Pierre Foret was on the intensive two-year study programme that prepares students for France’s elite professional schools when he built a facial-recognition tool that scanned videos from public broadcasters on YouTube.
The artificial intelligence (AI) software identified politicians and calculated their airtime during France’s 2012 presidential election (it is a requirement that candidates gain equal media exposure during the campaign). The aim was to see if an algorithm could do this as proficiently as a human: it could.
Mr Foret then decided to apply his technical skills to take advantage of growing opportunities in the financial sector. Last year he began a $72,920 master of financial engineering programme at Haas School of Business, at the University of California Berkeley — a 12-month course that replaced the final year of his engineering diploma from French school Ensae ParisTech.
The Berkeley Haas programme provides students with mathematical, statistical and computer science skills to succeed in modern financial markets. These include how to price assets and build algorithms that scour markets for patterns to inform trades, known as quantitative, or quant, investing.
It is easy to see why quant funds are an attractive proposition for financial graduates such as 24-year-old Mr Foret. In the decade to 2018, quant funds grew steadily, with assets more than doubling to $962bn between 2009 and 2017.
The Berkeley Haas programme launched in 2001, with one cohort of 46 students. Linda Kreitzman, its executive director, says an extra cohort was added this year in response to student demand. This increase in interest was boosted by hedge funds seeking AI and programming skills to support their quant investing operations.
Ms Kreitzman says most of the graduates want to work in investment. Of the 67-strong cohort of full-time students in 2018, 65 received job offers. Of those who accepted roles, 39 per cent went into asset management, 29 per cent to investment banking, 14 per cent to fintech and 9 per cent to hedge funds.
Their average base salary was $118,530. Ms Kreitzman believes this pay level reflects the school’s reputation and high demand for “quants” who build trading algorithms. By comparison, the average salary for 2017 Berkeley Haas MBA graduates who went into financial services was $125,227.
There are risks associated with jobs in quant finance, however. Chyng Wen Tee, academic director of the MSc in quantitative finance at Singapore Management University’s Lee Kong Chian School of Business, says some traders may be supplanted by machines. “AI can analyse data and make decisions far faster than humans can,” he says.
That said, Prof Tee is optimistic about opportunities for students. While machine learning is good at analysing static data, he says, financial markets are dynamic. “Algorithms can struggle to separate the signal from the noise. You still need a human to step in and recalibrate them,” he says.
Even if machines do not replace humans, career prospects still rely on financial markets. But quant funds overall lost 5.6 per cent last year amid broader market declines, according to Hedge Fund Research.
These short-term woes are unlikely to be a threat to financial training because the skills taught — including how to spot, communicate and manage risk — make graduates employable in a wide array of industries, says Antoine Jacquier, director of the MSc in mathematics and finance at Imperial College London. Most of the 50 students on this year’s full-time course are on an optional five-month internship, he says. Half are at banks and the rest work for buyside companies (which purchase stocks and other financial products), or consulting and tech companies in AI-related roles, financial modelling or asset pricing.
Mr Foret did a three-month internship as part of the Berkeley Haas course with SumUp Analytics, a San Francisco-based data company co-founded by Laurent El Ghaoui, who teaches a class at the school. It was with him on the masters programme that Mr Foret developed an AI training method.
Mr Foret says this work — alongside being a finalist of a prestigious data science competition with his classmates — helped him win a place on a 12-month training programme with Google AI in New York from July, where he will research machine learning.
Not all quant programmes have seen a rise in applications, however. The University of St Gallen in Switzerland has struggled to attract students for its masters in quantitative economics and finance, says Christian Keuschnigg, the programme’s academic director. About 30 students enrol each year, compared with around 100 for other finance programmes at the school, he says.
“As our programme is extremely technically demanding, many prospective students shy away from doing the extra work,” says Prof Keuschnigg, though the upside is that they rarely drop out or have their applications rejected.
Applications for BI Norwegian Business School’s MSc in quantitative finance, launched in 2018, have also been slow, says associate dean Costas Xiouros. He expected 25 students to start the course in its first year, but only seven did so. He says this was because some students were worried employers would not value such a new course. Numbers are better this year, with 14 candidates enrolled so far.
Prof Keuschnigg says the risk is that St Gallen might not enrol enough students to justify the cost of running the quant course, though he does not suggest a closure is imminent. But he remains optimistic for students: “Demand for quants is outstripping supply.”
No comments:
Post a Comment