When investors buy futures, they tend to gravitate toward either discretionary or systematic managers. Which type is better?
The Alternative Investment Management Association debated the question yesterday in Toronto with the help of Scot Billington, co-founder of Covenant Capital Management (a systematic manager) and Frank Maeba, managing partner of Breton Hill Capital Management (a hybrid discretionary-systematic manager).
Billington used some non-investment examples to argue the benefits of systematic models. For instance, Billy Beane, immortalized in the movie Moneyball, used mathematical formulae to build winning baseball teams.
And in the 1990s, economist Orley Ashenfelter created a formula for determining the quality of red wine in France. Based on rainfall and temperature, it correctly predicted prices 90% of the time.
Read: How to outperform
Billington favours a systematic model because:
- he knows he can’t predict prices;
- he can test hypotheses and eliminate what won’t work; and
- it removes emotion from decision-making.
He jokes clients should fire him if he ever tries to base his decisions on opinion rather than his model’s output, adding it can’t be swayed by optimism or pessimism.
“I want to make the same decision October 3, 2016 as I would have July 9, 2002,” he says.
As for the critique that systematic managers lack common sense, Billington says using models allows him to mathematically formulate his intuition, experience and past successes.
“If there’s a volatility I want to buy, why not codify it, test it and apply it every time?” he asks. “If there’s a market anomaly that’s exploitable and persistent, why not codify it?”
The argument for discretionary
Maeba counters too much automation can be dangerous. He once heard a French quant manager compare his systematic model to an airplane. The 100 PhDs on his team were the fuel and the sophisticated computers were the cockpit.
“The airplane could fly itself,” the manager said. “But,” Maeba responded, “I wouldn’t get on that plane unless there was a human in the cockpit.”
Maeba adds, “humans will still be more responsive [than models] to structural changes in the markets,” and that discretionary managers have broader toolkits.
For instance, in 2008, Maeba saw CDS spreads widening to 1,800, and stopped trading them as part of his fiduciary duty to clients. A model may not have accounted for that responsibility, he argues.
He acknowledges some people think discretionary managers lack discipline, but says it’s possible to blend judgment with discipline.
Maeba gives the example of Green Mountain Coffee Roasters, a stock many people have shorted because they believe it’s overpriced. His fundamental analysis bore out that assessment, but since Green Mountain has been rallying, a lot of people have lost money. So he’s followed the momentum upward, explaining it’s possible to hold an opposite conviction while following a trend.
He adds that there’s always going to be discretion involved in setting the rules for any systematic model.
“Rules imply the rulemaker is an expert,” Maeba says.