trades yesterday
- #873
- Edited 11:06pm Aug 28, 2018 8:53pm | Edited 11:06pm
{quote} Thanks for you input Noshali, I have not been able to get a GBPUSD trade on a 1M chart. It seems this EA is more or less built for 5M and 15M.
Profit since Aug 8 when I started using the Set files GBPUSD M1..Using autolot
- #874
- Aug 29, 2018 12:00am Aug 29, 2018 12:00am
GBPUSD M1 set file analysis....
Here is how I do my testing and try and optimize.
We all know that grid type trading is VERY Dangerous. So how do we minimize risk.
Entries should be very good…if entries are good then formation of grids will be better. I try and force less and less entries. This causes a problem that the EA trades very seldom and reduces our profit. To overcome that I use lower timeframe and use auto lot so it trades bigger lots as the balance grows.
Following is the analysis on GBPUSD M1, set file is on post 800. My way of testing could be totally wrong so if anyone feels that there is a better way..please share your opinions as would love to hear from you…please punch holes in my analysis…I am open to any criticism. I don’t claim to be an expert and still have a long way to go to really understand forex.
I back tested from 2016 and looked at the MAXIMAL DRAWDOWN. This is the number that tells me minimum equity I require for the account to not blow up. I normally fund double that. Also, if the markets behave in a similar pattern as they did from 2016-todate. BACK TEST is an INDICATION and not a GUARANTEE.
Please see the attached screenshots…I have tested with 99% tick data…slippages…variable spread, ICMarkets commission and latency. I use a 10K balance when back testing over a longer period so that I can see the max DD. If I test with a 1K balance and the account blows up then I have to test again with a bigger balance. Back testing using the following settings is very very slow so better to test in one go.
0.01 lot is used…this will give me the base line. Now if I use auto lot the DD will increase accordingly. I use auto lot settings that ensure that the first trade opened on the account will be that of 0.01 and will go up as account balance increases. Any DD that happens on the account should be in ratio to the increase in the account balance.
According to the test, I need a minimum of $1,188 for the account to not blow. So double this amount to $2K and then test with auto lot.
For optimization, I will test from 2016 to date and take the worst month (highest DD). I will only optimize on that one month and test again. I don’t optimize on longer periods as that is too curve fitted and results tend to be bad on a live account.
Well that is all I do. I feel these settings are good for a live account with a MINIMUM BALANCE OF $2K
0.01 lot testing from 2016 to 2017
2018 Analysis with AUTO LOT 1% you need a minimum of $1487 for account to not blow. Account doubles..you should allocate $2000 to one set file...you will sleep well at night.
Here is how I do my testing and try and optimize.
We all know that grid type trading is VERY Dangerous. So how do we minimize risk.
Entries should be very good…if entries are good then formation of grids will be better. I try and force less and less entries. This causes a problem that the EA trades very seldom and reduces our profit. To overcome that I use lower timeframe and use auto lot so it trades bigger lots as the balance grows.
Following is the analysis on GBPUSD M1, set file is on post 800. My way of testing could be totally wrong so if anyone feels that there is a better way..please share your opinions as would love to hear from you…please punch holes in my analysis…I am open to any criticism. I don’t claim to be an expert and still have a long way to go to really understand forex.
I back tested from 2016 and looked at the MAXIMAL DRAWDOWN. This is the number that tells me minimum equity I require for the account to not blow up. I normally fund double that. Also, if the markets behave in a similar pattern as they did from 2016-todate. BACK TEST is an INDICATION and not a GUARANTEE.
Please see the attached screenshots…I have tested with 99% tick data…slippages…variable spread, ICMarkets commission and latency. I use a 10K balance when back testing over a longer period so that I can see the max DD. If I test with a 1K balance and the account blows up then I have to test again with a bigger balance. Back testing using the following settings is very very slow so better to test in one go.
0.01 lot is used…this will give me the base line. Now if I use auto lot the DD will increase accordingly. I use auto lot settings that ensure that the first trade opened on the account will be that of 0.01 and will go up as account balance increases. Any DD that happens on the account should be in ratio to the increase in the account balance.
According to the test, I need a minimum of $1,188 for the account to not blow. So double this amount to $2K and then test with auto lot.
For optimization, I will test from 2016 to date and take the worst month (highest DD). I will only optimize on that one month and test again. I don’t optimize on longer periods as that is too curve fitted and results tend to be bad on a live account.
Well that is all I do. I feel these settings are good for a live account with a MINIMUM BALANCE OF $2K
0.01 lot testing from 2016 to 2017
2018 Analysis with AUTO LOT 1% you need a minimum of $1487 for account to not blow. Account doubles..you should allocate $2000 to one set file...you will sleep well at night.
- #875
- Aug 29, 2018 2:40am Aug 29, 2018 2:40am
GBPUSD M1 set file analysis.... Here is how I do my testing and try and optimize. We all know that grid type trading is VERY Dangerous. So how do we minimize risk. Entries should be very good…if entries are good then formation of grids will be better. I try and force less and less entries. This causes a problem that the EA trades very seldom and reduces our profit. To overcome that I use lower timeframe and use auto lot so it trades bigger lots as the balance grows. Following is the analysis on GBPUSD M1, set file is on post 800. My way of testing...
very useful stuff here. Everyone should thank you
