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

RE: [amibroker] Invitation -- suggest your favorite objective function



PureBytes Links

Trading Reference Links




I 
don't know for sure, but I think the point about objective function was missed 
completely.    If you are looking at a 3-D graph, what is it 
graphing?    Net profit, CAR, CAR/MDD?   Whatever it is 
graphing is your objective function.   Your objective function could 
be a formula with six to eight metrics represented.   Most of the ones 
I use have more than four.    So, using a 3-D graph and looking 
for broad, flat areas is a great idea as long as you have decided what it is 
that you want to graph.
<BLOCKQUOTE 
>
  <FONT face="Times New Roman" 
  size=2>-----Original Message-----From: advenosa@xxxxxxxxxxxx 
  [mailto:advenosa@xxxxxxxxxxxx]Sent: Wednesday, October 22, 2003 
  8:30 AMTo: amibroker@xxxxxxxxxxxxxxxSubject: RE: 
  [amibroker] Invitation -- suggest your favorite objective 
  functionThere might very well be, but I don't believe 
  you need that to pick yourparameter values. If the peak area of the 3-D 
  graph is flat and broadenough, and you see no local maxima that could have 
  been due to chance,then any parameter you select within that broad peak 
  will do just as wellas any other over time. That is the essence of 
  robustness, and your systemought to be able to respond to small market 
  changes without having to beconstantly re-optimized. This is not to say 
  that re-optimization is not anappropriate thing to do occasionally (and I 
  don't know how to define theword "occasionally"). Al 
  VenosaOriginal Message:-----------------From: Dave Merrill 
  dmerrill@xxxxxxxDate: Wed, 22 Oct 2003 07:27:03 -0400To: 
  amibroker@xxxxxxxxxxxxxxxSubject: RE: [amibroker] Invitation -- suggest 
  your favorite objectivefunctionI understand what you're 
  saying, but doesn't it seem like there might wellbe some useful numerical 
  measure of the flatness of the 3D graph, forexample?dave  
  I think there is a human component that has to be included in 
  systemdevelopment. I am a believer in mechanical systems from the 
  standpoint oftrade execution. However, when you are trying to develop your 
  mechanicalsystem and to decide how best to implement it, you have to make 
  somejudgments that I don't think can easily be automated into a 
  single,multidimensional matrix out of which comes a simple, black and 
  white answer.There will always be shades of gray that require human 
  judgment. That's whyI sent along that figure (just an example). You have 
  to develop your owndecision rules that are right for your personality, and 
  those require you toexamine a multitude of objective functions to aid in 
  your decision. There isa software package called Powerkit, wherein the 
  developer has included 3canned systems that the user can tweak with any 
  combination of parametervalues he wants to optimize. When the optimization 
  is done, a 3-dimensionalgraph automatically appears showing how the 
  parameters affect profits in3-dimensional space. You pick your parameter 
  values based on the flatness ofthe 3-D graph, which is a measure of 
  robustness. TJ once told me that heplans to develop a similar 3-D 
  graphical capability that will blow Powerkitout of the water. That is 
  probably on the lower part of his priority listright now, but it will come 
  eventually.  Is that a sufficiently vague answer to your 
  question? :-))  Al Venosa    ----- Original 
  Message -----    From: Dave Merrill    
  To: amibroker@xxxxxxxxxxxxxxx    Sent: Tuesday, October 21, 
  2003 8:23 PM    Subject: RE: [amibroker] Invitation -- 
  suggest your favorite objectivefunction    
  that's how I've approached it in general too, eyeballing the variousresult 
  outputs manually. but if possible, I'd like to systematize it, partlyso it 
  can be an object comparison measure, and partly so it can become apart of 
  the trading system itself.    do you have any sense of 
  how that might be done? it doesn't seem thatthe example you gave boils 
  down to simply weighting and summing the variousindividual 
  metrics.    any other ideas, anyone, on systematically 
  combining them some other waythat captures more of what we do when we 
  eyeball?    dave      I 
  think there are numerous ways to evaluate your system rather thanrelying 
  on one objective function. For example, the other day, I was doing 
  aportfolio optimization to find out the optimum no. of positions I 
  shouldhave open, given a certain amount of starting equity. I then plotted 
  theresults of a few of the output parameters TJ now gives us in an 
  optimizationoutput. See attached png file. If you were simply using one 
  objectivefunction (CAR) without considering anything else, you would 
  choose 5 openpositions because that's where CAR max'd out. However, MAR 
  (=CAR/MDD) wasstill rising as was UPI (you want that to be high), while UI 
  (you want thatto be low) was still falling and RAR/MDD was still not max'd 
  out yet. So,when you look at 7 open positions, you see that UI had pretty 
  much bottomed,UPI, RAR/MDD, and MAR had finally topped. So, in this 
  particular case, theoptimum no. of open positions would be selected to be 
  7, in my opinion,rather than 5. So, using this tiny little example, I 
  think you need to lookat several different objective functions and decide 
  on the best strategybased on the total story they all tell. I also like to 
  use expectancy, whichI think is an extremely important and useful 
  objective function. Sorry forthe multiple answer, but I don't think there 
  is only one criterion uponwhich to judge a trading 
  system.      
  Al--------------------------------------------------------------------mail2web 
  - Check your email from the web at<A 
  href="">http://mail2web.com/ 
  .Send BUG REPORTS to bugs@xxxxxxxxxxxxxSend 
  SUGGESTIONS to 
  suggest@xxxxxxxxxxxxx-----------------------------------------Post 
  AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx (Web page: <A 
  href="">http://groups.yahoo.com/group/amiquote/messages/)--------------------------------------------Check 
  group FAQ at: <A 
  href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html 
  Your use of Yahoo! Groups is subject to the <A 
  href="">Yahoo! Terms of Service. 







Yahoo! Groups Sponsor


  ADVERTISEMENT 









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



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.