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

[amibroker] FW: Optimize/OverOptimize



PureBytes Links

Trading Reference Links




<SPAN 
class=125061001-09102003>[other reply to my questions, from Dave Chamness. - 
dave merrill]
<SPAN 
class=125061001-09102003> 
<FONT face=Tahoma 
size=2>-----Original Message-----From: David 
ChamnessSubject: Re: Optimize/OverOptimize
It's OK to post my replies.
 
Equities may trade based on a specialist, or a 
small group of frequent traders, in effect market makers.  Change the group 
and price behavior may change.  Try to find a system for GE.  You 
may have better luck with a smaller stock.
 
Commodities are traded by many people who do not 
need to make a profit, such as hedgers and governments.  Currencies and 
interest rates trend because Alan Greenspan does not want to look like an idiot 
jacking rates up and down in a random walk.  So he lowers interest rates 
repeatedly until he is done.
 
Personally, I trade commodities, but I keep 
searching for stock systems.
 
Dave
<BLOCKQUOTE dir=ltr 
>
  ----- Original Message ----- 
  <DIV 
  >From: <SPAN 
  class=125061001-09102003> dave 
  merrill 
  Subject: RE: Optimize/OverOptimize
  
  <SPAN 
  class=968300905-01102003>thanks for clarifying, much appreciated. is it ok w 
  you if I forward your reply(ies)to the AmiBroker group where steve posted your 
  original? let me know.
  <SPAN 
  class=968300905-01102003> 
  <SPAN 
  class=968300905-01102003>one area intrigues me still: 
  <SPAN 
  class=968300905-01102003> 
  <SPAN 
  class=968300905-01102003>if we do find a market where a simple rule set works 
  well, why would you think that's so? because of some inherent property of 
  the stock itself that makes it non-random, different from other issues where 
  that rule fails? or is it another random walk phenomenon, unlikely to persist 
  at all? if that's so, it seems completely pointless to trade equities at all, 
  no different from gambling.
  <SPAN 
  class=968300905-01102003> 
  <SPAN 
  class=968300905-01102003>why do you think commodities act differently? because 
  prices respond more to real-world changes (supply/demand and factors that 
  influence it, etc) than to the raw emotionality that seems to drive equities? 
  if so, that implies we should look to fundamentals for more non-random trends 
  in equities, but not much in that dimension except news spikes seems to drive 
  valuation very much. <FONT face="Courier New" color=#0000ff 
  size=2>how do we resolve this apparent lack of 
  perceivable order, other than trading commodities instead?
  <SPAN 
  class=968300905-01102003> 
  <SPAN 
  class=968300905-01102003>thanks again for your thoughts, very 
  interesting.
  <SPAN 
  class=968300905-01102003> 
  <SPAN 
  class=968300905-01102003>Dave Merrill
  <SPAN 
  class=968300905-01102003> 
  <BLOCKQUOTE dir=ltr 
  >
    <FONT face=Arial 
    color=navy size=2><SPAN 
    ><FONT 
    color=#000000>Answers are in the text below.  Contrary to Steve's 
    statement, I have only one degree, BS Mechanical 
    Engineering.
    <SPAN 
    > 
    <SPAN 
    >Dave 
    Chamness
    <SPAN 
    > 
    <SPAN 
    >-----Original 
    Message-----From: Dave 
    Merrill [mailto:dmerrill@xxxxxxx] <SPAN 
    >Sent: <st1:date Month="9" 
    Day="29" Year="2003"><SPAN 
    >Monday, September 29, 
    2003<SPAN 
    > <st1:time 
    Hour="12" Minute="32"><SPAN 
    >12:32 
    PM<SPAN 
    ><SPAN 
    >To: <A 
    href="">dec@xxxxxxxx<SPAN 
    >Subject: 
    Optimize/OverOptimize
    <SPAN 
    > 
    
    <SPAN 
    >Dave, I 
    hope it's ok to contact you on this. steve karnish posted a presentation of 
    yours on optimization that I found very interesting, though I'm 
    afraid I don't get all of it. this is a topic I'm thinking about pretty 
    much constantly these days, with quite a bit of accompanying frustration. 
    IMVHO, most of the world gives way too much weight to optimizations that 
    seem like curve fitting to me, but I haven't figured out how to move beyond 
    that.
    
    <SPAN 
    > 
    
    <SPAN 
    >a couple of 
    questions, if I might:
    
    <SPAN 
    > 
    
    <SPAN 
    >- can you 
    explain the scatter plots on slides 3 and 4? what exactly is plotted on x 
    and y? the punch line, which I'm too ignorant to see, is that the system 
    fails with out of sample data. the one part I understand, I think, is that 
    the correlation coefficient, presumably between in and out of sample 
    results, is poor. is that right? how does the plot itself show 
    this?
    <SPAN 
    > 
    <SPAN 
    ><FONT 
    color=#000000>They show the In-Sample gain as % of perfect trading on the x 
    axis versus the out of sample gain on the y axis.  Each data point is a 
    separate stock with a separate system.  In sample gains were 15% of 
    perfect on average.  Out of sample were near zero on average.  
    Perfect trading wins all close to close changes.  There are 2 years in 
    and out of sample.
    
    <SPAN 
    > 
    
    <SPAN 
    >- slide 24 
    mentions "Trend Following on Commodities", as "100 day lookback, trade 34% 
    before breakout". I don't understand what this means. something about MA or 
    EMA(100), maybe, but what's the 34% piece? how does it get around the 
    parameter settings limitations that sink other systems? is this method, or 
    something based on related principles, tradeable in stocks and/or 
    mutual funds?
    <SPAN 
    > 
    <SPAN 
    ><FONT 
    color=#000000>Breakout buys a new high, sells a new low.  Near Breakout 
    trades sooner.  34% before breakout buys in the top third of the 100 
    day high-low range, sells in the bottom third.  Specifically, the 34% 
    means 34% of the high-low range.
    
    <SPAN 
    > 
    <SPAN 
    >- how would 
    I compute the daily standard deviation of the S&P500, in AmiBroker for 
    instance, in a way that gives the same .95%/day figure you 
    mention? is that the average std dev of daily close price change over 
    some specific period of time? I ask so I can generate comparable 
    figures for other markets.<SPAN 
    class=968300905-01102003> 
    <SPAN 
    ><SPAN 
    class=968300905-01102003> 
    <SPAN 
    ><FONT 
    color=#000000>Compute the standard deviation of all the close to close 
    changes. 
    <SPAN 
    ><SPAN 
    class=968300905-01102003> 
    
    <SPAN 
    >
    
    <SPAN 
    >- the 
    parameters I get optimizing today compensate for transient market behaviors 
    that will eventually end, and eventually it will do very poorly. but if 
    those behaviors persist, at least somewhat, for a little while, might 
    the system to do better than average in the short term? if so, is constant 
    re-optimization worth exploring, or even switching whole trading systems in 
    a mechanical way based on recent performance?
    <SPAN 
    > 
    <SPAN 
    ><FONT 
    color=#000000>I find little tendency for trading systems to work in the 
    future.  Try to identify a simple nonrandomness.  Try to find 
    markets that simple systems work on.  Don't pick an impossible market 
    like S&P 500 and try to fit a complex bunch of rules to 
    it.
    <SPAN 
    > 
    <SPAN 
    ><FONT 
    color=#000000>Commodities have long term trends.  Stocks show short 
    term 2-10 day reversals.   
    
    <SPAN 
    > 
    
    <SPAN 
    >thanks 
    again for writing and sharing this. makes me wish I lived somewhere 
    near the meetings you haunt...
    
    <SPAN 
    > 
    
    <SPAN 
    >dave
    <BLOCKQUOTE class=Section1 
    >
      <SPAN 
      >Dave is an Agilent, 
      triple-degreed, engineer.  Two weeks ago, he presented this work to 
      our Denver Trading Group's weekly meeting (actually, this group meets 
      every Thursday and most Saturday's).  Once a month, I 
      moderate a SIG on mechanical trading (and I haven't seen less 
      than eighty people in the room since I've been 
      attending).
      
      
      <SPAN 
      > 
      
      <SPAN 
      >Although, I don't agree with 
      certain aspects of his presentation and I somewhat object to his assigning 
      my name to the "Karnish System" (it has become a bastardized off-shot 
      of my work), I still believe that there is a lot of merit to aspects of 
      his work.  The "Karnish System" has become the moniker for systems 
      (along the front range of <FONT 
      face=Arial size=2><SPAN 
      >Colorado<FONT 
      face=Arial size=2><SPAN 
      >) that stochastically 
      smoothes a momentum oscillator that initiates buy and sell 
      signals using symmetrical triggers.  
      
      
      <SPAN 
      > 
      
      <SPAN 
      >I neither want to endorse, 
      defend or criticize Dave's work...but, offer this for group members to 
      stimulate thought.
      
      <SPAN 
      > 
      
      <SPAN 
      >Take 
      care,
      
      <SPAN 
      > 
      
      <SPAN 
      >Steve






Yahoo! Groups Sponsor












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.