Strategies back test not favorable anymore
Forums › ProRealTime English forum › ProOrder support › Strategies back test not favorable anymore
- This topic has 5 replies, 4 voices, and was last updated 7 years ago by
Louwrens.
-
-
03/06/2018 at 1:16 PM #64577
Hi
I am struggling quite a lot to develop a strategy for the SAF40 that trades a few times a day. I only want to be in the market for a few hours a day at most. I like to take small profits for now as I dont have lots of capital to risk. I have build many strategies that looked great when back tested. I then ran a few of them with some success. One of them ate 60% of my capital and then i stopped them to re-asses them. For some or other reason the back test does not look that great anymore. I dont know why it changed but if it looked the way it does now i would never have started them. I know I should test them in a demo account rather than live, but I am eager to trade.
Are there any short term strategies for the SAF40 already tested, if not what would be the indicator to use to develop one. I tried all the popular ones. I even bought a strategy from a company and tried to code it in, but without any success.
03/06/2018 at 1:47 PM #64580Hi,
I think we all try to do something like you. Some tips what I painfully learn during the process:
– use mini contracts for testing your algorithm live.
– set your maximum lost per trade
– If thinks are very bad, use the function “quit” inside the coding for deactivate the system
– if you use multiple variables and combination, then your program is prone to be overoptimize
-if you lost get the most of it and keep programming other scenarios for improve your favorite system
And do not give up. This is what keep me going on: if we make one system , at least one system! than win X euro per week then you can use it in different markets and time frames then you have X€ times (number of markets) times (number of time frames) …mmmmm yummi yumi!
1 user thanked author for this post.
03/06/2018 at 1:59 PM #64581Hi Leo
I think this is probably my problem. I am trying to make the system so save that at some point it does not work anymore or trades only once a month. Why the back testing looks different now than it did a week ago is still very strange.
I quiet enjoy building these systems and testing them, so will keep on going. I am just stuck at the moment. Not sure how to approach and new system as it feels like I tried all the combinations of technical indicators. Oh and a lack of coding knowledge also affects this.
if you use multiple variables and combination, then your program is prone to be overoptimize –
03/06/2018 at 2:05 PM #64582Why the back testing looks different now than it did a week ago is still very strange.
Use tick by tick mode (always) and set correctly the spread while backtesting your strategies.
03/06/2018 at 2:16 PM #64585Hi Louwrens
Don’t be afraid to look at the general Market Structure surrounding the trade entry when it executes.
By that I mean if a trade has opened Long and the bars around the entry and higher timeframe bars are screaming Short then stop the Strategy, check / improve code, run in Demo and live to fight another day??
If you can’t read Market Structure then use the info on the link below.
Cheers
GraHalhttps://www.investing.com/technical/indices-technical-summary
03/06/2018 at 9:38 PM #64624 -
AuthorPosts
Find exclusive trading pro-tools on