Fixed versus Trailing stops - experiences ?

[rog1111]

Hi all

I have only quite recently amended my fixed stop strategy in favour of a trailng stop (based on trade prices). For the chosen strategy, backtesting with 18 months data shows that the fixed stop gives better returns in terms of raw pips, albeit with greater drawdowns, although there is very little in it. The object of the exercise is to achieve a more consistent short term return & smoother equity curve.

Does anyone have any comments, theory or experiences to relate on fixed vs trailng stops ?

rog1111

 

[Roberto]

I haven't really done proper backtesting of this for my various strategies, and I imagine (maybe rightly, maybe wrongly) that there would be little overall difference. To me, the decisive factor is that I tend to find anything other than a trailing stop too stressful. I've always used a trailing stop, although the exact ways I've done so have changed quite a bit. At the moment, for some of my set-ups I close profitable positions in either two or three separate places, adjusting the stop-loss on the remaining part(s) even more often than that; for others, I trail the stop-loss but don't take any "early profits".

 

[sidinuk]

Psycologically a trailing stop is easier to trade as it locks in profit and prevents a situation where a good profit can turn into a loss with a fixed stop.

Overall though, it really depends on the particular system and what you are trying to achieve - I have systems that use trailing stops and others that use fixed.

 

[JonnyT]

I haven't found a system yet that benefits from a trailing stop.

JonnyT

 

[TWI]

I have been using a trailing stops since I remember. Way I work it is based on points draw from maxPnL.
I find this best option rather than a step approach that moves stop every time PnL increases by X or one that looks at % drawdown from max. Using number of ticks from max PnL is most generic.
Also find looking at PnL after x bars useful. e.g. if after x bars PnL < y then "do something".
Trailing stop is also useful when there is a big figure day such as non-farm. Often will really narrow up trail Friday morning and use it as stop reverse into the friday close. On the whole taking the break around a big figure will make you money into the close. I always close out the friday reversal before the close. If it happens to move the way I am already positioned then I will tend to leave it open.

 

[rog1111]

Thanks all for the replies.

I agree with JonnyT's observation, in terms of gross PnL. For the systems I have backtested, at least , stops nearly always diminish overall system performance. However I am prepared to forego quite a bit of annual profit in order to reduce my 'psychological drawdown', as pointed out by sidinuk.

twalker, simple observation (of GBP) backs up your idea of a time/profit constraint of some sort, I will look at it further and thanks for the idea.

Ideally I would have liked to use a sliding trailing stop ie. one that gradually gets a little closer as PnL improves, but I could find no easy way to implement this hands-off in TWS without using the API, so a simple trailing stop seemed like a reasonable way to achieve all the objectives. Only time will tell whether it makes money !

rog1111