[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [amibroker] AmiBroker 4.50.10 maintenance update



PureBytes Links

Trading Reference Links




<FONT face="Times New Roman" 
color=#000000 size=3>The number of parameters in a trading system: Looking for a 
Balance?  Questions to the board:  1.  Each 
parameter introduced into a system represents a measure of  control, 
which limits the reliability of the outcome. The more control  that is 
introduced into a trading system, the less reliable the will be.  In a 
statistical term, you are loosing degrees of freedom. So the more  
variables added like stops, logic filters and indicators (RSI, 
MACD,etc)  the more control you have asked and the less reliable will 
be the result.  This is true in general because it depends in the 
cross/joint correlation  between independent variables (ie:indicators). 
A well designed experiment  will check for possible joint correlation 
among independent variables.     In conclusion, the best 
system approaches are the ones which use a     very low 
number of variables (2 to 5), if possible. The more 
control     parameters the more likely a pair may be 
highly correlated.  a.  Is the above statement 
correct?  b.  Does logic filters (buy  if ..., exit if...) 
limits the reliability  of the outcome?  c.   If 
yes, how can I calculate the joint correlation with a RSI and a  logic 
filter?
<FONT face="Times New Roman" 
color=#000000 size=3>   <FONT face=Batang color=#0000ff 
size=2>d. <FONT face="Times New Roman" 
color=#000000 size=3> what it might mean if you can get good results with a 
lot of variables, which is contrary to the popular 
wisdom?  2.  Designing a neural 
network  to have a minimal degrees of freedom  (DOF) is critical 
when attempting to forecast financial markets. If a  neural network has 
too many, the model may produce excellent training  results but could 
generate unreliable forecasts when presented with  unseen 
data.  a. Is the above statement correct?  3.  
How much data to consider for Optimizing parameters? I think it  should 
be a balance between long ad short period:      i) 
If you take a small period you might leave a side certain 
market      characteristics and the optimization 
might not hold in the future.      ii) if you take 
a long period you might overfit and little data 
left      to walk forward testing.  With 
that said I would only reoptimize the system in real trading if  
results are not good let´s say in the last 4 weeks or the DrawDown  
reaches a certain amount.  This is a reactive trading system not a 
continuously optimized system.  a. What do you think of my 
reasoning?  b. Any comments?


Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx 
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html





Yahoo! Groups Links
To visit your group on the web, go to:http://groups.yahoo.com/group/amibroker/ 
To unsubscribe from this group, send an email to:amibroker-unsubscribe@xxxxxxxxxxxxxxx 
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.