As emerging commodity trading advisors go, few have anything on Peter Matthews. Matthews, of course, is not new to the game, being a founding partner in legendary Mint, the first CTA to reach the $1 billion threshold and the original program behind the Man guaranteed fund empire.
But in 2002, Matthews, encouraged by his wife Jackie, had the itch to get back in the game. “She saw that there was a lot more that could be done, that Mint wasn’t the end and that I had the ability to create something new and better and she and I thought we should do it,” Matthews says. “That is what is exciting about it. I feel about it today the way I felt about Mint in 1980.”
Matthews, a researcher by nature, was not satisfied with knowing trend following worked. He needed to know why, so he began a research effort that would eventually lead to his latest program PJM, which stands for Peter and Jackie Matthews. After operating PJM as part of global macro hedge fund Caxton for two years, Matthews took it independent in 2008.
Mint was the classic trend following CTA, using what Matthews refers to as the statistical approach to trading. In his new effort he uses a more scientific approach.
“What I wanted to get at the second time around is, why did it work? What is it about markets that makes trend following work? If I can understand then I can take the next step, to create a better way of doing what I did at Mint,” Matthews says.
He deconstructed markets to the basic elements and found four characteristics: 1) markets are made up of heterogeneous agents (different traders trading different models with different knowledge and risk tolerances) 2) they use a feedback process (each new price can alter what a trader will do next) 3) they are based on limited resources and 4) they are self organized (there is no central authority setting price).
Matthews recognized from his knowledge of scientific research that these elements are a characteristic of a Complex Adaptive System (CAS), which he says means that many assumptions regarding markets need to be scrapped. “There would be no talk about efficient markets and random walks and normal distributions,” Matthews says. “In [a CAS], equilibrium is the rarity. You will have bubbles and crashes.”
The most interesting implication is that in a CAS, the future is entirely unpredictable. “It really means we can’t possibly know what is going to happen in the future,” Matthews says, adding, “It turns the world of investment management on its head because it says there is no such thing as investment.”
While he saw this as a major revelation, it gave him comfort in what he had been doing all along. “It told me that trend following would never be dead because the very nature of [CAS] is such that they will have large standard deviation moves. Anything can happen and you need to be on the right side of those moves to capture excess returns.”
His revelation confirmed that trends exist, but challenged the notion that systems could be developed to predict them. Controlling risk becomes the most important element of trading. “That is ultimately how I developed what I consider the next generation of quantitative systems, which is a risk management system rather than a trend-following system. How do I think about risk in the correct way to make sure that I am a survivor, that my clients are survivors,” he says.
“Trend following is a reasonable solution to the [CAS] problem but it is not the best solution. I was able to take what I have done at Mint and say that was good, it worked, clearly diversification is good, clearly using stop losses are good, clearly riding the winners is good. Not having a bias towards long or short is
But he realized his focus needed to shift. “How do I have a rule for getting in and out of markets that minimizes my risk when I am wrong? I built a model around this concept of picking points to go long or short where the cost of error would be the smallest.”
The model is in all 58 markets it trades all of the time. The size of the position is based on the risk. He will take a much larger position for a signal with lower risk as defined by his model. “We look at low risk as such a great thing that we should do a lot more of it and high risk as such a bad thing that we should do almost none of it.”
Matthews has the advantage of the experience and history of Mint. Comparing PJM’s performance against Mint’s over the same period showed the new program produced similar returns — a little better than a 20% plus per year average, but with much less risk.
“The pain that you feel along the way in getting those returns is much lower. It kind of solves that problem the CTA industry has always had. If I can cut down on that chop and get those returns with less pain that is a better mouse trap.”
The program began trading customer funds in August 2008 and is up 13.6% in its first 12 months. “I’ve got something that is the next generation that I know is different from what other people are doing,” he says.
Click here for Peter Matthews' article on the Ulcer Index.