After a pandemic-stricken year that catapulted some of the most aggressive retail traders to millionaire status, the idea of turning to computer-driven quantitative hedge funds for outsized returns can seem unnecessary.
Renee Yao, the founder and portfolio manager of quant hedge fund Neo Ivy Capital, begs to disagree.
“For the past six months, there’s been a lot of retail flows into the market and the retail flows love those high-volatility, high-returns penny stocks,” said Yao, noting that meanwhile, blue-chip stocks attracted little inflows.
That discrepancy breeds the danger for a correction — something that Yao’s fund is always monitoring so that it can act fast in a chaotic market environment. An example of such drama was the Dow’s 13.7% decline in March followed by its 11% gain in April.
“If you use the traditional quant investing methodology, which you basically train your models with historical data, your model is going to get lost, because those are very rare events,” Yao said.
That’s why so many traditional big-name quant hedge funds were hit with poor performance in 2020. For example, Jim Simons’ Renaissance Institutional Diversified Alpha fund was down 32% while Ray Dalio’s Bridgewater Pure Alpha II fund declined 17% last year.
How does an artificial intelligence strategy work?
Yao’s fund, which manages about $200 million in client assets, has generated a consistent 10% in annualized return since its inception in June 2016 thanks to its artificial intelligence-driven model.
“The whole purpose of AI is to train the computer to think and analyze like a human being; it’s not entirely reliant on historic data,” she said. “So our model was able to quickly pick up the change in market regime and deliver positive returns for both March and April last year.”
Instead of training the computers with historical data and hoping history repeats itself, Yao and her researchers write their own artificial intelligence algorithm that automatically generates trading ideas for the fund 24/7, non-stop.
It does so by collecting the relevant data, cleaning the data, processing the data-generated forecasts, and then translating the forecasts into tradable targets, according to Yao.
“For example, the AI might say that today Elon Musk sent a tweet about Tesla, it’s a good tweet, it’s a piece of good news so we think the Tesla stock will go up,” she said. “So it’s going to give a positive conviction to the Tesla stock.”
She continued: “Then the algo will go ahead and incorporate that conviction into the risk constraint we have, and then translate that conviction into how many shares of Tesla we want to buy, and then sent the order to the market.”
With no human discretion in the stock selection process, the fund holds a diversified pool of about 1,000 stocks.
A new generation of quants
Yao represents a new generation of quant hedge fund managers who are embracing artificial intelligence to harness the infinite possibilities in financial markets.
But she technically grew up in the world of traditional quantitative investing.
After finishing her master’s in statistics at Columbia University, Yao started her career as a quant researcher at hedge fund giant Citadel before joining Millennium Partners and WorldQuant as a portfolio manager.
“As I continued to work in traditional quantitative investing, I started to observe that the space has become very crowded, and more and more people are trading similar ideas,” she recalled. “So we were trying very hard to find a new angle, a new source of alpha that is uncorrelated to the traditional way.”
“Once you build this AI system, the system will automatically find new ideas for you, which completely frees up the human resources,” she said. “We don’t need to hire hundreds of thousands of Ph.D.s like some of the second generation traditional quant funds do. Instead, we just build this system that will automatically do all the jobs of the Ph.D.s.”
She added: “In terms of business model, this is a good business model. In terms of the results, obviously, the results are quite good compared to traditional quantitative investing.”
As the quant investing landscape evolves, Yao continues to upgrade the system and technology that support her hedge fund. Aside from work, she is also trying to support the industry by helping to groom the next generation of quants.
“Whenever I have time, I would actually go back to Columbia and attend some of the seminars they hold at the business school and statistics department,” she said. “Whenever I have the opportunity, I will go there and try to encourage more women to focus on science and mathematics.”