After graduating from college in 1985, Jeff Mills took stock of his financial situation and future. The way he saw it, things were pretty simple. “I didn’t have an inheritance, I couldn’t count on winning the lottery and I didn’t have brilliant business idea at the time,” the 36-year-old Mills recalls. “I’d studied economics and computer science in college, so trading seemed like a natural fit.”
He applied his talents to researching the existing trading literature and writing his own programs to test the ideas he came across (a long-time friend of hedge fund manager Mark Boucher, Mills helped test some of Boucher’s ideas).
The first book he read was Edward’s and Magee’s Technical Analysis of Stock Trends. Although he found valid ideas, he was uncomfortable with the subjective aspect of the analysis and wanted things to be more mechanical and concrete. He was less enthusiastic about other ideas he came across, though.
|“I would read books by so-called experts and test their ideas and see whether they worked or not. About 90% of the time, they didn’t.”|
“I would read books by so-called experts and test their ideas and see whether these things worked or not,” he says. “I found about 90% of the time their ideas didn’t work; during certain market conditions they totally break down. What you can do with about 10% of their work, though, is add some kind of longer-term or more intermediate term indicator to bolster their approaches, and you might actually get something that could make you money.”
Thus began a trading journey that continues to this day. But he didn’t rush into anything. “I’m slow to commit money to a new idea,” he explains. “When I started out, I paper traded for a year before hitting the real markets.” He began trading in 1987, and only became a full-time trader in 1992.
Mills has developed well over 100 trading systems (he currently monitors 120 to 130 of them every day), and he constantly researches and tests new ideas. Many of his ideas about trading are reflected in the software he developed, Investigator, the relative strength search function of which is included in TradingMarkets.com. He trades his own account exclusively.
Jeff, who lives in northern California, has an easygoing manner, and he approaches trading very carefully and systematically. He told me about some of the key elements of successful trading approaches he has discovered over the course of his career.
ME: You’ve been testing and trading for quite a while, and you’ve been exposed to a variety of approaches. What are the kinds of ideas you’ve come across that work?
JM: Most of the systems I test on a daily basis have three components: a short-term component, intermediate-term component, and long-term component. The short-term component typically involves some kind of bar pattern, either a standard bar pattern or point-and-figure pattern. I haven’t found many candlestick patterns useful.
|“When you’re trading equities, volume is a godsend.”|
An example of an intermediate-term indicator might be a stochastic low. Also, when you’re trading equities, volume is a godsend: You can see the money flowing in and out, whether average volume is increasing, whether a breakout day from a trading range occurs on huge volume, if on-balance volume is increasing, or if there are divergences.
Finally, I always use a long-term indicator. I have one called the Gann trend indicator (GTI) that mechanically defines the long-term trend, whether it’s up, down or none. (Note: The GTI essentially measures the distance between reversal points defined by consecutive up or down closes; by adding or subtracting a percentage of these moves from the relative high or low points, you can also define stop-loss levels.) Or, I’ll use relative strength–looking for very high relative strength values for uptrends, etc.
One thing I’ve found particularly useful is “delta relative strength,” which basically tells you which stocks or futures are moving up or down the relative strength lists over time. It’s kind of momentum of momentum. The nice thing is that it identifies markets most people wouldn’t be aware of–those in the middle of the group that are up-and-coming. I always have at least one of my long-term indicators going in my favor when I make any trade.
I’m definitely a trend trader. There are systems out their for sideways markets, but I just don’t buy into them.
ME: When you talk about the long-term trend component, do you generally have a longer-term position on in conjunction with that trend and then also trade shorter-term positions, or do you just trade short-term positions in the direction of that longer-term trend?
JM: I do both, actually. Say the long-term trend is up and the market has a high delta relative strength, then I’ll definitely be trading in the long direction. What I’m looking for are mini pullbacks against the main trend to jump in.
|“I think your exit strategy is more important than your entry strategy.”|
An example is what is usually called a key island reversal day. I’d look for the stock to be trending up, and on the low of the island day, I’d want stochastics below 30. So, I’m combining short-term pattern, the intermediate term stochastic, and the longer-term trend.
ME: Do you actually find stochastics useful?
JM: It helps a little bit in these kinds of situations, but there are other indicators that are better overall. I have perhaps 120 to 130 systems I follow every day, and stochastics is in maybe 10 of them.
ME: What else do you trade?
JM: I trade breakouts from trading ranges. If there’s a long consolidation period, especially on weekly chart and volume has been building up, you can make a lot of money on those moves.
ME: Wouldn’t you just consider that a trend-following system? Do you have a money management or position management system that determines when you take profits or add on to position? When do you get out of that kind of trade?
JM: I think your exit strategy is more important than your entry strategy–by a long shot. I always have a stop on every trade, so before I make it, depending on the system and the exit technique associated with it, there is always defined point at which I’ll get out of the trade.
I often use the Gann indicators I mentioned to get me out of trades as well. It’s a mechanical way to define trend and it also acts as a great stop-loss technique. It allows the market to move in the direction of the trend but it also allows for the minor dips against the trend so you’re not stopped out. It works like a simple trailing stop-loss on steroids, if you will.
ME: Would it really determine the trend differently than, say, a breakout or moving average?
|“Finding trading systems is brutal–there aren’t that many out there that work.”|
JM: Very much so. First, it adjusts for volatility. Also, with a lot of trend-following systems you’ll enter long on Monday because the trend is up, and then on Wednesday you’ll find your trend indicator is telling you Monday is part of a downtrend because of the new data. The Gann trend indicator would still indicate Monday as being up if it was up before. It actually works better for short-term trading, while what I call the Revised Gann trend indicator is better for intermediate- to longer-term trading because it lets the market flow in your direction a little bit further. But it won’t move the stop up until the market has been able to fluctuate a little bit.
One thing you have to keep in mind with these trend indicators, though, is that when you’re testing a system, if you’re going to use it as an exit technique–maybe a long stop–if the trend is not “up” or “none” on your entry, you’ll get stopped out immediately.
Another important thing I look at is how the delta relative strength is behaving. For example, if the delta relative strength drops below a certain level and I’m long, I’ll get out.
ME: Are your techniques purely mechanical?
JM: Yes. I’m a 100% system trader. I’ve been trading these kinds of approaches since about 1987.
ME: What’s your trading day like
JM: The end-of-day stuff is all automated. What I spend almost all my time doing is research. I don’t want to lead people on. Finding systems is brutal–there aren’t that many out there that work. One thing I see in a lot of trading magazines are articles on systems tested on just the S&P, or one other market; statistically, that’s completely insignificant. You need a bare minimum of about 10,000 trades–entry and exit–to get an idea of how something works. The systems I trade work over all time periods; my price data goes back to 1968 on stocks, bonds, futures.
ME: What markets do you primarily trade?
JM: I mostly trade stocks and stock options. I’ve traded a little bit of futures–currencies act almost perfectly, on a technical basis. I like stock options because of their money management advantages–you know exactly how much you’re going to lose up front. It’s a good way to control risk.
|“I’m slow to commit money to a new idea. When I started out, I paper traded for a year before hitting the real markets.”|
ME: What kinds of option strategies do you use?
JM: I’m really simple. I try to buy options that are just out-of-the-money, about six month from expiration, and look for a 20% move in the stock over the next two months. Nothing fancy at all–but you have to be trading a system with huge upside potential. This probably works with about 20 of my systems; for the other systems, I’ll trade the underlying equity.
ME: How do you manage positions?
JM: One thing I do is if I have what I consider a large option position, I’ll cash out half when I’m up 100% just to make sure I’ll at least break even on the trade. I also use simple trailing stop losses in addition to the Gann trend indicators I mentioned before.
Trading is all money management. If you get nickeled and dimed a lot, it’s no big deal. It’s when you take big hits that you get in trouble. Most systems are only going to be reliable about 60% of the time. Getting little dings is just part of trading–it’s actually a good thing to get out even though the market sometimes rewards you for doing the wrong thing. I’ll often get stopped out, and sure enough, the market will take off in my direction I wanted it to go. It’s frustrating, but the good thing is that you still have money left to play the game.
ME: What’s your typical trade like?
JM: I trade maybe 10 times a week on average. A successful trade lasts typically anywhere from two weeks to two months. But again, this is selecting systems that are intentionally geared toward this time frame.
ME: Do you ever use fundamentals?
JM: Fundamental trading is actually a great way to go, I think. My problem with it is that first, there’s not enough fundamental data to do any serious historical testing on, and second, a company can look great and have great management, great products, great everything, and the price can still be flat or go down.
Also, how do you decide when to get out? When the product fails? When there’s a management shake-up? Once there are fundamental databases adequate to the task, I’ll be adding to my work. Ignoring fundamental factors is kind of dumb.
ME: What are the most important things aspiring traders should be aware of?
|“Becoming a successful trader is a long, slow ride. But if you’re persistent, you’ll win in the end.”|
JM: Money management is the most important thing in the world. Always have a stop, and always, always, always stick to your stop. Second, if you’re just getting into technical analysis is to buy tons of historical data–as much as you can. You want your technical system to work over all market conditions.
I learned the value of sticking to my stops early on–that’s one thing rookie traders trip up on. They suddenly think they’re Nietzsche’s Superman and they can predict the market–then they feel like they don’t need to honor their stops.
Don’t look for get-rich-quick schemes–it’s not going to happen. If you’re going to be a successful trader, it’s going to take a long time. What I’ve found with my systems is that you’re not going to find the stocks that are going to go up 1000% overnight. You’ll succeed with tried and true approaches: Try to capture the middle 60% of a move, and don’t try to predict tops or bottoms. It’s a long, slow ride. But if you’re persistent, which is another key thing, you’ll win in the end.
Be flexible. One interesting thing I found is that if you’re a system trader, testing things that don’t make sense sometimes makes sense. For example, one time I was giving a demo of my software and just throwing random things in there, and found a really successful system. It’s only happened once, but it shows you it pays to keep an open mind.
Finally, trade something you’re comfortable with. Anytime fear or greed gets in there, it’s just a recipe for disaster.
ME: What do you risk on an individual trade?
JM: Typically, I’ll start out with 5% of my total capital, find out what the statistical probability of the success of the trade, and then I’ll multiply the 5% times the probability, and that’s what I’ll risk on the trade. So, if the total equity is $100,000 and the probability of success is 60%, I’ll risk $3000 on that trade. That’s just the way I do it.
The idea is that you’re going to get dinged constantly, so you want to limit your exposure. If you have one great trade, it makes up for 10 bad ones.
The goal is to keep playing the game.