Trend Following Systems Follow The Trend

trend following systems

trend following systemsIt has been about three months since we started following these six simple systems. Over that time, we have seen the systems jockey for position among themselves a bit, mostly based on how the general market impacted their long positions. The systems that were long did great on the up weeks and lousy on the down weeks. On the other hand, the mean reversion system seems to be building more consistent profits, although it is doing so at a much slower pace.

Can we take anything away from this? Not really. It’s only three months, so the sample size doesn’t really mean much of anything.

However, it is interesting to note that during uptrends, the long term trend following systems can mirror the returns of buy and hold investors. Of course, the major difference is that each of the trend following systems has some sort of mechanism to get them out of their long positions if the market begins to decline. Buy and hold investors have no such protection.

Comparing Our Systems

trend following systemsIt has been very interesting to watch how theses different systems have been separating themselves over the past few weeks. Since the market has been in a long-term uptrend for the entire duration of this experiment, it is no surprise that the long-term trend following systems that have been long for the entire experiment have been the most profitable.

It has also been very interesting to watch how the more complicated systems have repeatedly hurt themselves by getting out of their positions two quickly or getting caught in short positions on big up days. This leads us to wonder if perhaps those systems are “too cute,” but it could also be a product of this unique market environment. These results could be very different if they were tested on a different three month period.

Eventually, the market will change direction and we will get to see how these systems all perform in a different environment. For now, the long-term trend following systems are leading the pack. However, if we experience some sort of crash and our short term systems protect us, they could make up for all of the lagging they have been doing in a very short amount of time.

System 1 – Buy & Hold

Overall Return: +4.59%

2 Weeks Ago: +1.27%

trend following systemsThe buy and hold investors have had a field day over the past two weeks. The SPY only registered two down days in that time and the up days have been tremendous. Weeks like these allow buy and hold investors to get very comfortable with their positions. This sets them up to rationalize their losses later.

Like we say every week, the biggest flaw with System 1 is that it has no way to protect itself when a downtrend inevitably occurs. No one really knows how much longer this current uptrend will last, and when it eventually ends you will want to have some mechanism to protect your profits. It is also important that your system has some form of defense against the next black swan event, which will surely happen but no one knows when.

System 2 – IBD Market Calls

Overall Return: A: -1.77%, B: +2.19%, C: -8.37%, -4.10%

2 Weeks Ago: A: -3.27%, B: +0.62%, C: -8.12%, D: -3.98%

Here is a quick reminder about the different versions of System 2:

  • System 2A – Long on Confirmed Uptrend, Cash on Market In Correction
  • System 2B – Long on Confirmed Uptrend, Cash on Uptrend Under Pressure
  • System 2C – Long on Confirmed Uptrend, Short on Market In Correction
  • System 2D – Long on Confirmed Uptrend, Cash on Uptrend Under Pressure, Short on Market In Correction

trend following systemsSystem 2 had a very interesting couple of weeks. Two weeks ago, versions A and B were sitting in cash and versions C and D were short the SPY. Once again, we saw the short versions get caught in their short positions when the market rallied. Versions A and B simply missed out on some gains, but versions C and D got smacked by being short on those big rally days.

Since IBD called a Follow-Through Day on Wednesday, October 16, all four versions of System 2 went long the SPY on the next day. Since then, they have all benefitted from the markets continued upward progress. The means that versions A and B tacked on to their gains for the year, while versions C and D made up a portion of the losses they took from getting caught in their short positions.

Moving forward, a shift in market outlook to Uptrend Under Pressure would cause versions B and D to move into cash positions. With the amount of strength that the market has shown over the past two weeks, it is unlikely that this will happen, but me have to be prepared in case that it does.

System 3 – 10/100 Moving Average Crossover System

Overall Return: +4.59%

2 Weeks Ago: +1.27%

trend following systemsSystem 3 continued to mirror the performance of System 1 over the past two weeks. I was originally concerned with the fact that System 3 has not outperformed System 1 over the past few months, but that is not really what it is supposed to do.

As a trend following system, System 3 should trade in the direction of the long term trend. While it has come close a few different times over the past few months, the long term trend has always continued to be up. Because the long term trend has been up, it makes perfect sense that System 3 has mirrored the results of buy and hold investors.

The differentiating factor will be when the market eventually declines. At that point, System 3 will execute its mechanism to get out of the market while System 1 will continue to hold onto its long position.

System 4 – 89/13 Day Breakout System 

Overall Return: -5.16%

2 Weeks Ago: -5.98%

trend following systemsThe SPY closed at a new 89-day high on Thursday, October 18. This signaled System 4 to enter a long position at the close of the following day. This means that the system missed out on a good chunk of the move, but did capture the tail end of it from the past week. The system is also poised to benefit from any additional gains that the market might put in over the next few weeks.

The SPY is currently trading more than 10 points above its 13-day low, so it is not very likely that System 4 will get stopped out in the coming week. It is more likely that the system will continue to hold its position and take what the market decides to deliver.

System 5 – 3 Day High/Low Mean Reversion System

Overall Return: +1.22%

2 Weeks Ago: +1.22%

trend following systemsThe past two weeks have not been kind for mean reversion systems looking for oversold conditions. There have only been two small down days over that time, so we didn’t even come close to seeing three down days in a row. The price also has not closed below the 5-day moving average over the past two weeks. For those reasons, System 5 has simply been sitting in cash waiting for a new opportunity.

With the market running so fast for the past two weeks, it is entirely possible that it pulls back to catch its breath in the coming week. This could set up an opportunity for System 5 to make a quick profit if an oversold condition develops. In order for that to happen, we would need to see the price close below the 5-day moving average after three straight days of lower highs and lower lows.

System 6 – Simple Moving Average Systems

Overall Return: A: +4.59%, B: +4.59%, C: -1.86%, D: +4.59%, E: +4.59%, F: -7.99%

Last Return: A: +1.27%, B: +1.27%, C: -4.96%, D: +1.27%, E: +1.27%, F: -10.92%

Here is a quick reminder about the different versions of System 6:

  • System 6A – Long above 200 day SMA, Cash below 200 day SMA
  • System 6B – Long above 100 day SMA, Cash below 100 day SMA
  • System 6C – Long above 50 day SMA, Cash below 50 day SMA
  • System 6D – Long above 200 day SMA, Short below 200 day SMA
  • System 6E – Long above 100 day SMA, Short below 100 day SMA
  • System 6F – Long above 50 day SMA, Short below 50 day SMA

trend following systemsAll six versions of System 6 were long the SPY for the past two weeks, so each of them were able to capture the same gains as System 1 and System 3. Because versions A, B, D, and E trade off of longer-term signals, they have all continued to match the returns of System 1 and System 3 for the duration of this experiment. It will be interesting to see when and how these systems eventually begin to differentiate themselves.

The shorter-term versions C and F are lagging their longer-term counterparts. It appears as though trading based on the 50-day moving average is opening our capital up to a significant amount of whipsaw issues. This is even more obvious in version F because of its short component, which doubles the impact of those whipsaws.

In the coming week, it would take some pretty significant losses for any of these versions to generate any signals. While losses that significant are improbable, they are certainly not impossible.