Interview with Jared Broad, CEO, QuantConnect – Part II

Click here to read Part I.

When QuantConnect (QC) Founder and CEO Jared Broad launched his firm in 2012, he knew that a great many members of quant forums and communities were having trouble getting started as traders. He knew this because until two years earlier he had been one of them. That was when he quit his day job and put all his time into building a quant engine.

Broad estimated there could be millions of investors as passionate as he who, because of day jobs and families, lacked the time or resources to build infrastructure themselves. Capturing even a small piece of that market would amount to a sizable operation.

Broad notes that at the time, only a few dozen firms in the world could support effective quantitative trading, and all their platforms were closed source. By embracing an open-source ethos, QC took a dramatically different approach.

“The world’s largest quant firms have thousands of engineers, but few have anything to do with quant research. They are dedicated to infrastructure, trading, connectivity and data sourcing. With our open-source, community-focused approach, we knew we could deliver the same kinds of services to anyone who wanted them, much more cost-effectively.”

True to his vision, QC today is an open-source quantitative research, algorithmic and trading platform. Roughly half their business still comes from individuals, but organizations now generate the other half. With QC handling the plumbing, operational costs are dramatically reduced for firms of all sizes.

“Our platform is a classic example of being incredibly simple for users and incredibly complex for us. Many firms move into quant trading assuming they can build their own platform. The problem is they don’t know what they don’t know. Even with plenty of resources, they soon start hitting roadblock after roadblock. The engineering cost blows up their budget, and they often don’t address certain key elements – things we only discovered through years of trial and error. When they deploy, they start losing money and don’t know why.”

He smiles. “This stuff is hard to do.”

Broad’s plan for the next 10 years is to keep adding data sources and code to QC’s core quant engine, called LEAN, in pursuit of his larger goal: inducing a shift in the quant industry as a whole.

“The big firms still build in silos and keep their code in-house. As far as we know, no other open-source platform is as close as we are to reaching critical mass in providing the same value on an open platform as the big firms have in-house.”

For clients of Interactive Brokers, QC hosts the IB Gateway Java app. When the clients run their algos, LEAN routes them through the Gateway and directly into TWS. Broad says QC now drives a significant amount of trading to IBKR.

Academia

Broad’s vision of QC’s future includes a commitment to academia.

“We either gift or heavily discount QC for professors,” he says. “Students perform research assignments on the platform and compete using sample data or live trading strategies. It gives them hands-on experience with quantitative finance. Professors can easily manage student progress with our collaboration features.”

As an example, Broad cites a professor in Duke University’s Master’s in Economics program. He was looking for a solution to quantitative education when he found QC’s all-in-one package: education, open-source code, 100+ data sources, paper trading capability and thousands of industry professionals validating the platform.

Support for the research papers published by academics and doctoral students in professional periodicals like the Journal of Finance is a recent addition. According to Broad, the results reported in many of those papers can be difficult for other quants to replicate.

“It’s such a significant problem in academia that it has a name: “The Replication Crisis.” We’re well-positioned to address that directly. We’ve started giving free subscriptions to anyone publishing a paper if they’re willing to publish their QuantConnect code along with their work. Once the code is on our platform, users can read the papers and reproduce the reported results with a single click. They can backtest it over and over and get the same outcomes, then experiment with modifications. Roughly 1,000 papers are published each year, and we’re offering 500 subscriptions. Depending on how many take us up, we could soon have half the industry’s annual research — all reproducible by any QC subscriber with a single click.”

Interested academic programs can contact QC at contact@quantconnect.com for more information.

Individuals and organizations can subscribe to QC directly at www.quantconnect.com or through IBKR’s Investors Marketplace.