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

[amibroker] Re: FW: Optimize/OverOptimize



PureBytes Links

Trading Reference Links

Steve,

No apologies necessary. I'm glad that we can have an open discussion, 
and that disagreements (no matter how poorly worded, perhaps the way 
I did) are not taken as personal attacks.

There's one disconnect that I find between the way you're viewing the 
world (as a professional trader) and the way I see it (as someone who 
does something else full-time, software in my case, and tries to 
enter the active trading world in spare time). Vast difference. In my 
kinda world, yes, people DO start by reading couple of books (I just 
used Dr. Elder's as an example) and/or by attending couple of 
seminars. BTW, I'd in fact recommend Dr. Elder's book to novice 
traders, but more for a candid discussion of psychology of trader and 
for basic understanding of crucial concepts like position sizing, 
discipline, etc. than the trading method itself. What I was trying to 
say is that people like myself buy the book thinking that its going 
to give them a magical formula that'll make them money. Worse yet, 
they implement it, find that it doesn't work, and then turn to 
optimization to "make it work". 

> Are you tweaking multiple variables inside your algorhythm or are 

I try not to tweak much. By sound strategy, I was trying to convince 
people that it is more important to "find a sound strategy and tweak 
it a little bit" rather than to "try a strategy and get it to work 
via optimization over past data". Most novice traders do the latter. 
MACD crossover is a great example. :-) Another example/trap is that 
people get over obsessed with trying to make a strategy based on 
single ticker's price/volume than to come up with one that also looks 
at the market conditions. This is especially true of fresh newbie's 
chasing the holy grail. I myself have been there a while ago. :-)

> *We have heard.  Most of the great stock/commodity traders are not 
interested in a high profile. John Henry owns the Boston Red 

Point taken.

> Direct all 
> flames at Dave. :-)
> 
> *Flames don't bother Dave, the Ft. Collins group will pour beer on 
him.  

Not sure which Dave you are referring to, since I think the person 
who gave the presentation is also Dave? I was referring to Dave who 
posted this, not to mention I was kidding. :-)

Thanks.

Jitu

--- In amibroker@xxxxxxxxxxxxxxx, "CedarCreekTrading" <kernish@xxxx> 
wrote:
> Jitu,
> 
> I apologize if I offended you.  I pulled up your orginal email and 
I'll try to be more specific with my points:
> 
> A typical wanna-be trader reads Dr. Elder's book. 
> 
> *Maybe, but, I don't think that's the typical path.  
> 
> Then he/she tries to implement it. 
> 
> *Sounds dirty or dangerous.
> 
> Of course, Dr. Elder, like many other "experts", never offers a 
completely working system,although the book offers, like many other 
books, good concepts well 
> understood by practically reading any other trading related book if 
> you have even half the brains of what it'd take for you to survive 
in 
> this jungle. Instead, just concepts. Now here, substitute YOUR 
> favorite book for Dr. Elder's book. Same thing applies.
> 
> *Amen Brother!!  This research and trading is bitch. I'm in 
that "half the brains" category and it's twice difficult for the half-
witted.
> 
> 
> So its upto me, an average person, to derive a profitable system 
from 
> his concepts. 
> 
> *I think you are on to something.   
> 
> Hmmm... What are the odds of that happening? After all, 
> HE had to resort to selling stuff to other traders to make a living 
> himself. Call me fanatically skeptical, but I find it funny how 
that 
> works.
> 
> *That Dr. Ehler must be some type of Jimmy Jones. Are you saying 
that there might be people who market useless garbage to the public?  
Please Jitu, don't drink the kool-aid.  
> 
> Now, you are thinking... What does this have to do with 
optimization? 
> It does, a lot. Because optimization, IMHO, is an attempt to make 
an  
> unproven logic work with the past data you have. 
> 
> *Some of us actually start with proven logic.  How many weeks, 
months or years does it take for logic to be "proven"?  There many 
traders and some products that can prove their logic and guess what?  
They've done it publically for long,long periods of time.  Maybe, 
you're turning over the wrong rocks.  "Looking for love/luck in all 
the wrong places".  People have things that work.  Many of these 
people share.  
> 
> Read that definition 
> again. That's all it is. At least, its a good thing that the base 
> concept is not "completely" random, if you read a book or two. But 
> nonetheless, its true. You start with a concept that you "think" 
will 
> work, then it doesn't, then you tweak it a little bit, it still 
> doesn't work, and then you finally turn to the software to tweak it 
> to death to make it work with the past data, a la optimization.
> 
> *I think you have a tweaking problem...I can refer you to a good 
doctor in Eagle County.  I just recovered from West Nile Virus.  That 
doesn't make me an expert on diseases, but, I believe you 
have "tweakitis".  The smarter they are...the harder they tweak.  
Some say:  "Don't tweak a little, because pretty soon you'll have to 
tweak every day!" I don't believe that part.  Every so often, I like 
to tweak-it-up.  
> 
> Are you tweaking multiple variables inside your algorhythm or are 
you optimizationing triggers, stops, targets, filters, qualifiers 
and/or other "goodies".   Remember the old saying:  "You can dress a 
pig in a prom dress, but in the end...it's still a pig".  If you 
start with a concept that doesn't work (across broad parameters) you 
have yourself a pig.
> 
> Think about it this way, if someone could become the greatest stock 
> trader just by optimizing a concept on the past data, wouldn't we 
> have heard about him/her by this time? That itself should answer 
your 
> original question.
> 
> *We have heard.  Most of the great stock/commodity traders are not 
interested in a high profile. John Henry owns the Boston Red 
Sox...so, I guess that is high profile.  Twenty years ago, he 
wouldn't even let me see what type of computers he was using.  I bet 
he is still very hush-hush.  It takes finding out who/what works.  
Finding out what works is work. 
> 
> IMHO, one should have a "sound" strategy in the first place. A 
sound 
> strategy should take into account ALL factors that the great 
traders 
> of the past have known to take into consideration. Bounce it around 
a 
> few people, if you're not sure that its sound. And THEN you can 
> optimize it a LITTLE BIT, and THAT is ok.
> 
> *As Keith Fitchen, the inventor of Aberration, likes to say and, 
I'm paraphrasing: "These book writers and system developers are 
drinking each others bath water".  I couldn't agree more.  I suggest 
that if your not sure that your approach is sound, that you don't 
even bother to "Bounce".  Optimizing is dope for system developers.  
Always remember to use drugs responsibly.
> 
> 
> All of this of course, is IMHO, a novice trader. 
> 
> I'm all for getting flamed by someone who is actually making tons 
of money just by  
> optimizing an original strategy that didn't work prior to the 
> optimization. 
> 
> *Why would someone want to start with a strategy that doesn't work?
> 
> 
> May be I'll learn something.
> 
> *That's the spirit!
> 
> 
> And if not ualready nderstood, no offense meant to ANYONE. Neither 
> people who optimize, nor people who are fans of Dr. Elder. 
> 
> *No offense taken.  Poor Dr. Elder.  First a Messiah, then scorned 
and flamed in forums across the world.  The shame of it all.  I'm 
trying to sell the story to Hollywood. Very sad.  
> 
> I'm in the "people who like to optimize" category...but, I'm mostly 
a recovering "person who used to like to optimize".  Once, I over-
optimized so hard...I almost died.   No kidding, it almost killed 
me.  So, obviously, you have to be really careful.   It can be 
deadly.  First, you do a little tweak here and a little tweak there 
and pretty soon you're a tweaking fool.  Suddenly, there's this ever 
growing need to "do a little tweaking"("on the 
tweak", "tweakin'", "doin' the tweak", "tweakarama", "tweaky 
deaky"...etc.). Bless me father, for I have tweaked. Welcome to 
tweaksville.  
> 
> Direct all 
> flames at Dave. :-)
> 
> *Flames don't bother Dave, the Ft. Collins group will pour beer on 
him.  
> 
> **Jitu, sorry I offended you.  Optimizing is a heavy subject.  I 
can understand your frustration.  Dave did a good job with the 
presentation and yes, he and people around him (using totally 
different approaches) make money trading mechanical systems.  These 
guys are "Crunch Monsters" and are powered up.  Guess what?  Dave 
gives a good optimizing seminar because he's also a 
recovering "tweaker".  
> 
> Jitu
> 
>   ----- Original Message ----- 
>   From: jtelang 
>   To: amibroker@xxxxxxxxxxxxxxx 
>   Sent: Wednesday, October 08, 2003 8:03 PM
>   Subject: [amibroker] Re: FW: Optimize/OverOptimize
> 
> 
>   Dave, if this thread goes on for ever, YOU will be the one to 
>   blame. :-) Here! I'll help you completely open the can of worms 
that 
>   you're attempting to open. :-P
> 
>   IMO, most, but not all, attempts to optimize are based on flawed 
>   logic. I'll present myself as an example, which should rhyme with 
>   most novice traders. A typical wanna-be trader reads Dr. Elder's 
>   book. Then he/she tries to implement it. Of course, Dr. Elder, 
like 
>   many other "experts", never offers a completely working system, 
>   although the book offers, like many other books, good concepts 
well 
>   understood by practically reading any other trading related book 
if 
>   you have even half the brains of what it'd take for you to 
survive in 
>   this jungle. Instead, just concepts. Now here, substitute YOUR 
>   favorite book for Dr. Elder's book. Same thing applies.
> 
>   So its upto me, an average person, to derive a profitable system 
from 
>   his concepts. Hmmm... What are the odds of that happening? After 
all, 
>   HE had to resort to selling stuff to other traders to make a 
living 
>   himself. Call me fanatically skeptical, but I find it funny how 
that 
>   works.
> 
>   Now, you are thinking... What does this have to do with 
optimization? 
>   It does, a lot. Because optimization, IMHO, is an attempt to make 
an  
>   unproven logic work with the past data you have. Read that 
definition 
>   again. That's all it is. At least, its a good thing that the base 
>   concept is not "completely" random, if you read a book or two. 
But 
>   nonetheless, its true. You start with a concept that you "think" 
will 
>   work, then it doesn't, then you tweak it a little bit, it still 
>   doesn't work, and then you finally turn to the software to tweak 
it 
>   to death to make it work with the past data, a la optimization.
> 
>   Think about it this way, if someone could become the greatest 
stock 
>   trader just by optimizing a concept on the past data, wouldn't we 
>   have heard about him/her by this time? That itself should answer 
your 
>   original question.
> 
>   IMHO, one should have a "sound" strategy in the first place. A 
sound 
>   strategy should take into account ALL factors that the great 
traders 
>   of the past have known to take into consideration. Bounce it 
around a 
>   few people, if you're not sure that its sound. And THEN you can 
>   optimize it a LITTLE BIT, and THAT is ok.
> 
>   All of this of course, is IMHO, a novice trader. I'm all for 
getting 
>   flamed by someone who is actually making tons of money just by  
>   optimizing an original strategy that didn't work prior to the 
>   optimization. May be I'll learn something.
> 
>   And if not ualready nderstood, no offense meant to ANYONE. 
Neither 
>   people who optimize, nor people who are fans of Dr. Elder. Direct 
all 
>   flames at Dave. :-)
> 
>   Jitu
> 
>   PS: A family emergency will prevent me from replying to this for 
>   couple of days, but I will follow it up as soon as I get a chance.
> 
>   --- In amibroker@xxxxxxxxxxxxxxx, "Dave Merrill" <dmerrill@xxxx> 
>   wrote:
>   > [other reply to my questions, from Dave Chamness. - dave 
merrill]
>   > 
>   > -----Original Message-----
>   > From: David Chamness
>   > Subject: 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
>   >   ----- Original Message -----
>   >   From:  dave merrill
>   >   Subject: RE: Optimize/OverOptimize
>   > 
>   > 
>   >   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.
>   > 
>   >   one area intrigues me still:
>   > 
>   >   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.
>   > 
>   >   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. how do we resolve this apparent lack of 
>   perceivable
>   > order, other than trading commodities instead?
>   > 
>   >   thanks again for your thoughts, very interesting.
>   > 
>   >   Dave Merrill
>   > 
>   >     Answers are in the text below.  Contrary to Steve's 
statement, 
>   I have
>   > only one degree, BS Mechanical Engineering.
>   > 
>   >     Dave Chamness
>   > 
>   >     -----Original Message-----
>   >     From: Dave Merrill [mailto:dmerrill@x...]
>   >     Sent: Monday, September 29, 2003 12:32 PM
>   >     To: dec@xxxx
>   >     Subject: Optimize/OverOptimize
>   > 
>   > 
>   > 
>   >     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.
>   > 
>   > 
>   > 
>   >     a couple of questions, if I might:
>   > 
>   > 
>   > 
>   >     - 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?
>   > 
>   > 
>   > 
>   >     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.
>   > 
>   > 
>   > 
>   >     - 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?
>   > 
>   > 
>   > 
>   >     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.
>   > 
>   > 
>   > 
>   >     - 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.
>   > 
>   > 
>   > 
>   >     Compute the standard deviation of all the close to close 
>   changes.
>   > 
>   > 
>   > 
>   > 
>   >     - 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?
>   > 
>   > 
>   > 
>   >     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.
>   > 
>   > 
>   > 
>   >     Commodities have long term trends.  Stocks show short term 
2-10 
>   day
>   > reversals.
>   > 
>   > 
>   > 
>   >     thanks again for writing and sharing this. makes me wish I 
lived
>   > somewhere near the meetings you haunt...
>   > 
>   > 
>   > 
>   >     dave
>   > 
>   >       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).
>   > 
>   > 
>   > 
>   >       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 Colorado) that 
>   stochastically
>   > smoothes a momentum oscillator that initiates buy and sell 
signals 
>   using
>   > symmetrical triggers.
>   > 
>   > 
>   > 
>   >       I neither want to endorse, defend or criticize Dave's 
>   work...but,
>   > offer this for group members to stimulate thought.
>   > 
>   > 
>   > 
>   >       Take care,
>   > 
>   > 
>   > 
>   >       Steve
> 
> 
>         Yahoo! Groups Sponsor 
>        
>        
> 
>   Send BUG REPORTS to bugs@xxxx
>   Send SUGGESTIONS to suggest@xxxx
>   -----------------------------------------
>   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.


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Rent DVDs from home.
Over 14,500 titles. Free Shipping
& No Late Fees. Try Netflix for FREE!
http://us.click.yahoo.com/mk9osC/hP.FAA/3jkFAA/GHeqlB/TM
---------------------------------------------------------------------~->

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 http://docs.yahoo.com/info/terms/