☰ Menu eFinancialCareers

Why the top jobs in finance will go to gig workers

algorithmic trading, trading algorithms, trading algos, algo developers, developers, programmers, coders, computer programming, computer development, Python, programming languages, hedge funds, quant hedge funds, quantitative finance, quantitative research, data science, big data, buy side, hedge funds, crowdsourcing, gig economy, crowd-sourcing

Consider a career as an algorithmic trading freelancer.

The crowdsourcing model for developing trading algorithms has gained traction on the buy side and is picking up momentum. Just as Uber and Airbnb have disrupted their respective industries, it is only a matter of time before “the gig economy” makes an impact on the sell side as well.

Tom Ducrot, the founder of venture-capital firm Fides+Ratio and a former executive director at Morgan Stanley, moderated a panel at the Battle of the Quants conference in New York that included four of the biggest players in the algo crowdsourcing space: Quantopian, Quantiacs, WorldQuant Challenge and CloudQuant.

These firms are in effect crowdsourced quantitative hedge funds – rather than employ in-house developers, they host competitions and invite freelance developers to write trading algorithms, which are then back-tested. They put real money behind the best of the bunch, and the trading algo development gig workers get a cut of the profits.

The way they tell it, this model is already starting to upend the asset management space and is likely to grow in significance. If you’re a great coder who would never be caught dead wearing a suit and tie or working at a bank or hedge fund, this could be your chance to get in the quantitative finance game – without changing out of your bathrobe or leaving the house.

Printer repairman by day, trading algo developer by night – if you’re lucky

Jonathan Larkin, the CIO of Quantopian, whose resume also includes J.P. Morgan, Millennium Partners, Nomura, BlueCrest and Hudson Bay, said he believes the addressable market of people with STEM-based backgrounds is around 25m globally. To date, the firm has attracted 140k people to its website who have used the research and data environment to back-test their trading strategies. A mere 25 have had their strategies licensed however: a success rate of just 0.02%.

“We don’t profile people when they come on board, before we license the strategies we get to know them and do full background checks,” Larkin said. “We’ve licensed strategies from 25 people living on five continents, and the common denominator is their technical experience in a modeling field, typically not financial services, but everything from academia to the oil and gas industry and someone whose day job is calibrating printer jets.

“If we license a strategy then we can act as a consultant, and our only revenue is as an asset manager operating strategies from our community,” he said. “It’s our intention to invest significant capital in these – right now it’s a maximum of $10m per algorithm but we’re trying to ramp up to $50m by the end of the year.

“We pay people a fixed percentage of P&L just like I used to do in a former life working at a multi-manager hedge fund.” Larkin omitted to mention what that percentage is.

The gig economy plus data science plus algorithmic trading equals crowdsourced hedge funds

Martin Froehler, the CEO of Quantiacs, formerly with IdeaLab Research, identified two big trends that are converging: first, the decentralization of labor spurred by companies like Uber and Airbnb; and second, data science becoming more mainstream to the point that it is now frequently taught in universities. His firm hosts Python coding competitions and offers machine-learning libraries.

“Crowdsourced algorithmic trading is at the intersection of these two friends,” Froehler said. “Quants get to retain their intellectual property, and the only risk that they run is that they don’t manage money, but if they do manage money, then there is unlimited upside.”

Coders without borders

With these crowdsourcing platforms available on the Internet, it democratizes the allows someone in a remote village in India, a farmer in Brazil and a video game developer in Russia to potentially monetize their trading algorithms, said Rich Brown, the CEO of WorldQuant Virtual Research Center a.k.a. WorldQuant Challenge. The firm has around 50k active users on its platform, the majority from academia. He says around 25k trading algos, which he calls alphas, have received assets.

“These coders want to work at home at their leisure to develop these strategies, and there are people all over the world pursuing these opportunities,” Brown said. “Even undergraduate students, if they are engaged in our consultant networks, then they can develop these alphas, and they are getting used.

“If you can come into these platforms, incorporate data, whether it is alternative or traditional data, get an understanding of this industry and apply your skills and get paid, that’s great.”

Quantitative research and data science applied to create trading strategies

Morgan Slade, the CEO of CloudQuant, formerly with Merrill Lynch and Citadel, said they officially put their website up publicly six months ago and now researchers in 70 different countries are using it already. He sees this crowdsourced development model as an opportunity for students, recent graduates and career-changers alike.

“We’re tapping into the new skills coming out of educational institutions and students’ and graduates’ new ways of looking at things, but there are also opportunities for experienced people to connect the dots related to the ontological relationships between the data and the stock markets and other assets,” Slade said. “There are huge untapped resources out there, and we try to engage with the researchers as if they were employees and support them as such.”

Two of the freelance researchers wrote trading algorithms that were so impressive that Slade hired them as full-time portfolio managers.

“We’re building [in-house] teams to support [freelance] researchers’ investment strategies, and we expect that to continue,” Slade said. “If we see people who are especially talented, then we might decide to have remote teams centered on individuals who stand out.

“To be successful, they have to master a prediction step, a portfolio construction step and a trading expression step – we give researchers the chance to take a stab at portfolio construction and trade expression on their own,” he said. “We help them with things they’re not going to be good at, such as trade expression – we help them get better execution and give advice on portfolio construction if they need it.

“People are spending their free time doing research, and if they weren’t getting something out of it, then they wouldn’t be engaged with our site, and there are also experienced traders who don’t have the ability to build what we have – people are eager to retain their IP, which we let them, so add it to our platform and we can both make money off of it.”

Photo credit: bowie15/GettyImages

Comments (0)

Comments

The comment is under moderation. It will appear shortly.

React

Screen Name

Email

Consult our community guidelines here