Not logged in
Viewing 7 posts - 1 through 7 (of 7 total)

Forums>StrategyQuant>Pre-Sales Questions>Best Settings for Ranking

  • #247211| Reply
    Participant
    23 Posts

    Im sure this may have been asked before, but Im wondering if anyone would like to share their favourite ranking for generating random trades, with a view of having the best long-term probability of passing further robustness tests.

    As an example, I am using (D1 EURUSD)

    Net Profit > 0 

    Average Win > Average Loss

    Max DD% < 40%

    Stability > 0.5

    Avg Trades Per Month > 1

    Ret/DD Ratio > 1.5

    Would anyone recommend changing these or adding to them?

    Thanks, M

     

    #247217 Reply
    Customer
    291 Posts

    i mostly use profit factor > 1.3, win% > 30%, RDD > 0.5/per year, number of trades > 30/per year

    if you dont tell us, how long data you are using, the ranking RDD 1.5 cant be analysed

    net profit > 0 is unnecessary, because you are using RDD > 1

    why to compare avg win and loss? its your choice..

    maxDD in percent doesnt mean nothing, because you are not telling us your capital and MM. i am always using ranking criteria which are not capital or MM addicted

    You want to be a profitable algotrader? Sharing of final strategies with real traders just started. Fill in this FORM. 1500+ final SQ3 and SQX strategies for members running on demo accounts DEMO ACCS. We provide also strategies for indices - DAX and DOW JONES, because we have realtick data from brokers.

    #247218 Reply
    Participant
    23 Posts

    That’s helpful, thanks Hankeys.

    I guess there is no strictly right or wrong answer to the settings, although some are logically incompatible with one another, like the net profit and RDD, as you mention. The RDD is then linked to losses over the testing period and makes more sense than net profit alone, so its more of a measure of fitness over time rather than an end number.

    The data I have is from 2003 to 2019, i.e. 16 years so I should set my RDD to 8, then?

    For MM I have 2% fixed amount of account and my capital starts at 500. I chose Max DD% as 40% since I saw a youtube video saying that one measure of a good strategy is that you dont need to incur a large amount of loss over the testing period. Should this Max DD be linked to the number of test years, like the RDD%?

    Also, you dont use stability at the generation stage to search for better equity curves?

    Thanks, M

    #247220 Reply
    tomas262
    Administrator
    1116 Posts

    Another approach to this would be to use as few criteria as possible. In such case I would only use these two: Profit Factor > 1.2 and # of Trades > X (usually at least 100 trades depending on trading timetrame) so there is statistical significance for the profit factor value. The problem with the criteria you mentioned above can be it can filter out very robust strategies while keeping strategies that look “perfect” but are not robust at all

    #247222 Reply
    Customer
    291 Posts

    i cant tell exactly the settings i would use fo D1 strategies, because i dont trade them and dont use them…i like exit on friday to avoid weekly gaps, and for D1 strategies this doesnt make more sense

    for EURUSD i will take 0.5 RDD per year as a minimum

    about stability – i dont use it, because good PF and RDD will make the EQ look good, stability will be for 100% bigger than 0.5

    You want to be a profitable algotrader? Sharing of final strategies with real traders just started. Fill in this FORM. 1500+ final SQ3 and SQX strategies for members running on demo accounts DEMO ACCS. We provide also strategies for indices - DAX and DOW JONES, because we have realtick data from brokers.

    #247225 Reply
    Participant
    23 Posts

    Thats great, thank you. I would not have thought of these approaches looking my strategies in a vacuum; I think I might be too caught up on the stability aspect (while even a high stability does not seem to always give that perfect equity curve – Ive got some at 0.81 which are still quite jagged)

    I think Tomas is right – too many criteria – Ive got many strategies which look great but do not pass the robustness tests, the trick to learn is to generate strategies which just do not look good cosmetically…

    #250161 Reply
    Customer
    224 Posts

    RDD

     

    Hello, tomas262,

    Is the RDD that they said Ret/DD Ratio in SQ?

     

    And is 1.2 for profit factor  the reasonable lower limit?  Have the calculation of  profit factor in SQ considerd the commission and slippage cost or not? (Futures markets)

     

     

     

     

Viewing 7 posts - 1 through 7 (of 7 total)
Reply To: Best Settings for Ranking

You can use BBCodes to format your content.
Your account can't use Advanced BBCodes, they will be stripped before saving.

Your information: