Backtesting seeks to estimate the real-world performance of specific high-frequency trading (HFT) strategies or models as if they had been employed during a previous trading period. Levyx developed its hardware-accelerated Risk Analytics Acceleration Framework to help customers realize more profit from an HFT financial model by speeding up the calculations associated with that trading model. Backtesting these HFT financial models during their development explores the profitability of a new trading model by running the model’s computations over a year of historical price data for a given security. These computations are particularly data- and compute-intensive, which makes backtesting a great candidate application for hardware acceleration using FPGAs.
FPGAs can accelerate many kinds of mathematical calculations including backtesting. Faster backtesting computations allow brokerages, investment banks, and hedge funds to test and deploy more robust and profitable trading models more quickly across their electronic trading and risk-management systems, which in turn reduces financial risk and increases profitability.
The Levyx Risk Analytics Acceleration Framework is based on the company’s highly scalable, low-latency Xenon data-analytics processing engine, which Levyx developed to manage the retrieval, processing, and indexing of very large datasets. The Xenon engine bypasses conventional file-systems and kernel buﬀers and directly performs SSD-friendly I/O operations to reduce I/O latency and maximize I/O bandwidth for storage devices. In particular, the Xenon data analytics engine allows application programs to treat storage space on Intel® Optane™ NVMe SSDs like RAM, which significantly speeds operations.
The Xenon engine also uses just-in-time (JIT) compilation to generate analytics code on the ﬂy and then executes that code on Intel® Xeon® CPUs for maximum efficiency and throughput with low latency. Optionally, the Xenon engine can push complex calculations to hardware accelerators such as the Intel® Programmable Acceleration Card (PAC) with Intel Arria® 10 GX FPGA to further increase performance.
For testing, Levyx created a demo with an 8 Tbyte dataset representing one year of trading data on the NASDAQ stock exchange. A sample financial model then calculated 10-minute moving averages over the year-long dataset and compared the most recent price tick to the moving average at each 10-minute interval. The financial model made trades when the last tick price was a specified distance away from the average based on a mean-reversion signal.
This demo traded synthetic stock, which is an options trading strategy that trades at-the-money call options and an equal number of at-the-money put options of the same underlying stock with identical expiration dates. Synthetic stock trading is a low-cost alternative to purchasing the stock outright, but this type of option-based trading strategy is computationally intensive. The backtesting demo application devoted more than 70% of its elapsed run time to the calculations while running on a CPU. Adding FPGA-based acceleration reduced the calculation overhead to just 30% and cut the overall run time significantly.
For many more details, see the White Paper titled “Why FPGA actually stands for “financial programming greatly accelerated,” available on the Intel® FPGA Acceleration Hub.
For more details about the Levyx Risk Analytics Acceleration Framework and the company’s Xenon data-analytics processing engine, please contact Levyx directly.
Legal Notices and Disclaimers:
Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No product or component can be absolutely secure. Check with your system manufacturer or retailer or learn more at intel.com.
Results have been estimated or simulated using internal Intel analysis, architecture simulation and modeling, and provided to you for informational purposes. Any differences in your system hardware, software or configuration may affect your actual performance.
Intel does not control or audit third-party data. You should review this content, consult other sources, and confirm whether referenced data are accurate.
Cost reduction scenarios described are intended as examples of how a given Intel- based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries.
Other names and brands may be claimed as the property of others.