Good morning guys,
I've spent a while scouring the forums, especially the 'trading systems' section, where I find an idea I like, code it up in to an EA and test it's performance. Doing this encouraged me to start my first 'live' account, trading the system found here : https://www.forexfactory.com/showthread.php?t=730458
I think the next step on from this is building an EA without a set strategy as such, and offering myself a challenge to almost build a base to work from in future. One thing that's always attracted me in coding is trying to replicate something that seems easily distinguishable to the human eye, but difficult to write a rule set to.
Looking through multiple youtube videos, posts on here and books available, I've opted to head in the direction of Price Action trading. Not entirely sure how profitable this is and I know that's hard to quantify due to its discretion based nature, but trying to automate channels, trend lines and SR levels really seems like an interesting task.
I will state that I may end up using some indicators; not necessarily to determine open/closes but more to save coding (to establish current trend, I could code a 'line of best fit' through the previous 10 closes and determine if its an upwards or downwards slope... or I could slap on a parabolic SAR or pair of moving averages, for example).
I think to start with, I need to get the functions coded that search for the different building blocks of price action trading. I will probably do them in order of perceived difficulty: 
Although it may not be exhaustive, it looks like a decent list to start from.
On top of this, I think it's useful to give my aims/rules/information for this project/journal:
I've spent a while scouring the forums, especially the 'trading systems' section, where I find an idea I like, code it up in to an EA and test it's performance. Doing this encouraged me to start my first 'live' account, trading the system found here : https://www.forexfactory.com/showthread.php?t=730458
I think the next step on from this is building an EA without a set strategy as such, and offering myself a challenge to almost build a base to work from in future. One thing that's always attracted me in coding is trying to replicate something that seems easily distinguishable to the human eye, but difficult to write a rule set to.
Looking through multiple youtube videos, posts on here and books available, I've opted to head in the direction of Price Action trading. Not entirely sure how profitable this is and I know that's hard to quantify due to its discretion based nature, but trying to automate channels, trend lines and SR levels really seems like an interesting task.
I will state that I may end up using some indicators; not necessarily to determine open/closes but more to save coding (to establish current trend, I could code a 'line of best fit' through the previous 10 closes and determine if its an upwards or downwards slope... or I could slap on a parabolic SAR or pair of moving averages, for example).
I think to start with, I need to get the functions coded that search for the different building blocks of price action trading. I will probably do them in order of perceived difficulty:
- Single/Double/Triple/etc. candlestick formations (in order)
- Horizontal support/resistance
- Trend Lines
- Equidistant channels
For the candlestick formations, I'm currently going to use the following:
Although it may not be exhaustive, it looks like a decent list to start from.
On top of this, I think it's useful to give my aims/rules/information for this project/journal:
- Beat 2.5% profit per year.. the current rate of my savings account. Seems minimal compared to the 'Watch me win 40% per day' threads on here, but that's not what this is about.
- Keep good-form, commented code (I'm self taught, so this will be judged against my own standards).
- Ensure EA is able to adjust between currencies/time frames with minimal user interaction.
- Update at least once per week with coding updates; this will start off much more frequent of course.
- Code in MT5 (MQL5) using AdmiralMarkets broker; I know MT4 is the go to, but frankly, I put time in to learning MQL5 to be future proof and see no reason to backdate.
- Initial 'Live' account test to be started by 1/1/19
Wish me luck guys, and I'll be in touch soon!
Will code in MT5 for free.. EAs only, can't be f*cked with Indicators
