A System.

No, a demo account would not be adequate. The market moves depending on open orders. For the system to be valid, you have to have real orders. Oander allows you to trade as small as 0.0001 pence a pip (not sure about their MetaTrader interface). So there is no disadvantage to using a real account.

Unless someone has already done the research, I reckon the system will get slaughtered.
 
No, a demo account would not be adequate. The market moves depending on open orders. For the system to be valid, you have to have real orders. Oander allows you to trade as small as 0.0001 pence a pip. So there is no disadvantage to using a real account.


Let's just get it rolling. One of the advantages of this system is lack of 'thinking' that is required, it's very simplistic and easy to follow.
 
trade params.jpg

Parameters.
 
Let's just get it rolling. One of the advantages of this system is lack of 'thinking' that is required, it's very simplistic and easy to follow.

What is the entry criteria? Manually open a trade in the direction of the preceding candle colour, then EA takes over?
This is probably the simplest way to do it.

Exit variable based on % of instrument as a trailing stop.
Stake size variable based on % of instrument value applied to.
Trying to keep all things equal would probably be best, especially when analysing results.

It would be good if LV did a walk through of all the processes involved in constructing operating and testing.

Some of us really don't know where to start with all this stuff.
As for coding :cry::rolleyes::eek:
 
What is the entry criteria? Manually open a trade in the direction of the preceding candle colour, then EA takes over?
This is probably the simplest way to do it.

Exit variable based on % of instrument as a trailing stop.
Stake size variable based on % of instrument value applied to.
Trying to keep all things equal would probably be best, especially when analysing results.

It would be good if LV did a walk through of all the processes involved in constructing operating and testing.

Some of us really don't know where to start with all this stuff.
As for coding :cry::rolleyes::eek:


Coding?:eek: I'll leave that to LV. Entry is based on colour of last daily candle. What we can do is go to the hourlies, if the hourlies are trading the same colour then we enter immediately. If the hourlies are trading the opposite we enter with a limit +/- 25 pips.
 
Coding?:eek: I'll leave that to LV. Entry is based on colour of last daily candle. What we can do is go to the hourlies, if the hourlies are trading the same colour then we enter immediately. If the hourlies are trading the opposite we enter with a limit +/- 25 pips.

yes yes, the EA has to start...what starts it ?
 
Guys don't get your hopes up, TBH I'm expecting it to fail.
Having messed about with HA before, my own personal view is its just another
lagging indicator.

Thats on a purely mechanical level.
Thing with most indicators is people do use them as a crutch,
or supplementing them with their own discretion.
Thats where the illusion of an indicator working comes from.

Remove that discretion, the kind of thing that decides price is stagnant,
and you just get loads of signals which burn any profit you've made and then some.
Still I'll crack on as it reveals the gap between discretionary interpretaion of an
indicator and its pure performance with no interference.

Thats why me and the hare bang on about the influnece of randomness
as a more sound basis for an algo.
Simply due to the fact that a dumb ass machine can handle that.
No interpretation finesse needed.
 
At the moment i have to do all the monitoring and place orders/trades accordingly. I'd like to make the system less manual.
 
At the moment i have to do all the monitoring and place orders/trades accordingly. I'd like to make the system less manual.

If this is something you do trade profitably at the moment, I'd guess
you are applying more subconscious discretion than you realise.

The best I've got so far is 38k profit in 2008 (100k trades static pos sizing),
every other year is flatish or dismal loss with this HA so far.
Thats the issue really, I just don't think its really possible to program
discretionary decision making, if you do try, more often than not
you end up with too many parameters, which results in greater
possibility of curve fitting.

As you do OK with it manually, it does tend to suggest you
are using greater discretion and overall sentiment.
Neither of which computers are very good at, not at this end of the scale anyway.
 
At the moment i'm monitoring the positions all day, although i'm not tied to the screen, a mobile phone app is useful. Trades and orders are being put through when the previous has finished, so i'm in the market virtually 24/7 100%.
 
At the moment i'm monitoring the positions all day, although i'm not tied to the screen, a mobile phone app is useful. Trades and orders are being put through when the previous has finished, so i'm in the market virtually 24/7 100%.

So you're always in.
How often do your stops get hit?
 
If this is something you do trade profitably at the moment, I'd guess
you are applying more subconscious discretion than you realise.

The best I've got so far is 38k profit in 2008 (100k trades static pos sizing),
every other year is flatish or dismal loss with this HA so far.
Thats the issue really, I just don't think its really possible to program
discretionary decision making, if you do try, more often than not
you end up with too many parameters, which results in greater
possibility of curve fitting.

As you do OK with it manually, it does tend to suggest you
are using greater discretion and overall sentiment.
Neither of which computers are very good at, not at this end of the scale anyway.


Cheers LV, thanks for the input.
 
Unless there is data indicating a positive return, I don't think it's worth the effort to build a system. Generally, if it doesn't work manually, it will not work systematically either.

As market moves are fractal in nature, if the system works, it should work in lower time frames also. Using lower time frames means you don't have to waste a large amount of time before knowing it is a dead end.
 
Coding?:eek: I'll leave that to LV. Entry is based on colour of last daily candle. What we can do is go to the hourlies, if the hourlies are trading the same colour then we enter immediately. If the hourlies are trading the opposite we enter with a limit +/- 25 pips.


I'm going to change this slightly. All trades will be entered on the next 0.25 cent in the desired direction.

If a pair has no open trade, and the pair is trading at 60, then a order will be placed either at 50 or 75.
 
I'm going to change this slightly. All trades will be entered on the next 0.25 cent in the desired direction.

If a pair has no open trade, and the pair is trading at 60, then a order will be placed either at 50 or 75.

If you do have a crack at it yourself, beware, automation can be a deep dark hole :LOL:
If what you do manually works, I'd just use the data for manual backtesting.
 
If you do have a crack at it yourself, beware, automation can be a deep dark hole :LOL:
If what you do manually works, I'd just use the data for manual backtesting.


I can see that u like ur coding and system building. I'm using HA as a basis for entry, btbh, it could be anything, a coin toss for example.
 
Top