Canada Startup Preps ‘Level 3' Indicators

Calgary, Alberta-based analytics startup Level 3 Data has officially launched a service for quantitative traders and hedge funds that provides an indicator of buying and selling pressure to improve trading returns and minimize risk, and is now offering a 90-day free trial of its data to attract early adopter clients.

The vendor's Buying & Selling Pressure Indicator uses a patent-pending algorithm that analyzes real-time tick data to consolidate volume at each side of the spread, providing a signal of supply and demand, as well as price exhaustion, which can be integrated into charts or captured as raw signals in one minute-intervals, or as historical data for firms to back-test the signal's effectiveness.

Level 3 was founded in 2007 by lead researcher Juan Guiterrez and president Don Soubolsky, a former stockbroker, who describes himself as an "accidental quant" after developing an algorithm that could reliably demonstrate buying and selling pressure by analyzing trading activity between the bid-ask spread.

"We're a startup, but we've been collecting data since 2007. The reason it took so long to prove our data is because it is very spikey -- and any spike in volume also causes a spike in our data. So only after several years of analysis and algorithm development were we able to identify our two data points," Soubolsky says. "Our system incorporates the benefits of tick trading data without the challenges of handling it.... The most powerful aspect of our data is the ability to identify price exhaustion. By identifying price exhaustion, you can filter out unprofitable trades.... And when used in conjunction with other risk metrics, it provides consistent returns."

The service initially runs on data on US and Canadian equities and exchange-traded funds, and though Soubolsky says the vendor has also received demand for the ability to apply the analysis to futures and commodities markets, he says there are no plans to expand it beyond equities, adding that firms can effectively get exposure to other asset classes via ETFs.

While a firm could create a similar algorithm itself, the advantages of using Level 3 are its simplicity and the advantage of gaining time to market over building it themselves, Soubolsky says. "We aren't asking an arm and a leg for our data... so the benefits are the convenience and complexity," he says, adding that clients can access the data on Level 3's servers using almost any programming language. "With just a few more lines of Matlab code, you can quickly and easily determine whether this has value."

Soubolsky says the next step for the vendor will be to get the service in front of clients that can provide feedback for Level 3 about how they use the service -- either via direct sales of its service or via partners. For example, he says Level 3 is in talks with an unnamed trading software provider about incorporating its analytics into the vendor's systems, which would potentially bring Level 3's service to a wider audience.

"We expect a combination of direct and channel sales. We're a small company, and we realize that there are a lot of potential clients out there, so partnering with another company would be very desirable for us," he says.