Getting out when proved (possibly) wrong

Hi,

Has anyone experimented with using a wider stop but moving the stop closer/getting out at market if the trade isn't profitable within a certain number of bars?

banco said:

Hi,

Has anyone experimented with using a wider stop but moving the stop closer/getting out at market if the trade isn't profitable within a certain number of bars?

Click to expand...

Yes did a lot of backtesting on this topic back in the late 90's

My findings were.

Using the stop as a percentage of the initial purchase price (Eg 5% of $20 = $1).
(1) 3-6% was too close over 50% were stopped out. The trade off was a much higher R/R when you did get into a trade.
(2) 7-12% was the sweet spot here 30-50% were stopped this appeared to be the balance point from my testing.
(3) 12-25% had the lowest stop out rate but you ran the risk of getting trapped in a trade that wallowed in the stop zone.

My testing didnt nominate a time period it was over the life of the systems testing.
Wouldnt be hard to add a time filter.
Would dearly love to get more time for testing questions like these.

banco,

I do this on my support buying trades. I also will exit the trade even if the stop has not been hit. My testing has shown that if buying the type of support I buy, ie 'good support' (support in a rising trend) if a move does occur then it is usually quickly, as in a couple of days. The longer sideways action takes place then the higher the likelyhood of a negative move, with a high percentage of large moves.

However I will add that this is what works for what I am trading, and I would suspect that trading different types of setups would have different results. What you need to do is test your system for past signals and see what happens if you employ this type of stop. Make sure you test in both bull and bear market conditions to check for variance to existing results.