David Hathaway is a systematic trader. Jon Farrin is a discretionary trader and expert in financial markets. The two have worked together for years and recently discovered that combining their styles creates a program much greater than the sum of its parts.
Hathaway and Farrin began working together at Wachovia in 2003. Farrin was a bond trader who eventually led Wachovia’s trading desk. Hathaway was a salesman for the products traded by Farrin and others but also a systematic trader looking to get out on his own. He studied under famous Turtle trader Jerry Parker.
“We worked closely together as salesman and trader building the Wachovia REITs business,” Hathaway says. “When Jon and I started to work together, we realized that our business philosophies, life philosophies and trading philosophies meshed really well together.”
Their idea of turning their different trading styles into a proprietary trading group at Wells Fargo (formerly Wachovia) ended with the 2008 credit crisis. Farrin moved on to hedge fund SAC and Hathaway would continue to perfect his diversified systematic approach.
All along they planned to pool their talents; by 2012 they were ready and CTA Sandpiper Asset Management was born. “We knew that long term systematic trend-following works, but that there are periods where it [struggles] and it is very difficult to weather those periods,” Hathaway says. “So you have two choices: You can build more models and in the long run dilute long-term performance, or you could find a different methodology to go along with it, similar to an asset allocation model.”
Rather than taking the path of other CTAs by adding short-term or countertrend models to core trend following, they added Farrin’s discretionary global macro approach that can and does benefit from many of the same moves but is non-correlated (0.2) to the core trend-following approach. “As a system developer, It is very attractive to build a mean reversion model and throw it in your backtest and look at smoother return streams,” Hathaway says. “However, it dilutes the long-term performance of the strategy, [so] I paired with someone who wasn’t going to dilute my P&L.”
Farrin says, “Some of our best periods are when both the model and my strategy is in a levered bet.”
Sandpiper’s risk management acknowledges that both strategies can be in the same positions, which are sized accordingly. Risk at Sandpiper is split 50/50 between the two models, which are not meant to hedge each other but act as independent non-correlated systems. “Jon is not here to be short gamma while the system is long gamma,” Hathaway says. “Jon is here to make money and do so in a different way.”
Farrin can trade any of the 50 markets included in Hathaway’s systematic program, but concentrates (80%) his positions in interest rates and currencies.
While most of Farrin’s macro trades are of longer duration, about one fifth are tactical macro plays around economic announcements.
If you ask for proof their approach works, they will show their 2013 performance. A difficult year for systematic trend-following, Sandpiper returned 23.95% on the strength of Farrin’s discretionary approach. The approach did not harm overall returns in the more trend-friendly year of 2014 when Sandpiper returned 33.5%.
“In 2013 we were both short gold but I got in earlier and out earlier,” Farrin says. “By definition if we have something that is moving in a parabolic fashion then we are both in, I will probably be taking profit earlier.”
“Jon will have different time horizons, profit objectives, stops — they may be tighter, they may be longer or shorter — [but] while we can be in a similar trade for a day or two, as market prices and dynamics change, the system will remain the same and Jon will be able to achieve his profit objectives or be stopped out,” Hathaway says.
“Jon is both fundamental and technical; he forms an independent macro thesis of the world using views on FX, economic data, interest rate differentials and other anecdotal and fundamental research,” Hathaway says. “That may coincide with what I am doing or it may not, but even if it does for a day or two or several weeks, it provides us with an advantage because we are not trading one set of parameters or beholden to one model.”
The two-model approach has worked through different market environments as Sandpiper has produced a compound annual return of 15.33% since August 2012 and is up 4.60% year-to-date through March.