admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
New Exits
Found some simple new exits today.
This is shown on my benchmark gsbsys2.1es with no stops at all.
Hope to add to GSB soon
| Code: | if currentcontracts > 0 then beginIf decision =-1 and sfDecision <1 then sell {short} ("sx**") this bar on close;
if decision =1 and sfDecision>-1 then buy to cover ("lx**") this bar on close;
end; |
For long, If secondary filter is no longer a buy, and the oscilators are a sell then exit long.
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
coming in build 50.89 in the next few days in the beta forum
|
|
|
coccigelus
Junior Member

Posts: 73
Registered: 11-7-2018
Member Is Offline
Mood: No Mood
|
|
That is a very notable improvement! Would be included the setting to remove the filter in the exit as well?
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
We talked at length on this, and i dont remember all the details of what you wanted.
All the options can be enabled or disabled.
But we still have the option to go from long trade, to short if indicators and sf say go short. This tends to almost never happen day trading.
Is this ok to you?
Reversal could also be stopped by limiting trades per day to 1
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Here is the results of the exits
es 30 minute, verify on 25 and 35
build from 1.1.2000 to 6.30.2015
results shown are out of sample to dec 11 2018
In summary exit 2 and exit 5 work well.
Significant increase in performance metrics, and decrease in drawdown.
Note verification is still helping. All metrics improved at the mild cost of number of trades.
But verification helped less. This is a good thing that I think is due to more effective exits.
What surprises me is new exits gave less trades, not more trades.
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
This is with the new exits. 50.80 version. Very aggressive exit number 6. There is no stop. This is just a system pulled at random.
Note the tiny losses. Such an aggressive stop will reduce overall pl I would expect.
System degrades a lot when the stop is removed
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Note the new exits. They work well but using them all works badly, plus is much slower to build systems.
|
|
|
engtraderfx
Junior Member

Posts: 98
Registered: 15-10-2018
Member Is Offline
Mood: No Mood
|
|
I tried using the next exits, for clarity, the above suggests exit 2 & 5 worked well (2=IRevSFFlat & 5=IFlatSFFlat), the numbering seems to have
changed in final release, is that equivalent to final exits 2 (IRevAndSFFlatRev) & 1 (IFlatRevAndSFFlatRev)?
Seeing variable results so far, some systems only see a few exits using these options which indicates limited improvement but some have a lot of
exits. Was wary of curve fitting by adding more variables to simple MOC exit, are we relying on WF to sort this out?
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by engtraderfx  | I tried using the next exits, for clarity, the above suggests exit 2 & 5 worked well (2=IRevSFFlat & 5=IFlatSFFlat), the numbering seems to have
changed in final release, is that equivalent to final exits 2 (IRevAndSFFlatRev) & 1 (IFlatRevAndSFFlatRev)?
Seeing variable results so far, some systems only see a few exits using these options which indicates limited improvement but some have a lot of
exits. Was wary of curve fitting by adding more variables to simple MOC exit, are we relying on WF to sort this out? |
51.x vs 50.89.
exit 2 is same as exit 5
exit 3 as exit 2
exit 5 similar but not the same as exit3,
others all different
IM not too worried about curve fit. The indicator result and or the secondary filter are both valid points entry and exit.
Things like exit2 where SF and indicators both are flat or reverse - seem common sense for a valid exit.
|
|
|
engtraderfx
Junior Member

Posts: 98
Registered: 15-10-2018
Member Is Offline
Mood: No Mood
|
|
Hi Peter, I am seeing systems with exit no's higher eg 12 or 13? I thought I had searched with 'none' only enabled. Doesn't seem to do anything tho.
|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by engtraderfx  | Hi Peter, I am seeing systems with exit no's higher eg 12 or 13? I thought I had searched with 'none' only enabled. Doesn't seem to do anything tho.
|
I dont know where you got that screen shot from, or the number.
Can you show the settings like I have in my screen shot.

|
|
|
admin
Super Administrator
       
Posts: 5069
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by engtraderfx  | Hi Peter, I am seeing systems with exit no's higher eg 12 or 13? I thought I had searched with 'none' only enabled. Doesn't seem to do anything tho.
|
I dont know where you got that screen shot from, or the number.
Can you show the settings like I have in my screen shot.
|
|
|