Hedge funds must 'embrace AI or die', says Project One founder

Project One, a 'next generation' AI-powered hedge fund has set it sights on generating returns similar to Renaissance Terchnologies' fabled Medallion fund, the granddaddy of quant funds. 

Project One, the brainchild of Andrew Sobko and Rami Jachi, is powered by a 100 per cent model-driven, alpha-learning, AI algorithm designed to pinpoint market demand projections while actively applying real-time data analysis insights without human interruptions. 

The Project One hedge fund, is targeting USD1 billion under management by 2021 and projects an average of 60 per cent annualised returns, similar to the performance of the Medallion fund which is famed for achieving returns of more than 66 per cent annualised before fees and 39 per cent after fees over a 30-year span from 1988 to 2018. Interested investors must meet the minimum USD1 million threshold.

Sobko and Jachi say the fund is built to fuel performance and provide resilience toward volatile markets due to global unrest or global pandemics that have roiled markets of late. The algorithm is designed with an alpha-learning AI model that continues to develop and apply all future predictive models without human involvement. Comparatively, traditional hedge funds rely on AI algorithms that become obsolete in a short period due to the need for human management in the process of how data is collected and analysed.

"If you missed out on Medallion, don't miss Project One," says Sobko, CEO and Head of Business Development, Project One Capital. "We've removed the error-laden human component from this fund and employed an AI algorithm with alpha-learning capabilities toward market analysis that adapts and applies its projective models in real-time. With the sophistication of technology applied in this fund, other hedge funds must embrace AI or die."

One of the challenges facing the use of AI in hedge funds is the inability of human programmers to keep up with its speed and sophistication. The Project One hedge fund removes all human interaction and management, relying on the algorithm's alpha-learning and adaptive technologies.

"Through our study of praxeology, there is no guessing," says Sobko. "We are fully aware of the facts associated with human behavior and involvement, which is why we moved to eliminate the error-prone component from our proprietary algorithm."

During testing of the Project One's proprietary alpha-learning AI market fund algorithm, a return of 160 percent was achieved over three months, reporting limited downsides. Unlike many other hedge fund analytical algorithms, Project One analyses, projects, and applies volumes of direct and peripheral market industry data and acts in real-time without human interaction.

According to Preqin Pro, hedge funds that use AI to help with trading have been outperforming the hedge fund benchmark with three-year cumulative returns at +26.96 per cent over the past three years.