John62
Junior Member
Posts: 23
Registered: 29-9-2017
Member Is Offline
Mood: No Mood
|
|
Exits
I find that GSB's focus is on entries. GSB's advice for testing is to use Market Close as an exit. I think there are better exits and would like to
test it. A good system consists of the combinations of good entries and exits. Is there any possibility in GSB to test the combination of entries and
exits (other as the stops build in)?
|
|
|
admin
Super Administrator
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Currently GSB is not optimizing exits. You can choose however the exit type.
ie moc, swing, stop and or profit target. After years of experience, I find exits are no where near as important as entries.
Not clear on "other as the stops built in"
I am open to doing this in the future, but thing there are more pressing features. ie stop & limit entries, more entry truisims (see the docs on this)
GSB is designed to minimise the end user doing things they think works well, but is actually harmful.
ie you could spend lots of genetic time optimizing exits, and the exit code in effect might do nothing and chew up a number of variables. Builder did
this sort of thing a lot. Things like days of week is deliberate to not be done genetically. Rather we use the approach of gurus like Andrea Unger.
(again in the docs- manually find the day of week market bias)
|
|
|
John62
Junior Member
Posts: 23
Registered: 29-9-2017
Member Is Offline
Mood: No Mood
|
|
What do you mean by swing? I do not see that option in GSB.
The exit does make a significant difference to win% and the P / L. I agree that you should not combine the entries in the same way as the entries. But
I would appreciate some simple logical stops. As a result, there are also more entries and possibly more opportunities a day. For example, I have good
experience in daytrading with the MA5. In combination with buying dips and limit orders.
Long post: If C> Average (C, 5) then Sell and vice versa for Short.
|
|
|
admin
Super Administrator
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
For swing, you just turn moc off.
Whats ma5? Tradestation symbol? I agree strongly that entry limits and stops
should be implemented. My guess is this is 1 to 3 months away.
Bottom line on exits is I'm open to the idea, and thinking about it.
I'm not sure for example if we let GSB build a system, then optimize exits.
Currently I can build systems from GSB, and live trade them the same day.
If I build exits that GSB thinks works, but they are not logically sound, or dont actually trade....
Ive then gone backwards in architecture.
You can always build a system, then later add the enhanced exit strategy.
Early GSB adopters had to put up with building an entry technique, then improving it so profits were high enough to be trade-able. Fortunately, those
days are long gone.
|
|
|
Carl
Member
Posts: 343
Registered: 10-5-2017
Member Is Offline
Mood: No Mood
|
|
I think ma5 means 5 bar moving average
|
|
|
admin
Super Administrator
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
average is in GSB already and the Genetic engine should fine the best combinations. Medium term we are going to have soft coded indicator range. That
way you can limit average to 5, for say 3 to 10 etc
|
|
|
John62
Junior Member
Posts: 23
Registered: 29-9-2017
Member Is Offline
Mood: No Mood
|
|
I mean Average as an exit, not an entry. Exist have a huge impact on your trading system. The result of MOC is always 1 trade a day. Especially if you
want more trade's a day or longer/swing trading. To show the effect in the attachment is a system with a MA5 exit and the same system with a MOC
close.
Attachment: Login to view the details
|
|
|
Carl
Member
Posts: 343
Registered: 10-5-2017
Member Is Offline
Mood: No Mood
|
|
Hi John62, i
In the pdf you posted October 14th, did you compare the two different exit methods on a GSB generated strategy or one of your own strategies? Thanks,
Carl
|
|
|
John62
Junior Member
Posts: 23
Registered: 29-9-2017
Member Is Offline
Mood: No Mood
|
|
One of my strategies, just to see the difference in exits.
|
|
|
John62
Junior Member
Posts: 23
Registered: 29-9-2017
Member Is Offline
Mood: No Mood
|
|
Hi Peter. You said: "Bottom line on exits is I'm open to the idea, and thinking about it." Did you make progress in thinking about exits?
Exits (and entries) also depend on your trading style (counter, trend, swing, long). At the moment GSB generates mainly daily trend signals (because
the stop is MOC). If you want a (intraday) countertrend or swing system you need other stops/logical exits. Will it be possible to generate systems
for a specific trading style?
|
|
|
admin
Super Administrator
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
GSB will make whatever style suits the market. For example secondary filter CloseD weight can be -1 or 1. This will invert the function.
GSB can and does generate longer term systems if MOC is false.
Additional exits are in the pipeline, not long after additional secondary filters.
|
|
|
![[*]](images/davis/default_icon.gif){kind=icon}
