New MA MTF indicator

[SteveHopwood]

Hi @all.

Dont use the TMA, there seems to be a coding jumble in it.

I feel your pain.

Nobody better to sort it out than you.

[pacinvest]

Ok I will start the ball rolling here on the use of the TPB in the MTF MA bars indy.

I decided to begin my testing round with a really simple approach to make sure I understand this indy that Thomas has given birth to. I am afraid that this first test doesn't even scratch the surface of the potential of what the indy can do, but I needed to start from the basic way I trade the Two Pole Butterworth filter (TPB) and then build on that with more complex ideas.

So here is what TPB study test 1 look likes. I have prepared some visual backtest charts for you and posted the set files for this first test.

This is an H4 entry Day exit strategy
It uses only two time frames
Entry is on agreement of H4 and Day signals
Exit is on end of day signal only

A rule I use for this test is entry on first H4 signal in agreement with Day signal and let the trade run until the Day signal ends. It would be cool to be able to set the criteria for when the yellow wavy appears as in this test.

What this simple approach does is let us have a bit of look at TPB for starters but also importantly demonstrates the value and importance of the next higher time frame to entry and exits. Obviously the next step is to test this against a sea of other possibilities this indy offers.

For the TPB period parameter I used 15 for both timeframes 1 and 2. The other parameters are default.

Lets look at some charts. In the chart AUDUSD H4 I drew some bands to show each trade and the estimated value in pips for each trade is there in white above the entry signal.

There are two more examples this time GBPJPY and EURUSD. I didn't draw in the bands to show the trades just put the estimated pip value for each trade above the entry signal.

Well pretty good result, I think if you spend some time with the TPB you will see the potential. I will carry on with the some more tests, but basically even this test 1 looks awfully good. If you were running this strategy during a good trending period there would be some huge profits, however it pleases me that the strategy gets through some pretty sea saw and quiet periods, that is important. Also there were no large losses and nearly all trades had bugger all draw down. There are definitely ways to fine tune entries and exits beyond simple approach. I wouldn't mind having a simple EA to trial this strategy or better yet a basket EA. A big thanks Thomas for your insightful approach and skills programming this indy.

Have fun guys.

[celte83]

Morning all,
how do you trade with this indicator? Because it repaints, it recalculates. Normal since we are in MTF. It recalculates because of higher TF.
Mark the arrows and reload the indicator or close and reopen Empty4.
Some arrows disappeared

Not easy to trade

cheers
jf

[tomele]

Morning all,
how do you trade with this indicator? Because it repaints, it recalculates. Normal since we are in MTF. It recalculates because of higher TF.
Mark the arrows and reload the indicator or close and reopen Empty4.
Some arrows disappeared

Not easy to trade

cheers
jf

Hi.

The indi doesnt actively repaint, but if you use it in a wrong way, the historical values differ from the realtime values. Like in most MTF systems.

I have written an explanation in post 1. I recommend reading my "UPDATE 8-11-17" to everyone. You will get a better understanding of this indi and how to use it correctly.

The TPB smoothing method doesnt have this problem, so everything that @pacinvest has presented is valid.

Cheers.

[celte83]

thanks a lot tomele for these explanations updated post 1 and here

santé !
jf

[pacinvest]

Hi All

Have two more TPB tests running with two MA's working and one with three time frames. I will try and enter these trades signalled on my test Demo account but will be missing lots of entries unfortunately doing this manually. The beauty of fuzziness and wavy lines is working away in these next two tests. There is some clever maths going on under the hood here. I am trying to check signals in real time too, but unfortunately my trading time is quite limited at them moment. Will share some more from tests 2 & 3 soon. Happy to entertain comments and suggestions and 'criticism' on working with TPB.

Thomas what's coming from me is I want to be able to generate a yellow wavy for exit signal based on only one of the higher timeframes, while using multi- time frame fuzziness and multiple MA's etc for the entry on the lower chart. In very simple language this is essentially trading on a higher time frame but nailing good entries on a lower time frame. In my manual trading I try to do this and use tighter stops to leverage the trades higher or multi-level trade once protective breakeven stops can be placed in early trades of the higher timeframe run. All these things become possible with a MTF approach that works well. Or put another way we attempt to just simply enter on lower time frame and let trade run the move on the higher time frame.

Thomas sort me out here if I am confusing the function and programming of the yellow wavy. Or how best to use it.

Cheers
Vince

[tomele]

Hi @all.

Some news about my TMA struggling in post 1. Read the UPDATE 10-11-17 and download the new versions.

Have fun.

[pacinvest]

Hi Thomas

Part of my little journey here is to make sure I understand the parameters and what I can do with them, so I am afraid I have to ask some basic questions.

1) The parameters Timeframe1MAsToUse & Timeframes To Show

If I want to have calculations done on timeframes 1 and 2 only and I want to show bars for timeframes 3 & 4 but not use them in calcs I would input the parameter as follows:

(Note: have set periods for all four timeframes, i.e. not set to 0)
Timeframe To Show: 1234
Timeframe1MAsToUse: 12
Timeframe2MAsToUse: 12
Timeframe3MAsToUse:
Timeframe4MAsToUse:

When I try this what happens is the indy seems to have a problem and prints no bars or signals on screen.
I tried putting a 0 input for timeframes 3&4 but that didn't chance the result.

Can you shed some light on this? Am I trying to do something not intended for this version of the indy?
________________________________________________________________________

2) The Input parameter Strategy to Use
I can't get my head around the difference between options AllBarsAgree and MoreBarsAgree, basically I don't have a feel for what the MoreBarsAgree parameter does?
________________________________________________________________________

3) I don't really understand how the Threshold Percent parameter? would love to understand that a bit better.
_______________________________________________________________________

Have had some fun looking at the Hurst set file and comparing it to TPB, that Hurst setup is definitely worth looking into I am impressed. Will post a few charts later.

Entries vs exits

Lastly in tests so far I am impressed with entries as long as market is moving and I think the potential to signal consistently good exits is great too, which is critical to an EA and not easy to achieve. But I think in some strategies the way you calculate an entry might be quite different than how you would calculate exits. Taking one example I like the idea of lining up all the ducks (higher timeframes), trading with trend direction and then having an entry that is quite sharp on the trading timeframe (shortish period and based on only one upper timeframe maybe) . And then in contrast I like the idea of letting trades run on a higher timeframe based signal and in doing so not having to deal with pull backs etc on the trading timeframe. Does this make sense and would it be a huge problem to basically separate entry and exit signals in this case arrows and yellow wavy signals? I'm also am excited about the ability of this indy to signal multi-level trading, but I'll park that for further down the track, is related to enetry/exit option diferences though.

Thanks for the work your doing here Thomas.

[goldcourse]

Hi Thomas,

Is it possible to add push notification alert in this already brilliant indicator?

Thank you.
Vincent

[tomele]

Hi Thomas

Part of my little journey here is to make sure I understand the parameters and what I can do with them, so I am afraid I have to ask some basic questions.

1) The parameters Timeframe1MAsToUse & Timeframes To Show

If I want to have calculations done on timeframes 1 and 2 only and I want to show bars for timeframes 3 & 4 but not use them in calcs I would input the parameter as follows:

(Note: have set periods for all four timeframes, i.e. not set to 0)
Timeframe To Show: 1234
Timeframe1MAsToUse: 12
Timeframe2MAsToUse: 12
Timeframe3MAsToUse:
Timeframe4MAsToUse:

When I try this what happens is the indy seems to have a problem and prints no bars or signals on screen.
I tried putting a 0 input for timeframes 3&4 but that didn't chance the result.

Can you shed some light on this? Am I trying to do something not intended for this version of the indy?
________________________________________________________________________

2) The Input parameter Strategy to Use
I can't get my head around the difference between options AllBarsAgree and MoreBarsAgree, basically I don't have a feel for what the MoreBarsAgree parameter does?
________________________________________________________________________

3) I don't really understand how the Threshold Percent parameter? would love to understand that a bit better.
_______________________________________________________________________

Have had some fun looking at the Hurst set file and comparing it to TPB, that Hurst setup is definitely worth looking into I am impressed. Will post a few charts later.

Entries vs exits

Lastly in tests so far I am impressed with entries as long as market is moving and I think the potential to signal consistently good exits is great too, which is critical to an EA and not easy to achieve. But I think in some strategies the way you calculate an entry might be quite different than how you would calculate exits. Taking one example I like the idea of lining up all the ducks (higher timeframes), trading with trend direction and then having an entry that is quite sharp on the trading timeframe (shortish period and based on only one upper timeframe maybe) . And then in contrast I like the idea of letting trades run on a higher timeframe based signal and in doing so not having to deal with pull backs etc on the trading timeframe. Does this make sense and would it be a huge problem to basically separate entry and exit signals in this case arrows and yellow wavy signals? I'm also am excited about the ability of this indy to signal multi-level trading, but I'll park that for further down the track, is related to enetry/exit option diferences though.

Thanks for the work your doing here Thomas.

Hi Vince.

Sorry for answering late, but I was sick and now we have renovations going on with dirt and chaos all over the place. Re your questions:

1.) Just enter "Timeframe To Show: 12". Otherwise the indi waits for status from timeframes 3 and 4, but they have no MAs to calculate. I will possibly add some code to avoid that problem.

2.) MoreBarsAgree simply means that the majority of bars agree. If you have 3 timeframes, 2 must agree. If you have 4 timeframes, 3 must agree.

3.) The "Threshold Percent" is the core of decisions in fuzzy logic. On every single timeframe you have 20 binary results if you have 5 MAs and all 3 conditions (slope, price/averages, averages/averages) activated. All those results are (equal-weighted) added up and scaled to a score with range of -100 to 100. So if all results say "up", the score would be 100, if 10 are "up" and 10 are "down" the score would be 0, and if all 20 are "down" the score would be -100. The threshold defines the value that must be exceeded to consider the overall score an "up" or "down" signal. If the threshold is 25, all values from -25 to 25 would be "no signal". A value of 26 or more would be an "up" signal. A value of -26 or less would be a "down" signal. Always for one timeframe.

4.) I have to think about the exit part. Maybe a second threshold for exits or different thresholds for the timeframes. I dont want to mix up the fuzzy logic with discrete logic. You can observe the four bars and ride your trade as long as the higher timeframe bar has a valid signal. Or did I miss something?

Cheers.