Non-Traditional Data Sources Enhancing HFT Decision-Making

Simply executing trades faster than the competition was once enough to succeed in the world of high-frequency trading (HFT).

HFT firms have always invested heavily in fast processing technologies, low-latency connections, and co-location data center space in an effort to shave every millisecond off of trade speeds. However, the proliferation of high performance computing (HPC) technologies has driven the industry to a point where lightning-fast trading is achievable for even the most modest HFT firm.

Jia Chen of Arista Networks summed up this shift by saying, “The race to zero is over. There is no more room to compete just on speed.” While processing speed will always be important in HFT, high-intelligence trading algorithms incorporating Big Data analytics and real-time market insight are now equally important in this fast-paced world of finance.

With low-latency connections to traditional market data becoming the norm, traders are increasingly turning their attention to non-traditional data sources for insight that could affect capital markets. These streams of structured and unstructured data are helping traders react to the market in real-time and uncover new advantages that might not be possible with traditional market data and price feeds.

Today’s HFT firms are enhancing decision-making using data from a variety of non-traditional sources:

• Social Media – By feeding data streams from social platforms like Twitter into trading algorithms, traders can filter out the noise and identify relevant tweets, trends, and events that are likely to impact the market.
• Weather – Changes in weather have been proven to impact trading decisions, making weather forecasts another data source to incorporate into trading models to predict investor behavior and pricing trends.
• Sentiment Analytics – Automated news sentiment analysis is gradually becoming a valuable component of an HFT strategy, helping to predict trading volumes, forecast price volatility, and provide a broader view of a specific company or market sector.

HFT firms can bolster competitive advantage by exploring and analyzing information from all available sources, however they’ll need to first ensure their computing infrastructure can support the growing demands of Big Data. Unstructured data, especially from prolific platforms like Twitter, flows continuously and changes rapidly, making HPC solutions capable of quickly processing large and complex workloads a key enabler for today’s HFT firms. Powerful hardware and software solutions optimized for HFT workloads can help traders speed innovation and operate at the pace of the market as new and expanding data types make their way into the trading decision-making process.

As the high-frequency trading industry continues to fine-tune its use of data analytics in automated trading models, traders must find new ways to manage data from non-traditional sources in order to make intelligent trading decisions. High performance computing architectures built to support HFT workloads will provide the necessary support for Big Data and enable firms to differentiate their trading strategies, gain a better understanding of market dynamics, and secure competitive advantage.