In his 32 years as CEO and founder of the futures research and trading firm Brandywine Asset Management, 26% of the markets traded with Mike Dever’s strategies have lost money.
“Most people would say, ‘I’m not going to trade those markets, they don’t work in that strategy,’” Dever says. “Well, that’s just wrong. But I’m fine with that. I don’t need people to agree with me.”
Dever says he’d actually prefer if people disagree with his trading methods—it’s what gives him an edge.
“Other traders aren’t rational players,” he says. “They’re just full of biases and that’s what creates opportunities. They don’t follow a disciplined, rational approach to trading, which for us means a systematic process.”
Brandywine’s systematic process is rooted in two core concepts of Dever’s: return drivers and predictive diversification.
A return driver is “the primary underlying condition that drives the price of a market,” as Dever put in his 2011 best-seller, Jackass Investing: Don’t do it. Profit from it. Sound, rational return drivers can be combined with relevant markets to create effective trading strategies. These trading strategies can be combined to create truly diversified portfolios—that’s where predictive diversification comes in.
Predictive diversification, the setup for Brandywine’s portfolio allocation model, is the concept of using past data to create a portfolio that will match past performance as much as possible, whether good or bad.
Dever says he first came up with this concept while working with researchers to develop the portfolio allocation model for Brandywine’s Benchmark program in the late 1980s. He had initially started trading futures in 1979 after developing a computerized trading program while studying at West Chester University. His first trade was in gold options and he had to finance it by selling his car.
The researchers were given strategies and market performances that Dever had backtested, and were expected to come back with hypothetical allocations. They came back with large allocations to a few strategy combinations and no allocation at all to others.
Dever asked one researcher how he came up with those numbers. The researcher said it was the optimal allocation.
“There was a flash in my mind at the time and I realized he was right,” Dever says. “All these guys had created the perfect answer to the wrong question. Everyone was trying to create the optimal portfolio, and what I realized in that instant was all I care about is creating a portfolio that’s past performance is likely to persist to the future. I don’t care if the performance is good or bad—if I don’t have some predictability, I have nothing to evaluate.”
This is why Dever continued trading the markets that lost money. To do otherwise would be to betray his systematic process.
In the Brandywine Symphony system, the Benchmark program that he updated in 2011, no one market or return driver ever dominates the portfolio’s performance. There are well over 100 markets and dozens of strategies in the portfolio, so any single trade based on any market/strategy combination will impact the portfolio by less than 1%.
Because of this, the Symphony model only profits with proper capital, he says. Brandywine’s minimum investment level is $5 million. “Anything less than that and you end up with a portfolio that’s not truly diversified and is subject to substantial risks,” Dever says.