Search

How to Trade with Moving Averages, Part II

Jason Pearce

Jason Pearce

Jason Pearce is a 25+ year veteran of the futures markets. Since 1991, he has played the roles of retail broker and managing director of a brokerage firm, trader, market analyst, newsletter writer/editor and trading systems/algorithms developer. Jason is now actively managing money as an independent RIA.
Jason Pearce

By Jason Pearce

Recap

In a prior post on the subject of moving averages, we talked about what they are, the benefits of trading with them, and some of the common techniques that traders use.  They are a simple, yet effective, tool for traders that have stood the test of time.

However, this does not mean that moving averages are going to be the Holy Grail of trading.  While moving averages can be highly-effective, so much so that they may even start to take on a reputation of infallibility during those times when the markets are trending, they quickly lose their luster once the trend ends and the inevitable choppy period returns.  When that happens, the profits accumulated while riding the trend can be quickly burned up.

When the trend starts eroding and moving averages start spewing out false signals, a trader has to concentrate on playing really good defense.  That’s done by making sure you employ sound risk management, rely on diversification (both in terms of trading systems and the instruments being traded), and in some cases making adjustments in the trading parameters.

The first two points are non-negotiable.  You can’t be successful without them.  It’s the last point that is the gray area.  Adjusting parameters will work for some, but not for others.  That means you’ll have to test it for yourself and see if it makes sense to you.

In this post, I’m going to share with you a very simple solution for adjusting trade parameters.  This is an approach that has allowed me to move quickly and methodically to recalibrate after a false moving average trend reversal signal.  In other words, if I get knocked off the horse and it’s still running, I’ve got a way to get right back in the saddle.

Turn On the Radio

A useful analogy for this method is that the market is like a radio station and a moving average is your dial to tune it in.  The failed signals that occur when you’re using a moving average are like static; you sometimes need to adjust the dial when this starts to happen.

Suppose a market is trending higher and experiences the normal pullbacks along the way.  If the uptrend is still valid, the pullbacks should all be contained by a moving average that is trending higher with the market.  There’s no need to adjust the dial because Dr. Johnny Fever’s voice at WKRP in Cincinnati is still coming through loud and clear on your radio.

Eventually, though, the market will experience a pullback that corrects far enough to finally break the moving average.  Something has changed and this break is a sell signal.  The trader will now liquidate long positions and maybe even go short because it indicates that the trend has changed.

So what happens if the market recovers after the correction and goes on to score new highs for the move?  Well, one certainly can’t argue with the market and expect to win.  The only logical conclusion is that the recent break of the moving average was a false signal.

Something has still changed, though.  The failed moving average signal tells us that the countertrend volatility in the trend has changed, but the direction of the trend has not changed.  Therefore, you either need to be long again or at least flat.  There’s no justifiable reason for a trend follower to have any short positions on at this point.  Unlike a DJ on the radio, the market does not take requests.

When this sort of failed moving average signal happens and tells us that the countertrend volatility has changed, a trader has two choices: she can take the next signal on the very same moving average or else she can try a different moving average in order to adapt to the new volatility.

Your Options

Despite a failed signal, there’s certainly nothing wrong with using the exact same moving average for the next signal…and the one after that…and the one after that.  This is what a 100% systematic trader does.

Many professionals would advocate this systematic route as the market will eventually break the same moving average and not recover.  Then you’ll finally have the real trend change you were looking for.  It’s just a matter of how many failed signals you might get first before a successful signal finally materializes.

Nothing is certain in trading, except for this: you will have a string of failed signals from any and all trading systems, whether or not it’s based on moving averages.  Hence, the reason for proper money management and diversification.  Drawdowns are a fact of trading and you don’t want to risk depleting all of your capital during the whipsaw periods.

After a failed moving average signal, there is another option: you can use a different moving average to trigger the next trade signal.

Let me warn you right now that if you opt for the second choice, you could potentially be crossing a line over into discretionary trading territory and getting out of the purely systematic one.  However, just because you are acting like a discretionary trader doesn’t mean that you can just fly by the seat of your pants and that you don’t need any trading principles to guide you.  In my experience, the successful discretionary trader usually has just as many trading rules and principles to guide them as a systematic one does.

Moving the Dial

Whenever I take the second route and decide to use a different moving average after a signal failure, I immediately look for a slower moving average –meaning a longer moving average- that was not violated during the market break.  That might mean switching from a 20-day moving average to a 30-day moving average or from a 30-day moving average to a 50-day moving average.

The objective here is to get rid of the recent “radio static” by locating a smoother moving average that has not yet failed during the market’s trend.  This is done in order to accommodate the markets change in the countertrend volatility that occurred when larger corrective moves started to happen.

This begs the question: why not use this slower moving average to begin with?

The reason is because a slower moving average lags further behind the market.  This means that it will generate sell signals further away from the market tops and buy signals further away from the market bottoms.  And we all know that every trader endeavors to sell as close as possible to the top and buy as close as possible to the bottom.

Dealing with Trouble

If a trending market is making very shallow pullbacks, the trend is being contained by a faster moving average.  No problems here.  It makes perfect sense to use this faster moving average that is running close to the market’s current price.

It’s when those pullbacks start to become sizable that the trouble begins.

The faster moving average will start getting picked off left and right with one false break after another, causing a string of losses, only to see the market turn right back around and continue the trend.

In an attempt to avoid racking up even more false signals and a continued losing streak, a slower moving average can be used.  But the slower moving average is further away from the market’s current price, so you will most likely be selling at a lower price and buying at a higher price than you would with the faster moving average.

The trade-off here is one of speed vs. accuracy.

The faster moving average gets you in and out quicker, but generates more false signals.  The slower moving average generates less false signals, but it gets you in and out of the market much later in the game.

What we are attempting to accomplish here with the use of parameter adjustments is to have the best of both worlds by using a faster moving average when it works and then switching over to a slower moving average when the fast one start to experience turbulence.

This certainly does not mean that the slower moving average won’t eventually fail at some point own the road.  It will.  It just means that, thus far, its track record for the recent trend has not yet been blemished.  So the trader who switches to a slower moving average for the next trade signal will be risking capital on a parameter that has not yet experienced a loss during the current run.

Who Wants Pizza?

Let’s take a look at an example of how a trader might have used this method of “tuning in the radio” to trade Domino’s Pizza (DPZ) stock since last spring.

To keep things simple, the trades will be on the long side only.  No shorting!  Also, we won’t worry about the impact of commissions, slippage, position-sizing, etc.  End-of-day prices will be used to determine the hypothetical profits and losses for each trade.

Initially, the 20-day moving average will be our weapon of choice.  A two-day close above the 20-day MA is a buy signal and a two-day close below the 20-day MA is a liquidation (sell) signal.  For this example, trading will commence on May 1, 2016.

After gapping down to multi-month lows in late April of last year, DPZ finally bottomed out at $116.91 on May 4, 2016.  Three weeks later, the stock made a two-day close above the 20-day MA for the first time in over a month and triggered a buy signal at $121.29 on May 25th.  Time to place our online order and buy some pizza stock.  You want that with pepperoni and sausage?

On May 31st, Domino’s closed back under the 20-day MA.  The next day it manages to hang on by the skin of its teeth and close back above the 20-day MA by a mere two cents.  The long position stays intact.  Boy, was that a close call.

On June 27th, Domino’s completes a two-day close below the 20-day MA and triggers a sell signal at $122.08.  The result of this first trade is a profit of 79 cents-per-share.  Not much, but it beats a loss or a poke in the eye with a sharp stick.

Just three days later, DPZ triggers a new buy signal when it gaps up to a two-month high and makes a two-day close above the 20-day MA.  The entry price is $131.38.

The next sell signal finally comes on August 16th when Domino’s closes below the 20-day MA for two days in a row.  The exit price of $143.97 results in a profit of $12.59-per-share.  Now we’re cooking with gas!

However, a new buy signal is triggered just three days later at $147.28 and new highs for the move are reached one day after that.  It’s a good thing we got back in.

But wait just a minute.  Since the last two sell signals were quickly followed by new highs for the run, it appears that the 20-day MA is not doing all that great of a great job, after all.  Although it’s getting us in the trade, it has bumped us out too early.  We have to keep getting right back in the trade.  That doesn’t even take into account the “near miss” we had on June 1st.

To move the dial and slow things down a little bit, let’s increase the moving average parameters by 50% and see how the 30-day moving average has been performing on this run.

As it turns out, a buy signal via the 30-day MA was triggered at $124.82 on June 7th.  Although DPZ closed back under the 30-day MA on June 27th, it was a one-day event and not a two-day event like the 20-day MA experienced.  Therefore, the trade would have stayed intact.

Furthermore, the stock stayed above the 30-day MA during the mid-August dip as well.  So it would have been a winning trade for over two months now.  Based on this, it makes sense to go ahead and make the adjustment and now use the 30-day MA for our new trade parameters after getting long again at $147.28.

On September 9th, DPZ dropped to a three-week low and close below the 30-day MA.  Fortunately, it recovered the next day and a sell signal was avoided.

On October 5th, Domino’s had a three-day streak of dipping below the 30-day MA and then closing just pennies above it.  Our luck finally ran out and a sell signal was triggered on October 14th when a two-day close below the 30-day MA prompted an exit at $151.12.  This trade made a profit of $3.84-per-share.

Two days later, DPZ exploded higher and posted a new all-time high.  I guess this proves that Domino’s really does have fast delivery!

This fast turnaround indicates that using an exit signal by way of the 30-day MA break was premature.  Plus, it had some close calls back in early September and early October where some sell signals were narrowly missed.

To remedy this situation, we will slow things down even more with the moving average.  To do so, we will increase the moving average parameters by about two-thirds and examine the performance of the 50-day moving average during this multi-month run higher.

Initially, DPZ made a two-day close above the 50-day MA on June 17th and triggered a buy at $127.79.  It then made a two-day close below the 50-day MA and triggered an exit at $122.08 on June 27th.  The trade resulted in a loss of $5.71-per-share.

The stock made a second two-day close above the 50-day MA on June 30th and triggered a buy at $131.38.  Since then, DPZ has not once closed below the 50-day MA.  The pullbacks in September and October -which both established the lows for the month- ended just above the 50-day MA.

This indicated that, after initially stumbling out of the gate, the 50-day MA has found it’s footing on the second trade attempt and is currently tuned in to the market.  Therefore, after getting the new buy signal via the 30-day MA at $159.45 on October 18th, we will use the 50-day MA for any new trade signals going forward.

In mid-November, DPZ experienced a sizable three-day selloff that pushed it below the 50-day MA.  However, the stock never went below the 50-day MA on a closing-basis.  It recovered and went on to make new highs.  Looks like we’ve made the right decision so far.

Domino’s sold off again in December.  This time it did make a two-day close below the 50-day MA and triggered a sell signal.  The trade was liquidated on December 13th at $161.86, resulting in a profit of $2.41-per-share.

Nearly a month later, DPZ finally makes a two-day close back above the 50-day MA and triggers a buy signal on January 11th at $168.96.

The question now is “Do we use a two-day close below the 50-day MA as our exit signal again or do we adjust the parameters again and see what an even slower moving average would do?

If you look at the performance of the 75-day moving average (a 50% parameter increase from the 50-day MA) you will note that the two-day close below it on January 3rd was the first sell signal since the original buy signal was triggered at $132.59 back on July 1st.  Also, the steep correction in mid-November ended after the stock tagged the 75-day MA and started to recover.

In addition, the 50-day MA and the 75-day MA aren’t all that far apart at the moment.  Given the similar locations and the better recent accuracy of the 75-day MA, I would lean towards using the 75-day MA for new trade parameters.

Regardless of whether you pick the 50-day MA or the 75-day MA for your parameters on the new trade, the liquidation signal was triggered on April 10th at $173.75 when Domino’s made a two-day close below both the 50-day MA and the 75-day MA.  This last trade booked a profit of $4.79-per-share.

The Results Are In

Recall that the trade campaign on Domino’s stock was initially started ten and a half months ago with the 20-day moving average.  The objective of this strategy was to adapt to changes in the trend by finding and using the moving average that was not breached during the most recent countertrend moves.  This means that it contained the corrections in the overall trend and didn’t get suckered by a “Larry Bird head fake” that shook out the faster moving averages.

Had a trader “remained static” and just stuck with using the 20-day MA parameters for entries and exits the entire time, he would have been in and out of the market eight different times.  The results were four winning trades and four losing trades.  Not taking into account the commissions and position-sizing scheme, the net result from trading DPZ would have been a profit of $17.02-per-share.  That’s not too shabby.

However, the strategy of “tuning in the radio” would have only produced five trades.  Yet they were all winners and the net result was a profit of $24.42-per-share (not including commissions, slippage, etc.)

As you can see, the parameter adjustment strategy had one-third less trades than using the 20-day MA alone.  This would obviously have cut down on commissions and potential slippage.  But even without taking that into account, the net profit of the parameter adjustment strategy was still a significant 43% greater than that of trading with the static parameters of the 20-day MA.

This one trade example does not mean that a parameter adjustment strategy will always make money or even that it will beat the results of a static parameter strategy every time.  The point of the exercise was to get you thinking about the potential of having dynamic parameters that adjust to market conditions.

Even though it was mentioned earlier in the post that this sort of active management of the parameters was in the realm of discretionary trading, there’s no reason that it can’t fit neatly into the category of systematic trading.  A trader would simply need to determine the rules for when to adjust the moving average parameters and what parameters to shift to.  Once that’s programmed in, voilà, you have yourself a mechanical system!

Keep Fine-Tuning

Although the parameters for the trade signals are adjusted to progressively slower moving averages as the trend unfolds, I will still continue to track the faster moving average that I was originally using.  The objective is to eventually put the faster moving averages back in the driver’s seat.

If the market makes another correction or two that is once again contained by a faster moving average, it means that the countertrend volatility is calming down.  This allows a trader to keep “turning the radio dial” and adjusting the trade parameters back to the faster moving averages.  Perhaps this means that will be switching from a 75-day moving average to a 50-day moving average to increase the response time.  Then you might switch again from a 50-day moving average down to a 30-day or even 20-day moving average.

The end goal is to end up using the fastest moving average that is containing the trend without getting breached during the countertrend moves.  You have to be vigilant to keep dialing it in and making adjustment as the market shifts until you find the best-fit moving averages for the current market trend.

Do Your Homework

The various moving average parameters (20-day, 30-day, 50-day, 75-day, etc.) that were used in this post are not set in stone.  They are used to illustrate how the methodology works.  Go ahead and test all sorts of different parameters, from ten days to fifty days to one hundred days or even more.

Also, investigate the different types of moving averages.  Does a simple moving average meet your expectations?  If not, you can explore using this technique with exponential moving averages, displaced moving averages, and other types.

In addition, you should research how this approach performs on different timeframes.  Like any robust technical tool, this moving average parameter adjustment method works just as well on the weekly and monthly timeframes as it does on the daily and hourly charts.  You need to find the right moving average parameters, the correct types of moving averages, and the timeframes that best accommodate your own style of trading.

One Factor of Many

Moving averages have been around for eons and have helped many a trader accumulate a fortune.  They still work just as effectively today as they always have, even if the parameters have to be recalibrated from time to time.  The moving averages can serve as the core entry/exit triggers for a trading system or they could be used for a confirmation tool or some sort of filter.  There’s definitely some value to be found here by any trader.

It is important to remember, however, that the moving averages are just one of piece of the puzzle for successful trading.  By no means should they be considered the Holy Grail in trading.

To trade successfully, you need to set up the correct position-sizing matrix.  That’s created by good pyramiding rules and risk management.  Moving averages only tell you when to get in or out of a market, not how many shares, contracts, etc. to buy or sell.

What you trade is an important factor as well.  Choosing the markets that have historically worked well with your trading system certainly doesn’t guarantee future success.  But why in the world would you ever want to have exposure in trading a market or instrument that has not worked well with your system or methodology before?!

When you have selected the moving averages that best fit your trading style…

And you have created your position-sizing plan…

And you have decided upon a portfolio of which markets to trade…

Then you still have to maintain the right mindset in order to follow your plan.  Stay disciplined.  Do not stray from the Path of Righteousness.  If you can that, you’ve got a pretty good shot at trading success!


More Articles by Jason Pearce:

How to Trade with Moving Averages, Part I

Market Returns Do Not Equal Investment Returns with Leveraged ETFs

Is The Canadian Housing Market Bad for Canadian Banks?

2017: The Death Year for Stocks

Potential Bond Market Reversal Ahead

Related Topics