Collaborating with a programmer to create trading tools/strategies - Have anyone done this?

Discussion in 'App Development' started by Howard, Nov 23, 2019.


  1. can you say a few words what work a programmer needs to do ? for example, if you only use indicators, that is easier to code; but if you use candle context, it is extremely hard to code; or if you use trendline, it is also pretty hard to code identical to your naked eye since trendline is very subjective.
     
    Last edited: Nov 23, 2019
    #11     Nov 23, 2019
    Howard likes this.
  2. ET180

    ET180

    Even if you have it done by a professional, if the software provides a true edge and they realize it, then I would expect whoever you have coding it to steal your strategy. Sure, they will develop what you ask, but I would expect them to take the strategy. Although maybe it won't matter. Maybe hire a programmer from another country who would be less likely to use the strategy on the same market.

    Do you know any family members or friends who you trust and could work on it?
     
    #12     Nov 24, 2019
    Howard likes this.
  3. Howard

    Howard

    I use NinjaTrader for charting, execution and a bunch of my custom coded indicators.

    I use custom software because - to my knowledge - there does not exist anything off the shelf that offers what my custom software do for me.
     
    #13     Nov 24, 2019
  4. Howard

    Howard

    No indicators. Not candles per se, although I guess part of the logic I use could be perceived as candles. Simply put, it's an application which analyzes the market according to my own criteria/logic and makes this information available to me.

    Not sure if I can say anything more than that without revealing more than I want to.

    The application is already written and working.

    Yes. You're right.

    The last programmer I had to help me on optimizing a part of the application turned out to actually be a trader. Spooked me quite a bit. But I down played the entire thing and said it was just a hobby for me. I think the software provides an edge, but it's not something you just automate or which generates automated signals, although that can possibly be the outcome for a future version. It requires discretion to use.

    The company who first coded it are from Eastern Europe and have zero interest in the stock market. I can just tell from the way we communicate that they have no idea what I'm doing. That's why I stayed with them for so long, although I was unhappy about the results on subsequent updates.

    Seems like it's a tricky proposition to do this properly either way.
     
    #14     Nov 24, 2019
    Real Money likes this.
  5. qlai

    qlai

    I doubt this very much. It always puzzled me why people are not concerned with their IP being stolen overseas where there is no way to legally go after them. Cost is valid reason, but other than that. But I think very few people have anything really worth stealing that is non discretionary. Or the stuff that may be worth stealing may not be easily put to practice.

    I agree with you that if a developer has no interest in trading (very unlikely after crypto craze), there's usually a way to ask him/her to implement it in a way that is too vague to put it into practice.
     
    #15     Nov 24, 2019
  6. Howard

    Howard

    Doubt what?

    I've communicated a ton with these people. Trust me - they don't know what I'm doing, nor do they have any interest in the stock market.

    Trust me, I am concerned about my ideas being stolen, but what options do I have, really? They did sign an NDA of course, although I have no way to know if that's being honored.

    Probably true.

    Also true.

    What I have is certainly worth stealing, IMO. At least I feel it is since it's of value to me. But it's my logic, my (pseudo) code and my ideas. If some stranger who don't know anything about the markets (and even then) would come across it, he wouldn't know how to use. At least not simply out of the box. It would strike him as very complex and confusing...
     
    #16     Nov 24, 2019
    qlai likes this.
  7. SteveH

    SteveH

    Howard,

    Why don't you ask for help here in the programming aspects of what you want to do without giving away your algorithm and do the implementation changes yourself?

    C# programming is pretty easy. You're probably just caught up in using NT's library of function calls and maybe how to do certain comparisons. Just misdirect the comparisons you want to make. That is, pick common indicators not related to your algorithm and ask about their manipulations. Then, you apply the pattern of what's being done to your algorithm.

    Your software implementation sounds mature already. You can make the changes yourself. Whenever you make a change, just make sure you back-up the prior version which worked the way you wanted it to.
     
    #17     Nov 24, 2019
    Howard, ET180 and ironchef like this.
  8. ironchef

    ironchef

    Agree.

    Also, can't you split your method/program into subroutines, functions each only does a small part of the computation? Since you know how to program, you write the overall overlay that link all of it together so no one knows what exactly what you are doing?
     
    #18     Nov 24, 2019
    Howard and SteveH like this.
  9. ET180

    ET180

    Honestly, it might not be a bad idea to simply learn how to code in C#. It's a useful skill and will enable you to quickly try your own ideas. Honestly, only about 2% of the ideas that I have tried turned out to be useful. I've written slightly over 10 MB of java source code over the past decade. Many ideas that sounded good, but didn't work.
     
    #19     Nov 26, 2019
    Howard and SteveH like this.
  10. 10mb data, how many line of codes?
     
    #20     Nov 26, 2019
    Howard likes this.