Want to take on Wall Street? Quantopian’s algorithmic trading platform now accepts outside data sets
Quantopian has built an online playground for quantitative traders, making it easier than ever for millions of everyday engineers, mathematicians, and statisticians not on Wall Street, and thousands of professional quants, to build, test, and execute quantitative trading algorithms. The platform exited beta in January of this year, and at the time, challenged investors to run their algorithms against 10 years of historical market price data to see if they could gain an edge.
Today, the company announced Fetcher, which provides the option to import any external data set into the platform to better reveal market insights and, hopefully, generate additional profits. These data sets are likely to include commodity prices, currencies, short interests, obscure indexes, and derived trading signals, but the possibilities are endless.
There are more than 4 million data sets available through Quandl alone, not to mention Yahoo Finance and other platforms, as well as those that individuals personally gather through observing the market. As long as the data is organized as a time-series and comma delimited, Quantopian will clean, normalize, and sort the information against existing price data.
“You can imagine investigating things like how the unemployment rate of young professionals affects the profitability of hotel chains,” Thomson Reuters director of quant product strategy Jessica Stauth says. “This opens some very interesting new areas for algorithm development and makes it even easier to go from idea to trade.”
“With Fetcher, we’ve essentially turbo-charged the platform with the power to explore and execute on the furthest reaches of quant creativity,” Quantopian founder and CEO John Fawcett says. “The volume and diversity of the data available out there is unbounded and we’re expecting some stunning results as people bake it into their algos.”
Fawcett and his co-founder Jean Bredeche set out initially with the vision of creating a product for everyone not on Wall Street. This goal informed all their design decisions, with the result being an easy to use product that allows for simple iterative algorithm creation and testing. Consequently, but not surprisingly, it has proven to be an enormous hit among Wall Street traders and hedge funds as well.
“The process of researching an idea and then implementing it in the market is traditionally highly disjointed,” Fawcett says. “We wanted to shrink the creation cycle by running backtesting in real time to give visual feedback. The idea was to lower the slope for getting started. It turns out, the benefits of this approach are just as great, if not greater, for people already in the industry.”
Quantopian solves several problems for professional traders. First of all, it gives them a place to quickly and easily test new ideas. But more importantly, it gives these practicing quants a place to implement a different strategy with their personal money than they do with their clients or to take the first steps toward leaving their jobs.
Equally valuable are the community and mentorship relationships that are forming on the site, according Fawcett. For an introverted group that’s traditionally unwilling to share their secrets, Quantopian offers a rare place to gather among like minded individuals and break proverbial bread.
Regardless of whether a quant is professional or a hobbyist, they tend to be very sensitive about their intellectual property. Quantopian has gone to great lengths to ensure that all algorithms will be kept private and never takes possession of third-party data files, but rather runs them through simulations line by line without ever saving them. This method also makes it easier for the company to negotiate first-party data partnerships.
Quantopian is not monetizing yet, but the founders have a number of ideas they plan to implement in the near future. They expect that exploration and research within the platform’s first-party data will be free forever, however it’s likely that they will soon charge a subscription fee or transaction fees around live trading. Also, Quantopian is likely to evolve into a marketplace where enterprising quants can offer access to their algorithms or proprietary data for a fee, should they choose. When this becomes a reality, the company would participate in those transactions in some manner, as well.
Financial services is one of the most heavily regulated industries of all, something Fawcett and Bredeche hope to steer clear of however possible. That’s why they’ve taken the route of a technology platform, not a broker dealer. Quantopian integrates with the APIs of popular online brokerages and routs trades, but does not own or manage customer accounts. This may ultimately limit their income potential in certain ways, but will pay enormous dividends in terms of reducing complexity, liability, and hassle.
The Fidelities and Schwabs of the world care about two things only: Assets under management and transaction fees. As long as Quantopian drives both, the industry titans are unlikely to feel threatened by the upstart platform. Should this change, or should the brokerages see an opportunity to extract additional fees out of their clients by offering an algorithmic trading platform, Quantopian could encounter stiff competition – or a highly motivated acquirer. It’s likely too early to worry about either of these, however, as the startup is simply focused on delivering value to its users and continuing to improve the product.
On the startup side, Quantopian faces pseudo-competition from Motif, which offers those less mathematically inclined the opportunity to create themed investment strategies based not on data and algorithms, but on personal areas of expertise or interest. The company is utilizing a similar marketplace model where community members can market and sell their ideas.
Approximately 18 months after the first line of code was written, Boston-based Quantopian has grown its team to seven people, and announced a $2.1 million Seed round from Spark Capital and GETCO in January 2013.
“We’ve been measuring adoption in terms of site traffic, app usage, and followers to our open source project,” Fawcett says. “We’ve seen a huge uptick in algorithm development and forum activity in the last three months. Also, our reach has been great, with users following a weighted distribution geographically among major financial hubs, including London, Singapore, Hong Kong, NY, obviously, and even the Nordic countries.”
Quantopian stands out by democratizing algorithmic trading. This began by providing the individual access to tools and a community around data-centric trading previously unavailable anywhere. The addition of Fetcher stands to open up additional avenues of creativity and analysis, and paves the way for a forthcoming marketplace that could inject growth hormones into the platform. For any mathematicians, engineers, and hackers out there who ever wanted to test their mettle against Wall Street’s best, Quantopian is like a Disneyland for geeks.