I have written about a number of startups in the computing realm that have the potential to begin leveling the playing field in a number of finance-related categories.
Here's another new one: QuantConnect.
Its co-founder and chief executive, Jared Broad, recently got in touch with me and we had a great and lengthy conversation. Today he passed along a bit of news about what he and his compadres have built: They have added 15 years of US equities data and five years of market-trading data to their powerful algorithm backtesting platform.
Specifically they have brought in market-trading data — also known as tick data — from several sources, including Forex data from FXCM, sentiment data from Estimize and StockPulse (other data providers are listed on the website).
That data, when combined with their backtesting tools, can empower independent engineers to code, test and iterate algorithms — and should provide a wee bit of friendly competition to another site I'm a fan of: Quantopian.
The two sites essentially share the same mission. As Mr. Broad put it, he is seeking to democratize algorithmic trading by giving engineers access to free financial data, powerful cloud computing and strategy back-testing.
"We launched QuantConnect with the goal of bringing algorithmic strategies to the mainstream engineer and investor, so they can be empowered with cutting edge investment strategies," Mr. Broad said in a prepared statement that announced today's data integration. "By providing a platform with unlimited free financial data, we're giving access to tools that would typically cost $50,000-$100,000."
"This product is way ahead of retail options currently available," said Alejandro Cañete Báez, adviser and head quant at Pan Alpha Trading.
The idea behind QuantConnect (as with Quantopian) is to allow users to build and test algorithms in minutes rather than days.
QuantConnect is providing a secure repository for publishing and maintaining an application programming interface, allowing teams to better collaborate and upload encrypted algorithms.
"We are building a global network of engineers designing diverse new strategies, while helping investors across the world earn better returns with the strategies that suit them," said Shai Rosen, QuantConnect co-founder and chief operating officer. "We are working hard to make our vision a reality by building partnerships with data providers, listening to our quant community and continuously adding features to our platform."
I remain intimidated by high-powered forumlae and the very notion of the algorithm itself.
Mr. Broad, in our past conversation, tried to put my fears to rest.
“I don't think it is anything to be scared of, and we hope to empower people and make them not afraid of it; it is really sort of the next evolution [of financial tools],” he explained, adding that just as the larger financial community would not really want to go back to the slide rule or abacus, quants only want to see more sophisticated computational tools developed and available.
Mr. Broad said he had begun his professional career as a biomedical engineer, but three years ago embarked on a project where he was designing algorithms for a fund company.
It was then that he realized the “double work” most anyone doing similar work had to undertake. This included not just the algorithms themselves but usually construction of a test environment as well (and usually a proprietary one that remained the firm's intellectual property). So in August to November 2011 he and Mr. Rosen went about setting up QuantConnect.
Part of what intrigued me about Mr. Broad's story is how international it is in scope.
The two won a $40,000 grant from Startup Chile (yes, the nation in South America). Everything else Mr. Broad explained was pretty much bootstrapped.
Mr. Broad, a New Zealander, is among the million or so Kiwis that find themselves working or travelling/adventuring overseas at any given time. He explained that part of his reason for being in Chile was a familial connection, but added that Chile has a great balance between lifestyle and work ethic.
He also added that while he had garnered interest from hedge funds and some other angel capital sources early on, he and his partner had preferred to suffer now by bootstrapping their development work and avoiding potential conflicts that sometimes accompany heavy outside investment.
As for a few of the technical facets, the platform is based on C#, but the firm is adding other language support, which Mr. Broad noted can end up being pretty much whatever is demanded by users or potential users.
For now, QuantConnect is being hosted on Amazon, which Mr. Broad noted is a little bit pricey for a startup but at the same time provides for most any level of scalability the firm might require.
“That's how we can achieve the speed we do with the platform because of a layer added on top of Amazon,” he said. “In that way we can take terabytes of US equity data and spread it over a thousand-server Amazon environment and compute in parallel.”
That makes for backtesting an algorithm in minutes instead of hours, especially when one considers there is typically 20MB of data per stock, or about 4,000GB of data total in the platform's dataset.
“So the way I see it, as interstate highways have enabled commerce throughout the United States and the developed world, this incredibly fast infrastructure we are leveraging and building out in our platform is going to enable a lot of innovation,” he said.
He went on to explain that the site will have three levels of subscriptions. For the independent quant user community as well as independent engineers and developers, the site will remain free. For educators and professionals at institutions, there will be an as-yet-unpublished fee structure.
“Our plan is to monetize what we have built later through institutional subscriptions. Right now I just want to make it a beautiful product that helps people design algorithms,” Mr. Broad said.
For more information visit Quantconnect online, where you will also find a tutorial and quick-start guide.