Trend Gate or Filter

[nizar]

Hi everyone,

If I have a short-term system that works well in "trendy" markets and not in sideways markets where it will get whipsawed to death, how would I best define a "trendy" market?

I need to somehow develop some strict criteria (maybe based on price action for the last x periods??) that tells me if the market for a given day is likely to be a trendy market or not??

Hope that makes sense.

Any ideas would be much appreciated.

Nizar.

 

[Trembling Hand]

What time frame is your trend?

 

[julius]

Hi everyone,

If I have a short-term system that works well in "trendy" markets and not in sideways markets where it will get whipsawed to death, how would I best define a "trendy" market?

I need to somehow develop some strict criteria (maybe based on price action for the last x periods??) that tells me if the market for a given day is likely to be a trendy market or not??

Hope that makes sense.

Any ideas would be much appreciated.

Nizar.

Nizar,

I also have this problem with a couple of systems that trade hourly bars on the DOW & ES.

It's difficult to find filters which are worthwhile because usually price will follow a pattern of congestion, breakout, trend, congestion. Therefore the lag of the filter will often keep the system out of many of the best 'breakout' trades.

You could try widening the signal trigger to avoid whipsaws but this will obviously delay entries.

 

[tech/a]

Id say your exit filter is lagging.
ie returning close to your entry filter.

Try targets.
You need a trend to profit 1 bar,2,3 doesnt matter but you need a trend.

You need to define a point to point method to capture as much as you can.
I use price bars in whatever timeframe to do this.---works.

Then work with your trade management to suck as much out as possible.

 

[nizar]

What time frame is your trend?

Intraday, the strategy works off 5-minute bars but requires immediate executions for entries and exits.

 

[professor_frink]

Intraday, the strategy works off 5-minute bars but requires immediate executions for entries and exits.

Is this for forex(I think I remember you saying this was what you were looking at trading a while back)? If it is, do you have any filters for what time of day you take trades? There are times when forex is much more likely to be trendy, and times when it's more likely to be chop and slop

 

[prawn_86]

Is this for forex(I think I remember you saying this was what you were looking at trading a while back)? If it is, do you have any filters for what time of day you take trades? There are times when forex is much more likely to be trendy, and times when it's more likely to be chop and slop

Totally agree with this. From Franfurt open through to about an hour after London open is the most liquid for the pair I scalp. Thats 4 - 6pm Aus EST.

Also, i dunno if you can code it, but i find that when i open my charts if the range on the vertical axis is > 20 pips, you know it is a trending market and that there is a chance to make profits. But if the range is about 10pips then you know you are likely to get whipsawed and smashed because of brokerage

 

[julius]

Id say your exit filter is lagging.
ie returning close to your entry filter.

Try targets.
You need a trend to profit 1 bar,2,3 doesnt matter but you need a trend.

You need to define a point to point method to capture as much as you can.
I use price bars in whatever timeframe to do this.---works.

Then work with your trade management to suck as much out as possible.

Tech, I'm not following, could you re-word please.

Niz - can you post some screenshots of the whipsaws on the 5-min charts ?