the most brilliant investment minds in the history of Wall Street is, without a doubt, a great challenge for any manager or mere investor in order to enhance the investment portfolio. Until recently, it was something unthinkable and the only thing that was aspired to was to follow, copy or accompany the movements of the portfolios of the great Wall Street gurus, such as Warren Buffett, Stanley Druckenmiller or David Teppereven if it is to achieve a similar benchmark. But with the advent of artificial intelligence (AI) “everything” seems possible, or almost everything.
So last Wednesday, just as the Fed announced the rate cut, fintech startup Intelligent Alpha launched an AI ETF fund, powered by chatbots, that promises to harness the brainpower of the world’s most illustrious investment minds and their investment prowess. For some, it could be the market’s boldest bet to explore new AI tools to emulate the decision-making of Wall Street’s most legendary and iconic investors. In recognition, The new AI ETF pays homage to Jesse Livermore, one of the most legendary stock traders of the early 20th century, by naming its ticker “LIVR”.
The new Intelligent Livermore ETF fund is based on ideas and investment decisions generated by three language models (LLM) such as ChatGPT, Gemini and Claude, dubbed the Investment Committeewhich will be inspired by the thoughts and actions of famous Wall Street titans.
The VIP list of selected investors
Intelligent Alpha noted that the list of personalities the fund will emulate not only includes Buffett, Druckenmiller and Tepper, but also investors such as Dan Loeb or the old one known by Argentina, the “vulture” Paul Singeramong others, although the designers of the instrument explain that the fund’s holdings may not necessarily reflect the actual bets of those investors.
As explained Doug ClintonCEO of the Minneapolis-based firm, the idea was to recreate the basic principles of hedge funds where there are specialized funds today. While the portfolio will not necessarily reflect their investments in real life, it will be inspired by the investment strategies and philosophies associated with such investors. So The firm will send specific instructions to the chatbots, ordering them to create portfolios based on the personalities of these financial icons.so the trio of chatbots will generate between 60 and 90 global companies spanning a range of sectors, topics and geographies, including healthcare, renewable energy and Latin America, among others.
However, to avoid confusion where an LLM makes up information or misinterprets the information received, a human being from the fund finally intervenes for final supervision.
As the CEO explains, this is just to make sure that there are no hallucinations in the portfolio, such as a company that has committed fraud or some other serious issue, and also to ensure that the portfolio complies with any regulatory or compliance restrictions that the AIs have not taken into account when creating the portfolio.
ETFs and AI: More of the Same?
While the concept of an AI-managed ETF is at least, shall we say, bold, it is not entirely new as several hedge funds and ETFs have begun experimenting with AI to streamline investment processes. It just so happens that while experts have praised AI for its ability to process massive data sets quickly, eliminating the need for analysts to perform monotonous tasks (some hedge funds already use chatbots for research and investment processes, thus cutting down on the time spent on these activities), however, They acknowledge that their ability to consistently outperform traditional strategies remains unproven.. That is, it is still largely experimental and unproven. Therefore, they warn that There is still little evidence that AI is disrupting and displacing mass investment units, and there is still a lot to be solved when it comes to problems like chatbots making things up in their responses.
Furthermore, there is no evidence yet that AI-chosen investments have an advantage over passive investing. According to monitoring Of the 16 AI-focused ETFs in the US, only one is outperforming the S&P 500 in 2024the Franklin Intelligent Machines ETF, which returned 19% while the stock index gained 18% at the close of the last financial year. In addition, only one of them, the Global X Artificial Intelligence & Technology ETF, has seen significant inflows, with more than $1 billion this year. Next is the Roundhill Generative AI & Technology ETF which raised less than $120 million while the rest have seen minimal inflows or outright outflows so far this year.
From the experiment to the bottom
What sets Intelligent Alpha’s ETF apart is its use of LLMs rather than traditional machine learning models, Clinton explains, noting that most AI ETFs are based on older techniques, limiting their strategies to cluttered quantitative data. So By leveraging LLMs, Intelligent Alpha aims to overcome these limitations and create portfolios with a different kind of edge.
The idea for the Intelligent Livermore ETF came about in 2023 when Clinton began experimenting with ChatGPT to build investment portfolios and after seeing success in creating strategies that outperformed the S&P 500, he expanded his testing to 40 different investment strategies. These tests eventually led to the founding of Intelligent Alpha, which is linked to Deepwater Asset Management that manages approximately $400 million in venture capital and equity funds, where Clinton is associated with well-known technology analyst Gene Munster.
While the Intelligent Livermore ETF is Intelligent Alpha’s debut product, the firm has plans to launch a range of AI-powered offerings, including custom portfolios and hedge funds aimed at both retail and institutional investors.
Source: Ambito
