Best Practices for System Development?

• 09/03/2016 at 8:04 PM #12639
  Wing

  Hello all.

  I am going to come up with a better work process for building my systems. Right now, I tinker and backtest a few systems at a time until I’m done and either let them rest or run them live.

  I want to write down a plan for the process of how to make these systems. Specifically I have been thinking about the order in which I implement the following into the systems: Choice of indicators/signals, timeframe, market, backtest and forward testing (10-30 % out-of-data testing), money management (position sizing), Monte Carlo testing, time restrictions (only trading certain weekdays and time intervals).

  Does anyone have a set order in which they tweak each of these factors? To avoid curve-fitting, which factors should be optimized for last (eg. optimizing technicals before time) ?

  // Wing

  

  Wing

  Participant

  Topics: 6

  Replies: 159

  Been thanked: 77 times
  09/04/2016 at 6:08 PM #12672
  Nicolas

  Hi,

  I would say that firstly, you have to decide that how much time do you want to stay in the market. With this simple question, you can quickly organize what type of indicators (or any other calculation methods) and what timeframe you’ll play with.

  Then the risk reward/ratio is the most important thing in my opinion. In fact, the important question is “how can I exit loosing trade the best way”, while let profit run in trend following system can easily manage losses, mean reverting one often need to catch little money with very tight stoploss.

  All statistical tests need to be done with same amount of contracts, if they are ok, money management could be added to modulate risk.

   

  1 user thanked author for this post.

  Wing

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
  09/07/2016 at 9:50 AM #12769
  Stef

  Try this…

  5 users thanked author for this post.

  victormork, David, tapsomepips, Wing, Nicolas

  

  Stef

  Participant

  Topics: 20

  Replies: 52

  Been thanked: 15 times
  10/24/2016 at 3:44 AM #15386
  tapsomepips

  Hi Wing,

  Just looked at the pdf from Stef.  I’m sure it will take me a couple of days to digest!  But awesome all the same so thanks Stef.

  The issue with this game is the almost unlimited number of variables.

  To realise the holy grail will mean being available for every second the markets open until they close.

  What I have found useful is defining my availability to the market.

  Opportunities are available everywhere and at every time of day and night, alas we are not.

  Just because you present yourself to the market doesn’t mean that there is an opportunity there.

  I have tailored my personal opportunity window to the daily/weekly Euro session and look for opportunity within that.

  Longer periods require deeper pockets.

  I think it’s more about deciding the level of time commitment you are comfortable making to the market and then defining your strategy within that.

  Without your personal parameters, the market will always run you ragged.

  Sorry if it’s a bit off topic.

   

   

  

  tapsomepips

  Participant

  Topics: 3

  Replies: 7

  Been thanked: 1 time
• Author
  Posts