Sydney-based hedge fund startup Minotaur Capital, which has replaced traditional analysts with artificial intelligence, has beaten the global stock market in its first six months while significantly reducing research costs, according to a report by Bloomberg.
Founded by Armina Rosenberg and Thomas Rice, Minotaur achieved a return of 13.7% for its flagship fund, the Minotaur Global Opportunities Fund, in the six months ending January, outperforming the MSCI All-Country World Index's 6.7% return over the same period.
Rosenberg, who previously managed a global equities portfolio for tech billionaire Mike Cannon-Brookes and led JPMorgan’s Australian small-cap research at just 25, and Rice, a former portfolio manager at Perpetual, built the fund around an AI-driven investment strategy. Unlike traditional hedge funds, Minotaur employs no analysts, relying entirely on AI models to analyse stocks.
“AI is faster, cheaper, and more efficient,” said Rosenberg, 37. “We’re looking at about half the price compared to a junior analyst’s salary.”
Minotaur’s large language model processes 5,000 news articles daily, generating comprehensive reports of around 2,000 words on global stocks with high growth potential. The fund targets companies it believes can double in three years or achieve tenfold growth over the next decade.
The fund charges a 1.5% management fee and a 20% performance fee on profits. By the end of 2025, Rosenberg expects Minotaur to manage around AUD50m ($31m).
Where deep modelling meets mortgage credit
Greg Valli has been part of two of the three decades of Ellington Management Group. During that period, the firm has enjoyed a rapid rise…
More
Crowded Treasury shorts raise prospect of sharp squeeze
Growing bets on further rises in US Treasury yields are increasing the risk of a rapid reversal if incoming economic data points to a…
More
Hedge funds take record 7% share of Treasury market
Hedge funds have become an increasingly important presence in the roughly $30tn US Treasury market, helping absorb government debt as…
More