Topic: Acceptance Criteria Opinions Requested: R/DD

I am using an OOS period of ~7 months (IS ~15 months).

Minimum Number of Trades is 100 IS and 40 OOS.

What is the "Minimum R/DD" that may be "acceptable" in the OOS period.

Let me define "acceptable" in this case. This would be (close to) the minimum R/DD during this OOS that could occur for the strategy to still be considered for use in your portfolio pending further results or testing. For the lower R/DD values, this would be an "inconclusive results" conclusion. The strategies that fall within these range can be further considered or pruned altogether.

The reason I ask is because currently the Acceptance Criteria do not allow for a R/DD of less than 2. I think this may be a bit too restrictive for short-term periods and it this minimum should probably be dropped to 1 (maybe even less?). In general, R/DD gets larger with more data. A standard I've seen used for validation is R/DD of 0.5 per year.

So for example, in my 7 month OOS period, my EA has made a total profit of $1000 while incurring a $1000 drawdown somewhere along the way, that seems that it may still be a potentially viable strategy with a R/DD of 1.

Re: Acceptance Criteria Opinions Requested: R/DD

I suggest a minimum of 2.5 or 3

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

3 (edited by qattack 2017-08-21 02:15:07)

Re: Acceptance Criteria Opinions Requested: R/DD

Let me restate my question a bit...

2.5 or 3 would be fine for an IS period, where the generation has occurred. As we all know, we can find more than enough strategies that meet this simple requirement.

The OOS will be significantly weaker across the board. My thinking is that the OOS period would roughly reflect what would happen if we place a given EA on a demo account.

So to rephrase the question: For our plans to deploy 100s of EAs, after a full seven month testing period on a demo account, if you were pruning them ONLY using R/DD, what is the bare minimum value that you accept? Those with "bare minimum" values would be worthy of continued testing perhaps.

Now, a slightly different question: after this period, what is the minimum R/DD value that you would consider promoting an EA to a live account (for significant money), but this time using other stats as guides as well? Be realistic with regards to other stats, as you are not getting a R/DD of 0.5 and a Profit Factor of 2.0, for example.

Perhaps 2.5 or 3 is the answer; maybe we are generating enough super strategies that we can afford to toss out a boatload of OK, but not great possibilities.

Re: Acceptance Criteria Opinions Requested: R/DD

qattack wrote:

I am using an OOS period of ~7 months (IS ~15 months).

Minimum Number of Trades is 100 IS and 40 OOS.

What is the "Minimum R/DD" that may be "acceptable" in the OOS period.

Let me define "acceptable" in this case. This would be (close to) the minimum R/DD during this OOS that could occur for the strategy to still be considered for use in your portfolio pending further results or testing. For the lower R/DD values, this would be an "inconclusive results" conclusion. The strategies that fall within these range can be further considered or pruned altogether.

The reason I ask is because currently the Acceptance Criteria do not allow for a R/DD of less than 2. I think this may be a bit too restrictive for short-term periods and it this minimum should probably be dropped to 1 (maybe even less?). In general, R/DD gets larger with more data. A standard I've seen used for validation is R/DD of 0.5 per year.

So for example, in my 7 month OOS period, my EA has made a total profit of $1000 while incurring a $1000 drawdown somewhere along the way, that seems that it may still be a potentially viable strategy with a R/DD of 1.

There is 3 catch here.

First Catch

The reason I ask is because currently the Acceptance Criteria do not allow for a R/DD of less than 2

I can filter and use R/DD less than 2, just use the "down" keyboard key
http://i.imgur.com/9IfHU3S.gif



You have others filters to play with

now that the software has acceptance criteria for Complete backtest, In Sample (training) part and Out of Sample (trading) part you can easily define a way to get out this trap

http://i.imgur.com/QrtKEcg.gif
All Data:
Maximum equity drawdown : 5%

In Sample: Minimum profit per day > 30
OOS: Minimum profit per day > 30


I belive that R/DD its not a metric that you use by itself, You use it to improve somethint that is already good.

Vinicius Pereira, Portuguese Support.
Improve your trading with my strategies & signals on MQL5. High success rate & many followers. Check them out & join my EAS telegram group for updates.

Re: Acceptance Criteria Opinions Requested: R/DD

The reason I ask is because currently the Acceptance Criteria do not allow for a R/DD of less than 2

I'll fix that today.