Index futures automation

Discussion in 'Journals' started by hilmy83, Jul 3, 2022.

Can a fully automated trading strategy work in the long run?

  1. YES!

    56 vote(s)
    66.7%
  2. Hell naw.

    15 vote(s)
    17.9%
  3. I don't know, I got my own trading to worry about.

    13 vote(s)
    15.5%
  1. hilmy83

    hilmy83

    Decide to run the current bot per above parameters with 2 contracts. Starting point around 52k. Let's see how this thing does week to week.

    Drawdown might get a bit steep from time to time. No more "make money every day" mentality, need to focus on long term growth.
     
    #1021     Jun 18, 2024
    SimpleMeLike likes this.
  2. Good Morning hilmy83,

    I can only speak for myself when I say this but your strategy/algo makes money every year since 2006 until present year.

    If I were you, I would run this system, and not do a damn thing else in life regarding trading. That's for damn sure.

    Maybe do some gambling clicking the charts manual on the side. But your algo looks good to me.

    I build algos as well in the NQ and ES markets, and the truth is man, we can only bet on 1 or 2 algos per our capital.

    I know all this diversification bullshit sounds so good, but the harsh reality is , if you do not have +$500,000 diversification means nothing.

    So you doing good man. Bet on your NQ algo and hope for the best.

    I have 1 ES algos and 1 NQ algo, both been in drawdown since 2023, I am down about -$30K combined with them. What the fuck can I do, fuck it. Got to roll with it.
     
    Last edited: Jun 18, 2024
    #1022     Jun 18, 2024
    hilmy83 likes this.
  3. hilmy83

    hilmy83

    Check some backtests for the recent years, saw >50% dd. Freaking same strategy can just fail even after +10 years.

    I might as well go my traditional route, make money most days and have low probility of 30-40% dd.

    2023

    upload_2024-6-18_17-37-21.png
    upload_2024-6-18_17-37-36.png



    upload_2024-6-18_17-35-21.png

    upload_2024-6-18_17-35-38.png
     
    #1023     Jun 18, 2024
    SimpleMeLike likes this.
  4. hilmy83

    hilmy83

    2022
    Apr-Nov
    upload_2024-6-18_17-40-8.png

    yea fuck this lol
     
    #1024     Jun 18, 2024
  5. Hello hilmy83,

    Exactly, my NQ/ES strategy made money every year like yours, and been in drawdown since 1/1/2023 and still in DD. And I started running in 1/1/2023. Not sure what happened. Big profit targets and big stops.

    Hopefully your algo does not enter DD when you start it or your money will be tied up for a long time. It is best to start algo while in DD so odds of making money is higher coming out of DD.
     
    #1025     Jun 18, 2024
  6. NoahA

    NoahA

    What kinds of trades is it putting on when it losses trade after trade? With your algo that won every day, it was easy to understand that on some days, the contracts would pile up with all the add-ons and reversals, so it was only a matter of time until there was a loss on a big position.

    But looking at the equity curve you show here, it seems like it takes weeks to hit the huge drawdown, so this means trade after trade doesn't work. Are you able to look at all of these trades on a chart to see what is happening? You seem to have fairly equal amount of longs and shorts, so its not like you're just shorting in a bull market or going long in a bear market, and the market changed.

    Right now the guys on the ES thread are killing it, and all they are doing is waiting for a dip and going long. Seems super easy. But at some point, every dip that they buy will probably not recover. So is your algo doing something similar in that it can work for 3 months and then all of a sudden completely fail?

    Maybe the ES traders will learn to short every rally and continue the winning streak, but it will likely take a while before they switch tactics, and this change will probably not happen until there are at least 5 or 10 days where dip buying doesn't work. So it can take a while for the dip buying euphoria to subside.
     
    #1026     Jun 18, 2024
    ironchef and rb7 like this.
  7. hilmy83

    hilmy83

    Honestly I think that backtesting with 10 years of data is worthless. Early 2000s is not the same as mid 2000's, and it's not the same as the last 5 years. Participants are different, tech is differnt, volalitily is different.

    But I do believe ORB does have a small edge. The trick is to minimize the worst case DD.
     
    #1027     Jun 19, 2024
    SimpleMeLike and NoahA like this.
  8. hilmy83

    hilmy83

    I'm leaning towards dax over nq again..

    With NQ, there's much shorter period for price moves from cash open vs the DAX; where you have the euro and US session overlap.

    Dax has more consistent volatility throughout the session based on my observation.
     
    #1028     Jun 20, 2024
    SimpleMeLike likes this.
  9. Hello hilmy83,

    Why not run your algo on NQ and DAX?

    I assume they both make money in back test right?
     
    #1029     Jun 20, 2024
  10. hilmy83

    hilmy83

    I'll let you answer that question, why do you think I rather run on one instead of spreading it out on multiple markets?
     
    #1030     Jun 20, 2024
    SimpleMeLike likes this.