Is there any reason why you are using a bar count of 50 in the sixths indi and not 120 as Nanningbob uses in his indi?
It's been a long day so I could be missing something
Should it be 120?
I just looked at the difference between 50 and 120; there is a difference but not a lot.
in the Equity milliped thread I wrote a short comment...
This is the idea for the next EA generation!
It's very easy: no charts, no indicators, only position management... Maybe in a few months I know how to do that.
I've just re-read the 'NB 4H TRADING SYSTEM V. 5.0 PHILOSOPHY.doc' document from Nanningbobs thread and in this it recommends using a 4HR chart showing 1 months of price action. The sixths are then calculated from this. By my reckoning this will be 120 4HR bars.
I will have a play around with different settings to see what difference it makes.
Mac uses 168 for 1 month
I think eventually we may need a different look back setting for individual pairs.
1st pic shows results for 2 years backtest on UJ M15 using 192 bars (48 hours) & 720 bars (180 hours) as trigger points. So 192 bars is the best - right?
Now check out GJ with exactly the same settings .... hmmmm
The reason I added being able to fix the look back period is so that we can fine tune the entry points.
Sorry if this is dumb question but which EA were you using for these backtests? The only reason I ask is that I can't get anywhere near the results using the EA in post 1. Also, it looks like the results you posted show the lot sizes changing but the EA in post 1 has a fixed lot size.
The point of the post was to illustrate how using the same entry condition (number of look back bars) does not work very successfully if applied across the board to all pairs.
The EA used in the tests is an earlier version of NanBob's strategy available in the NanBob thread.
Mac, I thought I saw somewhere that you were using 168 bars. Is that no longer the case?
Things are getting a little confused as originally I was using the indi with a different EA which also allows slightly riskier trades to be taken between the top/bottom two lines.
Probably best to set a default at 120 bars look back on H4 (20 trading days) & adjust this as we go along.
I was just taking a second look at the code(has very few trades) The Bandwidth Indicator iCustom call seems to be short one parameter
In a way it is irrelevant; I do not see how to use Bandwidth to spot a bulge, so the default effectively turns it off. I have not given it much thought either - I am learning to trade this stuff manually and having a lot of fun doing so.
I have made one major change.
The NB MACD is a visual illusion that works brilliantly for manual trading.
Using the mathematically-accurate StochMacdSqualou for manual trading results in most trades not being taken because it so rarely passes the 85/15 levels - the reason this robot has taken so few trades.
On the demo, I changed the requirement to 75/25 and the bot immediately took the chfjpy buy that I had taken manually a short while before.
So, I have added two new inputs to the StochMacdSqualou inputs - OverBought and OverSold - so users can play with these levels for themselves.
Download the latest from post 1.
I tried to post here .....
now everything is okay.
Post back here so everyone can see.
this is what I posted in wrong thread:
I downloaded the latest version you posted earlier.
Putting on a demo FXCM.
Not open a position for an error 130.
SL was the setting of 0 (zero)
I put SL to 500 to try.
But still give the same error.
You know why?
I thought decimals in code, in this case are three.
is it possible??
I tried to post on your page but I did not authorize. Could you arrange it?
© Forex Factory