The Marathon Runner
When people talk about managed futures pioneers, they almost always mention Bill Dunn. He's been running his models for almost 35 years. In that sense, he's a long-term survivor. He is also a long-term trend follower. Here he discusses the strategy from the perspective of someone who's worked on it for several decades.
Dunn Capital Management now has a number of trading programs, including one that is very short term—a pattern recognition system called Mosaic. If you look at the composite return that goes back to the 1970s, you're struck by three things. The first is how well Dunn has performed over what must be one of the longest track records in CTA and hedge fund history.
Dunn's compound annual return is more than 19%. The S&P 500 total return for the same period is 11%. Had you invested $1,000 with Mr. Dunn as he started his firm in 1974 and stayed with him, today your investment would be $426 thousand. By comparison, $1,000 in the S&P 500 compounded to $37 thousand during that time.
The second noticeable feature is that there is almost no correlation between Dunn's returns and the stock index and what little correlation that exists is negative. Last year, as stock markets crashed, a Dunn pool that allocates to seven trading programs returned 127%.
The third notable attribute is that the Dunn programs have steep drawdowns along the way. True, at other times the S&P 500 has drawdowns of similar magnitude. But the thought occurs that it may be tough for clients to stay put when they confront the drawdowns, even though in retrospect they'd be better off not redeeming. The manager's take on this issue is below.
Mr. Dunn has a PhD in theoretical physics and early in his career held academic positions and developed and tested systems for the US military.
Opalesque Futures Intelligence: Has your strategy changed over the years?
Bill Dunn: Our strategy has not changed since we started in 1974. We've developed a number of trading systems and programs, but the basic strategy of when to buy and sell and how much to trade has not changed.
OFI: But haven't there been major developments in 35 years?
BD: There are more markets to trade now. Back then there were only about a dozen tradable futures, now there are over 50. In 1974 we did not have currency, interest rate or stock index futures, hard as it is to believe. Those markets coming into existence is all for the better. As markets were created, we've worked on how to structure the portfolio, so that has changed.
OFI: How do you know that the strategy will work in a new market?
BD: I never know the future, I can only know the past. But I test the past. If the new markets behave like past markets, then there's no reason to think you can't trade the same way. We're always happy to see new markets. The more the merrier.
OFI: Please describe your basic strategy.
BD: The strategy is long-term trend following. Typically the winning trades last several months. Losing trades are often closed out in a few weeks.
OFI: Do you change the models when markets change, as they did last year, or when there are new rules and regulations?
BD: Change is a constant in markets. We do not tinker with the model when the environment changes. We're always looking for change, to take advantage of it. There was a lot of change in the past year, so that meant lots of opportunity. The impact of any rule change will show up in the way prices behave and the models will pick that up.
OFI: What have you learnt from your experience?
BD: We've got better at risk management. We've always tried to control risk but over the years we've become better at it.
OFI: Do you interfere with the models when unexpected events occur?
BD: My opinion or for that matter anyone's opinion as to what will happen tomorrow does not count in our trading programs. Only the tested models matter. We can modify the models, but they're back tested and have to prove themselves before we'll implement them. We don't tell the system what to do; the market tells us what to do every day using only price data.
OFI: What's a good environment for your strategy?
BD: We make money when there is a reasonable number of trends – whether prices go up or down – that last for several months. By contrast, markets that just bounce around give us difficulty and losses. In 2008 there were large upward trends in the first half of the year, for instance oil and grains, and strong downtrends in the second half of the year. It was the best performance year in our entire history. I'm sorry we all have to pay the price for those trends, but they were great to trade.
OFI: What happens when there are no trends?
BD: If there are very few trends long-term trend following does not make money. But we never know how long that situation will last and when trends will develop. None of these situations are predictable. I can't predict what will happen and I don't think anyone else can either.
OFI: When do your models have difficulties?
BD: When markets are volatile but don't go anywhere in particular, many systems are fooled into thinking there's a breakout of a trend when really it is just a breakup! Trendless choppy markets are bad for us. But you'll have problems with any system some time or another. There's no perfect system.
OFI: Aren't trend following returns too volatile for many investors?
BD: What keeps people away from managed futures is principally ignorance. True, managed futures returns are volatile, but as we recently saw, so are other markets. Our returns have a strong upward trend over the years, but there are wriggles along the way. People who can't accept a drawdown should wait for the next high and then take their money out! But most of our clients ride out the long term and get very attractive returns.