Herbert & Mason

TL;DR

Built a custom trading system for energy futures specialists Herbert & Mason, enabling them to automate complex strategies, improve data quality, and optimise trade execution across varying liquidity conditions.

About

Herbert & Mason are commodity trading advisors (CTA) specialising in energy futures trading. With over 10 years of expertise in futures markets, they've built a reputation for successful manual trading strategies. Their decision to automate existing approaches while developing new ideas marked a significant evolution in their business model, allowing them to scale operations without sacrificing performance or increasing headcount.

Challenge

When Herbert & Mason approached me, they faced several obstacles in automating their energy futures strategies. After 10+ years of market expertise, they wanted to automate their trading strategies to reduce manual monitoring and also to be able to find new ideas and stress test their existing ones.

One of the key challenges was executing interdependent orders simultaneously across markets with different liquidity profiles. This also meant being able to effectively combine positions and to monitor them for signal generation, which added significant complexity.

The data challenge was equally troublesome. Less liquid markets provided poorer quality data, creating false signals during backtesting that showed phantom liquidation events that never actually occurred in the real market.

The team also needed a way to systematically test variations of their strategies to optimise performance and discover new opportunities.

Solution

I tackled execution first, designing a system that placed limit orders in less liquid markets while completing positions with market orders in more liquid contracts. This approach minimised spreads and trading costs while optimising entry prices.

For the data issues, I sourced higher quality tick data from CQG and built a pipeline to transform contract prices into both continuous and raw price series for accurate testing.

Finally, I created interfaces that allowed the team to easily select contracts and parameters for optimisation tests, making the entire system accessible to traders without requiring coding knowledge.

Result

The automated system enabled Herbert & Mason to develop complementary strategies that diversified their portfolio. This improved overall performance, increased Sharpe ratios, and expanded the capacity of their trading operation without requiring additional manual oversight.