Data One Global: Machine Learning (ML) Driven Execution for Systematic Trading

Data One Global: Machine Learning (ML) Driven Execution for Systematic Trading

CIO VendorAhmad Nahhas, Founder and Chief Architect
"Data One is the product of an overregulated financial industry that was created after the 2008 economic meltdown," says Ahmad Nahhas, Founder and Chief Architect of Data One Global LLC. When the financial crisis occurred, the major banks focused primarily on cost cutting in an effort to survive. Ahmad, who spent the majority of his career designing and implementing predictive quant models for several of Wall Street’s banks, set out to capture the opportunity created by this lack of innovation. In 2013, Ahmad formed Data One Global to create leading trading products for the financial services industry.

Ahmad leveraged his in-depth knowledge of machine learning, algorithmic design, and market micro-structure across multiple markets and asset classes to create Data One’s Directional Probability platform. The platform is built for efficient and dynamic research & development, alpha-generation, and machine learning-driven execution with a focus on hedging and risk management. Its next generation design allows it to consume massive amounts of structured and unstructured data to produce real-time probabilities for markets and individual securities--continuously learning new information, detecting patterns across millions of relationships, and standardizing behavioral tendencies through time. “There is no question in my mind that Data One is light years ahead of its time when it comes to infusing machine learning into conventional execution strategies. For the first time in my trading career, the LIVE Sharpe ratio and returns of our ML-driven execution are consistently outperforming the simulations using conventional execution algorithms,” adds Ahmad.

With Innovation comes Success

A large investment bank leveraged Data One’s ML platform to enhance their market making capabilities. Another firm is utilizing the Data One execution engine to achieve more efficient execution and enhance its returns.

Data One’s platform standardizes market behavior by systematically constructing security-specific DNA strings capable of predicting short term price movements

For one strategy, Data One was able to demonstrate an increased return of 1.5 bps per day utilizing its execution algorithms. In each case, the Company’s technology generated enhanced returns with substantially lower risk. Data One extends sophisticated technology capabilities to its clients:
• Probability driven price enhancements that dynamically adjusts to market-microstructure
• Dynamic volume arrays vs. static volume arrays in a traditional VWAP
• Passive execution while avoiding predator HFT algorithms

8 months analyses of Data One’s Live SPY (S&P 500) strategy executed through the Machine Learning based engine vs. conventional algorithms (TWAP / INLINE). Strategy is active 5-10 times a month.

“Data One deploys algorithms and creates effective execution solutions to solve each of its client’s complex problems. We have demonstrable data that showcases our technology’s outperformance versus conventional algorithms like VWAP, TWAP, or INLINE,” adds Ahmad.

With plans to expand horizontally and vertically, the firm is seeking to continue to grow its foothold in the equities space by providing enhanced execution solutions to satisfy many different client needs. “The goal is to become a niche player and create more success stories in the equities space and then expand to other asset classes and geographical areas,” concludes Ahmad.