Your Trend Following Questions – Answered!
Last week, I shared the details of how a simple trend following system can generate truly market-beating gains with far less risk than “buy and hold”. Since then, I’ve been getting a slew of emails from readers wanting to know more about how trend following works — and what types of trading rules they can plug in themselves.
Today, I’d like to tackle some of your specific questions before we start digging deeper into maximizing profits with this strategy…
First off, if you haven’t had a chance to read last week’s article, I’d strongly recommend taking a look to get up to speed on trend following. You can find it here.
Picking Your Trading Timeframe
Now, onto the first question from Jeff:
“In you article, why did you choose the 300 day [moving average]? Could one use the 20 day, 50 day, or any other timeframe?”
Those are excellent questions.
When you’re looking at a technical indicator like a moving average, I think it’s useful to break down what the metric really tells you. As a reminder, a simple moving average is a stock’s closing price averaged over a set number of days.
So, in the case of our 300-day moving average, a computer is just adding up the last 300 days of prices and dividing the result by 300. Every day, the most recent day’s data gets added to the moving average and the last day’s data falls off.
Because of the way they’re calculated, moving averages “follow” stocks. So you can just think of them as a tool to approximate a stock’s trends; when a moving average is pointed higher, the stock must be moving higher too.
And since trend following works by finding big market trends and then riding them higher, those moving averages are one of the easiest ways we have to pick out where a trend is changing.
When figuring out a good moving average to use for trend following, I don’t care if it’s a 20, 50, or 163 day MA — those numbers themselves don’t mean anything to me. Instead of thinking about the numbers, think about what the numbers represent.
A 20-day moving average, for example, gives us a trend that approximates a month’s worth of trading. A 50-day MA is closer to a quarter. And the 300-day average that I used last week is just over one year (remember, we’re using trading days, not calendar days).
As a trend follower, you want the time period you choose to synch up with the cycles in the market. And frankly, the best way to figure out which to use is good old trial and error.
Let’s be clear: not all averages work well. You might want to stay away from the 20-day and the 50-day moving averages if you’re using the simple crossover system I introduced to you last week. The 20-day would have generated 4.5% returns since 1997 and the 50-day would have earned 19.2%.
Both fared a lot worse than buy and hold because they weren’t synched up with the market’s cycles.
I’ll show you how to pick more optimized time periods in a future column…
Moving Average Crossover
Reader Alan has another interesting question I want to cover today. He wants to know about buy and sell signals when a 20-day moving average crosses a 50-day moving average line:
“Are there any general guidelines in this scenario? It looks like it could generate pretty good buy and sell signals similar to your discussion about the 300 day moving average…”
Using two moving averages together has been a popular trading signal for decades — and yes, there are general guidelines about the scenario…
Basically, the shorter the time period of a moving average, the “faster” it is. For example, since there are fewer days of price data being factored in, a 20-day moving average is going to react to price moves more quickly than the 300-day. In a faster moving average, each day has a bigger impact on its value.
To spot trend movement, you want to buy when a faster average crosses above the slower average, and sell when the faster average crosses below the slower.
But this approach is far from perfect. The biggest problem is lag. Moving averages lag price by half their time period: so a 20-day moving average lags the market by 10 days, and a 300-day moving average lags the market by 150 days. For that reason, using two MAs (instead of price crossing an MA) can give you buy and sell signals too late — especially if one or both MA is out of synch with the market.
While moving average crossovers can add some interesting flavor for traders looking for confirmation signals (added bits of information that support their trading decisions), they’re usually a whole lot less useful for mechanical trend following systems.
I went ahead and ran a backtest for that standard 20-day/50-day moving average to see just how it would work out in the SPDR S&P 500 ETF (NYSE:SPY), and the results weren’t pretty. The profit chart is below:
The chart shows an investor’s hypothetical profit or loss using the system for the last 15 years. One glance shows you that this strategy spent almost as much time at a loss as it did at a profit — and it failed to perform even half as well as a lackluster buy and hold approach would have over that time period.
It looks more random than useful.
Trading Like a Professional
There’s an important message in the scenarios we looked at today: while I’ve shown you trend following can be a powerful investing strategy, it’s not an approach that can be applied blindly.
Even though we’re using fairly “technical” inputs to generate buy and sell signals here, it’s critical to think about what a given moving average is really showing you before you start using it to trade.
It’s also important to know that we’ve only looked at moving averages to build our buy and sell rules so far — there are plenty of other criteria we can introduce to our system to refine it and protect it from market risks.
We’ll look at some of those later on.
Jonas Elmerraji, CMT
Questions or comments? Drop us a line at firstname.lastname@example.org.
Start your free Tomorrow in Review email subscription...We Will Not Share Your Email Address
We Value Your Privacy