LeBeau and Lucas, System Trading Commodities Traders
October 11, 2005
Futures Magazine
February 1996
Number 2, Page 78
TRADER FILE
LeBeau and Lucas: Truth in testing
Source: Futures Magazine, February 1996
Photograph by Michael Justice
By Jack Reerink
Chuck LeBeau, 57, and David Lucas, 50 tested every technical indicator under the sun during their seven-year partnership and came to a disturbing conclusion: Most are no better than flipping a coin.
"We tested how effective the market entry is to exit after n-days and see what percentage of the time we were on the correct side," Lucas says. "It’s amazing how few technical studies, pattern recognition or whatever you want to use are actually better than random. If you can get over 55% after five days, 50% after 10 days, you’re doing remarkably well."
Lucas got a bachelors degree from Central Connecticut State University in New Britain, Conn., and studied biological statistics at the State University of New York at Binghamton, N.Y. After safeguarding missile silos in North Dakota for the U.S. Air Force from 1966 to 1970, he started as a computer programmer at Signa Insurance in Hartford, Conn.
"After a while, unless you’re a real computer junkie, it gets old," says Lucas, who joined E.F. Hutton in Long Beach in 1984. "I wanted to do something that was a little more vibrant."
LeBeau got hooked on futures earlier. At California State University, he took an investment class taught by Charles Harlow, a second-generation commodities trader. Upon graduation in 1963, LeBeau was drafted in the U.S. Army. He worked his way up to captain, as commanding officer of the armed forces courier station at the U.S. embassy in Paris. After "tough duty in the trenches of Paris," LeBeau extended his tour to bring his wife over. He started with E.F. Hutton as a broker in 1967.
Lucas and LeBeau teamed up in 1988, publishing a monthly newsletter, writing a book and developing a trading system for their commodity trading advisor, Island View Financial Group in Torrance, Calif.
"At first I wrote a little program that would test 20,000 combinations of moving averages on different markets. We just ran the thing on an old 286 for weeks and weeks," Lucas says. "After months of trying to find the holy grail, we discovered the reality was if markets were trending, any moving average made money."
Instead they concentrated on determining the trend and gauging its acceleration. They found the best tool is directional moment (DI), the largest part of a trading day’s range outside the previous day’s range. The value will be +DI in a rising market and –DI in a bear market.
"If the upward moves outweigh the basket of downward moves by a certain amount, there’s a trend in the upward direction," LeBeau says. "If divergence between –DI and +DI reaches a certain level, in 70% of the cases the trend will continue in the present direction for 20 days."
LeBeau and Lucas enter a trade if the trend is accelerating, using the rate of change in Welles Wilder’s 14-day average directional movement index (ADX) measured over seven weeks. They attach little value to ADX levels.
"We found the levels are not nearly as meaningful as Wilder and we originally thought," LeBeau says. "The level merely tells you where you’ve been and has very little predictive value."
Lucas adds the ADX (on a scale from 0 to 100) spends most of the time between 10 and 40, with 20 or 25 being the average.
"Nowadays we’d rather be trading when the ADX is at a lower level, probably in the teens somewhere is optimum," he says. "Markets get more mature, so trends are often shorter and more violent."
After much research, LeBeau and Lucas built their system around directional movement only.
"It’s very easy to throw a whole lot of things in the hopper. But in reality you’re curve fitting and reducing the ability for the system to work in real time," Lucas says.
LeBeau and Lucas use a hard dollar stop, a trailing stop based on chart patterns and a volatility-based stop (over a single period). They calculate them each day and use the one nearest to the market at the close. In case of a big winner, a profit more than x-times the contract’s four-day average true range, they employ a profit-taking strategy.
"Where as the [volatility] stop is based on x-multiples of the four-day average true range, in this case it’s a fraction," LeBeau says. "It is to the point where we have to change our stop several times a day."
This enabled them to sell March Deutsche marks at 73.15 on March 8 (99 ticks from the contract’s high) for a profit of 571 ticks, or $7,100 per contract. They also bought back September yens at 102.82 on Aug. 16, for a 1,326-tick profit, or $16,580 per contract.
They have learned that – as in foreign affairs – in futures markets no action is often the best action.
"Unlike a lot of traders, we’ve come to appreciate being out of the markets," Lucas says. They were flat for a total of eight weeks in 1995 in their $8.7 million currency financial program, which trades five (IMM) currencies and three interest rates (up 48.6% through November).
Having found one indicator that beats a random strategy, LeBeau is searching for a fundamental beacon: The $400,000 fundamental program was down 9.04% in 1995. He got whipsawed in cocoa, missed the Canadian dollar’s rally and went with the wrong grain contract, soybean oil. LeBeau says: "Fundamental trading is difficult."
Copyright (c) 1996, Futures Magazine
Visit: TurtleTrader.com® Website
Visit: Bestselling "Trend Following" Book
Visit: "The Complete TurtleTrader" Book Review