These are mathematical models whose parameters remain a closely guarded secret, and in the world of investment funds, they have become omnipresent. “For us, 30% of our business portfolio is chosen by artificial intelligence [IA] », Specifies a manager.
In the world of investment, quantitative specialists (“Quants”) are managers apart: the big stars do not come from trading rooms, but from hard sciences, such as mathematicians or physicists, who rely on increasingly on AI tools, in addition to their own analyzes on the solidity of companies.
The most famous fund is that of mathematician James Simon, Renaissance Technologies. Its star investment, the Medallion Fund, has the status of “the best performing fund in history”, with more than 35% annual gains after management fees since 1988, according to the Wall Street Journal.
But its investment strategy is secret and only employees of the company can now invest in it. It obeys predefined rules independently of the economic situation, like all funds that claim to be part of this so-called “quantitative” finance: the manager’s assessment of the companies fades in the face of the algorithm’s perception of the market.
“The primary mission is to measure risk” in financial markets, explains Stéphane Levy, head of innovation at Chahine Capital. “Quantitative funds are required to be more diversified” and more selective, “because they work according to the statistical laws of large numbers,” he explains.
Its fund monitors 1,500 financial securities, analyzed once a month to determine the selection of companies in which to invest: 30% are even chosen solely by AI. “A review of our factors by hand would take months, now everything is faster,” points out Stéphane Levy.
“Crazy amplitudes”
Each manager develops his own method. “The financial markets, in their mechanics, have changed” in their approach and their reaction, thanks to “new products and new flows”, which lead to “movements with crazy amplitudes”, underlines Léonard Cohen, president and head of management of Ginjer AM.
The company has designed its own “risk matrix”, which advises this or that investment, sometimes even going against the grain of a collective intuition on the market, he explains. However, traditional analysis remains necessary in addition to this use of matrices.
For these managers, access to as much information as possible is therefore key. The financial markets are full of them. “A lot of data is public, but if we want to recover the order book to the millisecond”, that is to say all the transactions carried out on the markets, “it costs millions of euros, both for buy them only to process them,” notes Michael Benzaquen, a researcher from the world of physics who works at Capital Fund Management.
This fund, which claims $10 billion under management, has “more than a hundred” computer scientists, or nearly a third of the workforce, to integrate this information, as well as around 80 researchers.
With this strike force, “we are able to see more things” on the markets than the stock exchange operator Euronext, assures Michael Benzaquen.
Two Nobel Prizes
The quantitative technique is used in particular by hedge fundslarge, essentially American funds that do not hesitate to take risky bets and are not subject to the same transparency obligations as other types of investment funds.
Mathematical formulas are used to model the market in order to better predict its reactions and anticipate variations. With this, the model “detects a market inefficiency and attempts to exploit it” to profit from it, explains Stéphane Levy.
The main risk of this strategy is not integrating the right parameters into the calculations. In 1998, the LTCM fund came close to collapse during the Russian financial crisis despite the presence among its directors of two Nobel Prize winners in economics, Robert Merton and Myron Scholes. He had to be rescued urgently to avoid a domino effect.
In 2024, these funds are also in the sights of the Chinese authorities, who are tightening their control over financial market players. The “Quant” fund Ningbo Lingjun was suspended from the Shanghai and Shenzen stock exchanges in mid-February after selling a lot of shares in a short time.