We've been getting some great requests for column topics, so keep them coming. Today we'd like to address an idea that actually came from several of you - just in different e-mails (and different ways of saying it). The root of the question is just this.....how do you handle conflicting signals if you're committed to trading systems? It's a great question, so today we'd like to answer it with a quick lesson, and maybe a dose of 'reality'.
First, let's get everybody on board with what a trading system is. As it sounds, a trading system is just a mechanized way of generating buy/sell alerts. They're typically based on specific chart events, but there are also other ways of producing these unbiased signals. For this example, though, let's just assume that the trading system is based on a chart.
The only other thing you need to know - whether you're a novice or a pro - is that there are basically two kinds of mechanical trading signals. (This is possibly the most important thing to accept about trading, yet it's the first aspect that's forgotten.) The first kind of signal is one that assumes a trend will continue. Let's call those momentum signals. The other kind assumes that a trend will reverse. We'll just call those reversal signals. Here's why it's important to never forget this little duality........if one type of signal is right, then by default, the other type of signal is wrong. With that in mind, let's look at an example of each. After that, we'll illustrate how trying to apply both at the same time can be maddening, as well as unprofitable.
A MOMENTUM-BASED SYSTEM
Let's run through an example of a momentum-based system first.....a 'long only' system, just to make it easy. For our charts today, we're using TradeStation, but almost all charting software packages have a similar system-building function. The most common of all of these systems is a MACD (Moving Average Convergence Divergence) crossover system. Basically, these signals assume that once momentum has been developed, it will stay in place. The buy signal is generated on our chart below whenever the faster MACD line crosses over the slower one (we can't get into the mechanics of these lines today). These entries are indicated on the chart of the S&P 500 with upward pointing blue arrows. For our exit, we've just applied a trailing stop, indicated by downward red arrows. A trailing stop just means we'll make an exit when the index pulls too far back off its most recent high, but that stop level is continually adjusted to protect profits. Take a look at the chart, and then take a look at the summary of the systems success below that. (By the way, profitable trades are marked by a green line, while unprofitable trades are marked with a red line).
S&P 500 with MACD system trades (long only) - Daily
Click here to open chart in a new window.
Does the system work? On a hypothetical account that buys 10 option contracts per trade, the TradeStation software says it does work. Take a look at the system profitability report. Even though there were more losing trades than winners, the winning trades were almost three times as big as the losing ones were. The average trade gain here (including winners and losers) was $99.59 (an arbitrary amount - we just need to know if the system works, which it would for any account size.) Over the last two years, this system would have netted us $1693 in profits, if we were trading options. In other words, given enough time, this system is one generally worth using. (continued below)
System test results for MACD signals
Click here to open chart in a new window.
A REVERSAL-BASED SYSTEM
OK, so what about a reversal-based system? Let's build one using the stochastics oscillator - one of our favorites. Like above, let's just use long entries to keep things simple. The buy signals are marked with blue arrows, and the exits are marked with red arrows. Profitable trades are traced with green lines, while unprofitable trades are marked with red lines.
S&P 500 with stochastic system trades (long only) - Daily
Click here to open chart in a new window.
The bottom-line results are similar to those we saw with the MACD system. Using a reversal-based system, you would have netted gains of $1079 over the last two years with these signals.
System test results for MACD signals
Click here to open chart in a new window.
BUT WHAT IF YOU USED BOTH?
It's tempting to use both systems, isn't it? After all, even if there is some overlap in the signals, it couldn't hurt. And where there's not overlap, you'll have an opportunity for more gains than you would with just one system, right? Unfortunately, that's not true. The least profitable results were achieved when both of these systems were applied. The trade count more than doubled, but net profits actually went down! And more than that, this would have been a major burden to trade. Just look at the chart where both systems were applied - it's a mess.
S&P 500 with stochastic AND MACD system trades (long only) - Daily
Click here to open chart in a new window.
And the actual dollars made? Only a profit of $851 was achieved when both of these systems were used simultaneously - the worst results yet. Even when the entries were 'unlinked' with their original exits, the results were about as poor.
System test results for MACD and stochastic signals used together
Click here to open chart in a new window.
The point is just this.....besides being chaotic, following too many signals can be counter-productive (as we saw in our last multi-system results). That's not to say that we only follow one kind of system; we actually have several kinds of systems running at any given time. But any trader has to recognize how some mechanized signals can conflict with others. If you've got an oscillator and a momentum signal both 'on' at the same time, and you're waiting for both of them to agree, well, you've got a long wait. Over the long-haul, you're better served by using a system that works for you, and recognizing that there will be periods when that system fails. If you're an advanced trader, and can mentally handle both types of signals at the same time, we'd recommend that you keep separate charts for both systems. That way, one won't interfere with the other. In fact, some traders switch back and forth when the market environment changes.....which can be tricky to do. More than anything else, keep in mind that all systems will have periods of failure. If you can stick with them through those periods, you'll be far better off.
Price Headley is the founder and chief analyst of BigTrends.com.