Which Trailing Stop Is The Best?

Trailing stops are a simple way to exit a trade. The idea is you buy a stock and then trail your stop loss a little way below it as the trade goes on.

This way you can let the trend continue in your favor but lock in your profits once the stock turns round.

The trailing stop is a very useful exit for traders that simplifies the process of letting winners run but keeping losses small.

The Key To A Good Trailing Stop

The key to a good trailing stop is that it needs to be loose enough that the stock has room to trend upwards. But it cannot be too loose or you will give back too much profit when the trend changes.

For example, in the Tesla chart below, you can see that the 5% trailing stop is too tight. It doesn’t allow the trend to develop and we take too many trades instead of following the trend:

Conversely, the 50% trailing stop below is too loose. We don’t capture enough of the trend and end up taking a loss when we could have had a big gain:

The best trailing stop strikes a balance between the two. It depends on the situation but the 20% trailing stop (below) often does a good job:

Which Trailing Stop Should You Use?

To find out which trailing stop is the best I’m going to test a selection of them on historical data going back 30 years across more than 11,000 US stocks.

We are going to use a new 252-day high as a buy signal and then see which trailing stop produces the most profit while limiting risk.

The trailing stops we are going to test are as follows:

• Percent trailing stop
• ATR trailing stop (Chandelier)
• Moving average trailing stop
• Parabolic SAR trailing stop

1. Percent Trailing Stop

This is the simplest trailing stop. Whenever the stock trades X% below it’s in-trade high then we will exit the stock on the next day open.

For example, if we buy Apple at $100 with a 20% trailing stop and it hits a high of $200 we will exit if the stock drops back to $160.

The following table shows the effectiveness of the percent trailing stop following a new 252-day high in the Russell 3000 from 7/1990 to 1/2020:

As you can see, the 20% and 25% trailing stop produced the best return-to-risk scores with reasonable win rate and profit per trade.

2. Chandelier Trailing Stop

The chandelier trailing stop uses the average true range indicator (ATR) to position the stop a certain number of points away from the action.

The advantage of this technique is that it takes into account the volatility of the stock and places the stop a certain multiplier away.

For example, if Apple is trading at $100 and we use a 5 x ATR(21) stop, the stop will be placed 5 times the ATR(21) below the recent high. If the ATR is $5 then the stop will be placed 25 points away (5 x 5).

In this test we are going to use the 21-period ATR and vary the multiplier to test the effectiveness of the stop:

You can see that the results for the Chandelier stop were pretty consistent when using a multiplier of 5 or more. However, the lowest multipliers saw poor results. ATR(21) and ATR(21) * 2 produced losses.

3. Moving Average Trailing Stop

The moving average trailing stop works like this. Once we enter the trade (a new 252-day high) we will follow it with a simple moving average line.

If the trend changes and the stock drops under the moving average line we will then exit the trade on the next day open. The following table shows our results across different moving average lengths:

The moving average trailing stops produced a reasonable return-to-risk score in the 40-60 day range. However, it was not as strong as the percentage stop and the win rate was lower too.

4. Parabolic SAR Trailing Stop

The parabolic SAR indicator rises according to specified parameters. But unlike the usual trailing stop, PSAR continues to move higher even as the stock stays where it is or declines.

This means there is an element of time involved so essentially, the stock is penalized for not continuing the trend upwards.

The PSAR indicator is made up of two parameters, acceleration factor and max acceleration. These are usually set up as 0.02 and 0.2, however, I found those parameters to be too fast.

The following table shows our results for various permutations:

The Parabolic SAR indicator produced some good return-to-risk scores particularly with small parameters (much smaller than most traders use).

Which Trailing Stop Works Best?

The results shown above provide some answers as to which trailing stop works best in stocks.

• The best trailing stop by return-to-risk was the 20% trailing stop with a score of 0.57. This was followed by the 25% trailing stop and the 15% trailing stop.
• The best trailing stop according to average profit per trade was the 50% trailing stop with an average profit of 82.72%.
• The 50% trailing stop also had the highest win rate at 53.3%. However, the 50% trailing stop naturally has a high drawdown and trade duration.
• The Chandelier trailing stop did not perform particularly well with low return-to-risk scores across the board.
• The moving average stop was not particularly effective either.
• The Parabolic SAR indicator put in some decent scores according to return-to-risk.
• Overall, the 15%, 20%, 25% and Parabolic SAR trailing stops appear to work the best.

Conclusions

In this article we looked at various types of trailing stops and tested them on 11,000 US stocks back to July 1990.

We found that the percentage trailing stop (particularly the 20% and 25%) does a decent job of capturing upward trends in stocks while limiting risk.

Meanwhile, the Chandelier stop and moving average line produced disappointing results and do not provide much reason to use these methods.

These findings support my previous experience and it was no surprise to me that the percentage trailing stops performed strongly.

If there is a surprise in these results, it is the decent scores for the Parabolic SAR indicator.

This trailing stop looks like it has some merit and I intend to test it with my existing trading strategies to see if it improves performance.

Notes

The data used for this analysis comes from Norgate Data and includes historical constituents and delisted stocks so as to minimize survivorship-bias. Data also includes dividends, is adjusted for splits and corporate actions and includes transaction costs of 0.1% per trade. Risk-to-return is defined as the annualised return divided by maximum drawdown. Stock charts created in Amibroker.

---