Strategy creation

Discussion in 'Strategy Building' started by andrew.whad, Jun 14, 2020.

  1. Hello everyone,

    I want to ask for some input on my current research to make sure that I am on the right track and not losing my mind.

    After many attempts, I realized discretionary trading just isn’t for me I decided to use a systematic approach and build some strategies that I would trade manually. The strategies will be based solely on Price Action with no indicators, news, or releases taken into account for now. I am looking at the 15 min intraday timeframe.

    I trade mainly indices and right now I am focusing on the FTSE100.

    1 Types of strategies:

    I took 2010 and looked at every single day in an asset and found the following strategies can be applied during the European session:

    7:00 to 10:00
    European open breakout (to the downside or upside)
    First retracement of previous breakout (up until 10:00 AM)
    Slow grind (up or down) - days with low range (not the best as it ca can also be a slow grind with no clear direction)

    10:00 - 12:00
    Continuation of the initial breakout is very risky and many fake breaks occur (to be mainly avoided)
    Reversal of the initial breakout (prefered) with the final target being the opening price.

    12:00 - 14:00 US Comes in:
    Continuation or break of key levels in the original direction

    14:00 - 15:00
    Usually time to sit still and do nothing

    15:00 - 17:10
    Entering on a retracement if a clear direction is visible from US open.

    2 Strategy analysis
    For each one of the strategies above I take all the days when the strategy worked and compare them with the days when it didn’t work; I then try to find the the difference in price action and also check lower time frames.
    Some of the differences can be new lower highs to indicate a change in a downtrend and a sequence of green candles on 1 min and 5 min.
    This part is very time consuming!

    3 Strategy test
    I created an excel file where I can input the exact details of my strategy (once finished) and it will calculate total trades, wins, losses, everything and then optimise as needed. I plan to run these on data from 2011 to 2015.

    4 Trading test
    Once a strategy has been optimised I will trade it on historical data with Forex tester (not the best but couldn’t find a better alternative). I have an archive of live news and i can see how well i can implement my strategies in a “live” conditions and also how releases and fundamentals will affect them.

    5 Going live
    I will further optimise if needed but if I all goes well I will move that strategy to live markets and trade only that strategy.

    The only problem is that this process is really long but in my mind it makes sense. Sometimes I feel I making no progress at all and that I am just wasting my time; other times I have a breakthrough that motivates me to go further.

    In the end I want to look at one day in IF and THEN terms; if there’s a strong open to the upside between 8:00 and 10:00, between 10:00 and 12:00 I can do A or B or nothing and so on for each part of the day.

    Of course other parts like correlations, risk on/off theme, market mood and volatility matter but I am starting with the basics.

    I would love to know what you think about this and your ideas and suggestions.
  2. qlai


    Discretionary doesn’t mean no rules. What was the problem(s) that you think the new approach will fix?
  3. On some losing trades it was obvious why I lost, but on some I just couldn't figure it out, was it my psychology, the market behaved differently etc.
    With a system (if implemented correctly) it's very easy to see if it's producing the same % of wins as in the backtest/forward; also a clear system has clear rules so if you didn't follow 100% it's clear that the issue is with you / psychology.
    I think right now I am just looking for clarity because there is a world of useless information out there.
    qlai likes this.
  4. wrbtrader


    Discretionary trading means no automation. In fact, even if you use a little automation (e.g. auto stop upon entry) and any other aspect of the trade is via manual execution...

    You're still a discretionary trader.

    Also, if you're using manual rules like the ones you've stated and its not automated nor're still a discretionary trader.

    With that said, there's another type of discretionary trading / subgroup that's called intuition trading. Some refer to it as shooting from the hip, trading via your emotions or making guesses without any background from backtesting / simulation trading results to give reason for the real money trades.

    I'm curious about your below statement...

    • Why not 2011 to 2020 ???
    Is it because you're manually backtesting and 2011 to 2015 is just a next phase in which you will then proceed to 2016 - 2020 because each year is very time consuming ?

    Hopefully your backtesting results will be similar to your simulator results. If not, it could imply you're adapting your trade method in current market conditions without realizing it.

    The process is suppose to be long so that you can discover the nuances of your trade method instead of seeing them when you have real money on the line in stress market conditions that can be chaotic / volatile due to it being a Pandemic.

    As you gain more trade experiences with your trade can adapt / react properly when those nuances show their face during an open trade position.

    Last edited: Jun 14, 2020
  5. Thanks for the long reply WRB.

    I am aware that market patterns every few years or so and that's why it can be a bit hard to make money with and algorithm.
    You did make me think about discretionary trading; I guess my aim is a more systematic discretionary trading.

    That is correct, the plan was to test 2011 to 2015 automatically and 2016 onwards I would look at how news, correlations, and other events would affect the system and then adjust it more...

    So it's supposed to be a long process... I thought so!
  6. Chyu


    In my opinion, creating a strategy is a very important moment for a trader, which can have very important results.