In theory trailing stops provide a way for traders to limit losses and to lock in profits on individual trades. The basic idea of the trailing stop is that as a trade moves into profit, the stop level adjusts upwards in the case of a long (buy) trade or downwards in the case of a short trade.
In this way the downside is limited by the stop level but the upside is potentially unlimited. In other words trailing stops are a way to allow profits to run and losses to be limited.
In this article I describe how trailing stops work. I also test anecdotal evidence that trailing stops lower risk and result in higher profits. This is done by running back tests on two vanilla strategies both with and without trailing stops.
Trailing Stops – How they Work
There are several variations of the trailing stop used by forex traders. The most common are described here.
Standard trailing stop
With the standard trailing stop the trader sets an activation profit threshold in pips. Once the threshold is reached the trailing stop “kicks in”. The trading system places a stop loss just below (or above for a short) the current market price.
Unlike a regular stop loss the trailing stop will move as the price reaches new highs and the profit on the trade increases. With a buy-side position the trailing stop will only move upwards – increasing the profit. The reverse is true for a sell side position. The trailing stop will remain fixed if the price moves against the trade. The exit happens once the stop level is hit.
The diagram above illustrates a basic trailing stop system.
With the trailing stop the trader will also need to set a trail distance. The trail is the “distance” between the current market price and trailing stop exit point. So for example with a trail of 10 pips the trailing stop will float 10-pips below the market’s highest price since “kick-in”.
With some trading systems the trail point is allowed to move either dynamically or in fixed increments (step sizes either in time or price). With a dynamic trailing stop the placement of the stop can potentially move on every price tick. Whereas with an incremental trailing stop the level is only changed once the price changes by the pre-set step size.
Obviously one thing to consider is that dynamic trailing stops have a high overhead (on your trading software and on your broker who has to processes the rapid order throughput). With an incremental trailing stop the level is only changed once the price changes by the pre-set step size. With a time-based increment the level is changed only once per interval.
Trailing stop with cap
With a standard trailing stop the take profit level is usually left unset or at least set very wide so that the profits can run. This way the trade will usually exit when the stop loss is hit rather than when the take profit is hit.
For this reason some traders prefer to use a variation on this known as the capped trailing stop. With the “capped stop” the take profit level is also set dynamically. The cap is usually placed a small distance above (below for shorts) the market price. The cap is held fixed for a certain period of time (or price interval). If the cap or stop isn’t reached in that time interval the trade remains open.
Clearly both the cap and stop can’t be adjusted on every tick otherwise neither would ever be reached.
The benefit of this approach is that it can result in slightly higher profits. The reason being that there’s a certain probability that the take profit level will be reached. Recall that without the cap the trade will nearly always exit only through the stop loss which is set “beneath” the triggering price level.
Conditional trailing stops
When you use a trailing stop you “trade off” a portion of your profit in order to limit your losses. When you set a trailing stop there is always the chance that the market will not move further in the direction of your trade.
In this case the stop will be triggered at or below your current profit level. Clearly trailing stops work best when the market is moving in the direction of profit. For this reason some traders prefer to use what’s called a conditional trailing stop.
With this method the trailing stop only triggers when a certain condition is met. Otherwise the standard stop and take profit is kept. The “condition” is usually based on the market direction. Momentum based conditions are the most common. With these the trailing stop activates at a point in time when both the minimum profit level is reached and when the market is moving in the direction of profit.
A basic condition for example can be based on price change over a certain number of bars. For example, if the price moves at least +10 pips over 2 bars, the trailing stop will activate.
The reason this is used is that when the momentum is in the direction of the trade it’s more likely that the stop level will have a chance to move upwards within the next few bars. On the other hand if the momentum is moving against the trade there’s a higher chance that the trailing stop will be hit very quickly before having a chance to move upwards (downwards) and reach a better profit. In this case taking the profit immediately is usually the best course rather than applying the trailing stop.
Client side vs broker side stops
One thing to be aware of when using trailing stops is the difference between broker side and client side stops. With a broker side system the trailing stops are managed as part of the trade order. Once the order is placed and the trailing stop is set the broker’s trading system will monitor the trade and set the stop accordingly as the market moves.
By contrast client side trailing stops are executed by the trader’s software. Metatrader for example has a trailing stop function as do most other trading systems. Client side stops will only operate while the client terminal is open whereas broker side stops are set for the life of the trade. This means with client side stops you’ll need to keep your trading terminal open continuously.
Back Test Results
We back tested the trailing stop systems on two “vanilla” strategies. The first was a trend follower, and the second a breakout system. The tests were run over two durations namely short term (12 months) and long term (10 years). The maximum leverage used was 1:1 ($100k start balance) and the spread was set to 2.1 pips. We tested each on a single currency pair at a time (GBP/USD and EUR/USD) and used four variations:
- No trailing stop – using fixed take profit/stop loss points
- Standard trailing stop
- Trailing stop with profit cap
- Trailing stop with profit cap and momentum condition
In summary adding both the cap and the condition yielded better returns than using a regular trailing stop alone. This came at the expense of slightly higher drawdown. In the 12-month tests, the trailing stop with cap and condition performed better than using either a trailing stop or using no trailing stops. The strategy with “cap and condition” achieved a profit of $12,524 with drawdown of 2.71%.
With standard trailing stops the total profit was $11,281 with a drawdown of 2.63%. Using fixed ratio profit/stop losses achieved a profit of $12,191 with 2.67% drawdown.
Underperformance at longer Durations
Over longer timeframes both the trailing stop and modified trailing stop (with cap and condition) significantly underperformed the strategies using no trailing stops. Adding the trailing stop allows profits to run and this did result in a few big wins. The most profitable trade for the standard trailing stop was $372.02. The most profitable trade with cap and condition was $402.5.
However over time this benefit was negated by a lower per trade (average) profit. This is due to the trailing stop system having to “give up” a portion of profit in return for a limited loss. We tested multiple configurations and the trailing stop systems always resulted in lower profits over the longer time periods.
What’s also surprising is that adding the trailing stops didn’t reduce drawdown to any significant level. On the longer test the standard trailing stop reduced drawdown by just 0.18%. But this came at the expense of a 22% drop in profits. The trailing stop with cap and condition increased drawdown slightly by 0.03% but resulted in a 15% reduction in profits.
We also expected the trailing stop to work better with the trend follower as this is its natural territory. But the results were similar with both strategies.
Standard trailing stop test
With the standard trailing stop the trail point was set a certain distance below (or above for short) the current price level. The trail point was initiated as soon as the profit reached 0.3% – (averages 45 pips on GBP/USD or 35 pips on EUR/USD). The trail point was adjusted only once per interval (5-minute period). With the “non-trailing stop” strategy we used a fixed stop loss/take profit with these points being set at +/-0.9% and +/-0.3% respectively.
|Strategy||Trail distance||Total Profit||Avg. per trade||Trade count||Max drawdown|
|No trailing stop||#||$12,191||$28.29||431||2.67%|
Trailing stop with cap
This system added a profit cap above (below for shorts) the current price level. The trail level and cap was re-adjusted just once per time interval (5-minute bar). The trade exited if either the profit point or the trailing stop point was reached during the interval. This modifies the standard trailing stop system (above) where the trade will only exit when the stop loss is reached.
|Strategy||Trail distance||Profit distance||Total Profit||Avg. per trade||Trade count||Max drawdown|
|No trailing stop||#||#||$12,191||$28.29||431||2.67%|
|Trail with cap||20||5||$12,062||$27.99||431||2.78%|
|Trail with cap||20||10||$11,346||$26.32||431||2.77%|
Trailing stop with cap and condition
The trailing stop with cap (described above) was modified further to add a momentum parameter. The conditional trailing stop only activated if a momentum condition was reached. The condition was that the price distance between the previous two intervals (5-minute bars) was greater (or less for shorts) than the trigger level. For example with a 10 pip trigger the price needed to move at least 10 pips (up or down depending on trade side) between the previous two bars.
|Strategy||Trail distance||Profit distance||Trigger*||Total Profit||Avg. p/trade||Trade count||Max drawdown|
|No trailing stop||#||#||#||$12,191||$28.29||431||2.67%|
|Cap and condition||20||5||10||$12,524||$29.06||431||2.71%|
|Cap and condition||20||5||20||$12,485||$28.97||431||2.65%|
For the longer duration tests we ran the strategies over a ten year time frame. The results below summarize the highest achieved profits using the standard trailing stop and modified trailing stop system. The standard trailing stop resulted in a 22% fall in profits compared to using a fixed ratio stop loss/take profit system. The modified trailing stop approach (cap and condition) resulted in a 15% fall in profits over using the unmodified strategy.
On the longer tests neither the standard nor the modified trailing stop system reduced drawdown to any significant level. This was the accumulative effect of the running profits being lower compared to the vanilla strategy.
|Strategy||Total Profit||Avg. per trade||Trade count||Max drawdown|
|No trailing stop||$139,143||$22.76||6113||10.05%|
|Cap and condition||$121,195||$19.80||6120||10.08%|
|Standard trailing stop||$114,491||$18.74||6108||9.87%|
Are Trailing Stops Worth Using?
What we found in our tests was as follows:
- Trailing stops with caps outperform the standard trailing stop system. Adding the “condition” improved the performance of both systems.
- The performance gain of the two trailing stop variations comes at the expense of slightly higher drawdown. However given the large profit differential the trade-off seems worth it.
- In the short duration back tests (one year or less) using a trailing stop with cap outperformed a fixed ratio stop loss/take profit system.
- In longer duration tests all of the trailing stop systems significantly underperformed a fixed ratio stop loss/take profit system. The profit reductions we observed were between 15% but also up to as much as 30%.
- Over the longer duration (10 year tests) none of trailing stop systems reduced drawdown to any significant level.
The key point to note about trailing stops is that “locking in a minimum profit” always comes at the expense of a slight reduction in profit on each trade.
It is true that letting profits run, as trailing stops do, increases the odds of a few big wins. But over the long term the lower profit average on each trade becomes significant. This cumulative effect is most noticeable with high frequency trading strategies.
So are trailing stops worth using? Obviously there may be other strategies that work better with trailing stops. For example where the trailing stop results in trades being held open much longer and as such this could lower trading costs. This hasn’t been tested here. There may also be the case for using them for manual trades where discretionary input of the trader is available.