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

Re: [amibroker] Off Off Off Topic - Margin of Error



PureBytes Links

Trading Reference Links

 
I think you guys are splitting hairs..... since when did variance of  returns 
become stationary???....  there is always the possibility of the  next 
drawdown being the ruinous one.
 
Go read Ralph Vince.....  what does DD or variance mean, when the  there is 
always the possibility of a bigger DD looming....(ie. the scenario  that you 
never tested)....
 

Bottomline, stats and testing can give you some ideas of how one system may  
perform in relation to another system.....
 
Ultimately, I think you have to sleep at night, or the best traded system  
won't work...
 
For example, trade a reasonable system at the Optimal F, and see if you can  
sleep at night.
 
You also have to be ready to pull the plug on a trading system when you  
can't stand the DD. 
 
My thinking is quite different....
 
Comparing systems is more relevant...   the rest being money  management 
 
 
Kevin
 
 
 
 
 
 
 
In a message dated 11/8/2006 8:57:00 PM Central Standard Time,  
ftonetti@xxxxxxxxxxxxx writes:

It's not  about margin of error ... It's about the confidence level or 
lack thereof  for what the system will do with data it has not yet 
seen.

For a  simple OOS test I would probably split the data in half ... 
optimize the  system based on the first half of the data and then test 
the second half  to observe what the OOS performance metrics were ... 

For a more  thorough test, I would, as I outlined earlier, take the 
minimum amount of  data I could to generate what I felt were a 
sufficient number of trades  and then I would ...

- Optimize the system on that data
- Test some  period following the IS as OOS
- Roll the end date forward to the end of  the OOS
- Possibly roll the begin date forward by the same amount
- Go  back to the Optimize step 

I would continue to do this until I ran out  of data ...

This would show me the anchored or rolling WF / OOS results  for the 
greatest period of time and assuming the system was worth trading  
leave me with parameter values that could be used going  forward.

While I take his book as more of a guide then a bible, you  might want 
to read Pardo's relatively short ... Design, Testing &  Optimization 
of Trading Systems ...

--- In  amibroker@xxxxxxxxxxxxxxx, "brian.z123" <brian.z123@xxx>  wrote:
>
> The question is:
> 
> If I have 10 years of  data and that provides 500 Monday samples for 
> a Calender Effect study  would the margin for error be less or more 
> if I tested the sample in  one segment (500 trades) or undertook 5 
> tests with 100 samples and  then combined the results in some way?
> 
> If so, what is the  best way to use the 5 samples to get the 
> population outcomes.
>  
> No offence taken if you are too busy to answer.
> 
>  BrianB2.
> 
> --- In amibroker@xxxxxxxxxxxxxxx, "quanttrader714"  
> <quanttrader714@> wrote:
> >
> > One way to  approach this is with a form of the block bootstrap, by
> >  resampling  blocks of consecutive observations of random length 
>  with
> > replacement.
> > 
> > --- In  amibroker@xxxxxxxxxxxxxxx, "sebastiandanconia"
> >  <sebastiandanconia@> wrote:
> > >
> > > I only  offer this as a consideration when using such testing, 
> not as  a
> > > criticism of Monte Carlo Simulations.  A subtle but  significant 
> point
> > > (IMO) when using MCS:  They  may or may not be applicable to
> > > trading/investing, because  the markets don't always behave 
> randomly.
> > > 
>  > > An example of when a MCS would clearly be appropriate:  Let's  
> say you're
> > > a defense contractor manufacturing a  part for the International 
> Space
> > > Station.  The  part is critical, but because of limitations in
> > > engineering  technology it has a high failure rate, and there's 
> no way of
>  > > forecasting in advance if a part will fail.  However, although  
> the
> > > failure rate is high, it's also very  consistent.  Until 
> technology
> > > advances  sufficiently the only practical solution is to keep 
> plenty of
>  > > spares on-hand.
> > > 
> > > A MCS could  tell you what the optimal number of spares to keep 
> on-hand
>  > > would be.  The part failures are random, but MC could tell you  
> the
> > > likelihood of two, three, five, ten, etc.,  consecutive 
> failures.  You
> > > might determine  that there would only be a 1/100,000 chance 
that 
> 6
> >  > spares in a row would fail, so you might advise NASA to stock 
at  
> least 6
> > > spares at all times.
> > >  
> > > Some trading systems, though, will be successful because  they 
> take
> > > advantage of repeating sequences of  events, not random events.  
> Business
> > > cycles go  through a specific sequence, company growth follows a 
> certain
>  > > pattern from infancy to maturity, price trends/reversals follow  
a
> > > sequence, etc.  If trades based on reliable,  repeating patterns 
> are
> > > taken out of order by a MCS  such that a massive drawdown or a
> > > bankrupting series of  losers occurs, that can distort the value 
> of the
> > >  trading method by putting the trades in an order that wouldn't 
>  occur.
> > > 
> > > Soapbox alert!:)  Another  reason that "Why does it work?" is 
> such an
> > >  important question with trading systems, since a good answer to 
>  that
> > > question can lead to a good answer for another  important 
> question, "When
> > > WON'T it work?"
>  > > 
> > > 
> > > S.
>  >
>





Please note that this group is for  discussion between users only.

To get support from AmiBroker please  send an e-mail directly to 
SUPPORT {at} amibroker.com

For NEW  RELEASE ANNOUNCEMENTS and other news always check  DEVLOG:
http://www.amibroker.com/devlog/

For other support material  please check also:
http://www.amibroker.com/support.html

Yahoo!  Groups Links









Content-Description: "AVG certification"
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.1.409 / Virus Database: 268.14.11/542 - Release Date: 11/20/2006