Well, that is just how I was going about it. I thought "your" idea was even better. It certainly would be a time saver. There could be a difference between forward testing going forward from the date when the AIO was run versus running AIO up to a date in the past and forward testing to the current date. An important factor in this is the idea of "stationarity" in the stock's price action; more accurately, how long stationarity persists.
I just posted on the topic of how long stationarity would persist in a stock's price action. I intuitively agree with your assessment of around 6 months. But, this process parameter is probably non-stationary. :-)
I accept the idea that B-C's parameter set needs to be periodically re-optimized. The key is to recognize when the stock's chart has changed sufficiently to require this. Or, do we give up and re-optimize every time we look at the chart.
hl
Jawed Ashraf <[EMAIL PROTECTED]> wrote:
Jawed Ashraf <[EMAIL PROTECTED]> wrote:
Actually, I forward test in real time. Boring but true.
I've considered doing what you suggest, using a downloaded file and trimming
off "recent quotes" (number of days trimmed is part of the experiment) and
then trading forwards by adding days back in to the file and watching for
signals and so on.
Just never got round to it. LOL. I don't doubt it would be educational.
In fact for people new to B-C it's probably a key exercise to undertake.
Ultimately the lesson to be learned is that very few stocks submit to
optimisation within B-C. Search high and low for them, and trade them till
they run out of steam.
6-monthly results (or the month or so before) are usually when the steam
runs out, it seems to me.
Jawed
-----Original Message-----
From: Howard Larson [mailto:[EMAIL PROTECTED]
Sent: 20 January 2005 17:11
To: [email protected]
Subject: RE: [bestcharts] Re: BC versions 4.32/4.321
This sounds like classic back testing/forward testing using historical
quotes. B-C doesn't support that directly, so one must manually download
historical quotes from Yahoo; back testing and optimizing with a file having
a date in the past (e.g., Aug 10) as the last quote, then forward testing
with a file having a later date (e.g., the current date) as the last quote.
Consistent results for variable BT periods indicates to me that stock's
price arises from a process with stationary or nearly stationary parameters;
a desirable trait indeed.
Very good.
hl
Jawed Ashraf <[EMAIL PROTECTED]>wrote:
As a matter of interest, I look for stocks which optimise well with CCI and
SO (being the primary momentum indicators) over a number of different BT
periods. What I'm looking for is a gain/trade that doesn't wildly fluctuate
depending on BT period. I'm also looking for a gain/trade that's in the
region of 8%+, in order to allow for some gains to be eaten up by costs and
(intraday) timing (timing can easily alter gain/trade by 5% if a stock is
volatile).
I assess the gain/trade based on a *fixed* history of trades. e.g. if a
stock has been trading upwards since mid-August, then I'll use a BT period
starting around 10th August(say) and assess the performance of a stock's
AIOs for BTs of 20 to 90 all during that same time period (so an AIO for 20
days is back-tested from 10 August to see how well it matched the trend, and
so on for 30 days, 40 days, etc.). I'm looking for a "strong optimisation"
i.e. different BTs tend to produce the same results.
If a stock passes that basic test, I then test-trade it, forwards. If it
passes that test, then I consider trading it for real.
When you've done all this, you aren't left with many stocks that you think
are worth trading. The further back a trading pattern in the trend started
the more likely it's at the end of the road.
Jawed
-----Original Message-----
From: Howard Larson [mailto:[EMAIL PROTECTED]
Sent: 19 January 2005 20:17
To: [email protected]
Subject: RE: [bestcharts] Re: BC versions 4.32/4.321
My assumption is that the parameters, once set using the procedure
described, should remain valid for a period of time going forward. I don't
know how long that period is, but will be determined by change in market
mood. At any rate, I intend to do a study that compares the current signal
provided by parameters stored in the past with that given by redoing the
complete parameter optimization. The problem I have with daily
re-optimization is: How valid is the current signal if it will disappear in
the days to come? Another way to look at it is, if you accepted a signal in
the past, how valid is the current signal if the old signal no longer is
shown?
Answer:
Regards,
hl
Rik Rasmussen wrote:
Sounds like you are using the B-C tools to their maximum capability.
As long as you remain vigilant to market conditions changing, and update
your strategy accordingly, you should be ok.
My problem was with "one holds it constant over time" and "associate the
optimum number of quotes with the stock symbol ".
You really can 't do the first, which is why you can't do the second.
These considerations are NOT unique to B-C, but are true of all technical
analysis.
Rik Rasmussen
Best Charts Yahoo Group Moderator
http://groups.yahoo.com/group/bestcharts/
-----Original Message-----
From: Howard Larson [mailto:[EMAIL PROTECTED]
Sent: Wednesday, January 19, 2005 11:15 AM
To: [email protected]
Subject: RE: [bestcharts] Re: BC versions 4.32/4.321
Okay, we do have a diversity of opinion on this subject. To re-cap my
technique, I select a stock and run multi-AIO to determine the best
(optimum) number of quotes to use with BT. The optimum number for me is
when the greatest number of indicators are indicating maximum profit. I
then run AIO using the optimum number of quotes in BT and save the
parameters. I then hold all of this constant going forward. I believe this
methodology to be valid so long as the market stochastics are asymptotically
stationary. This is probably true except during episodes of technical
trading.
So, if the above is not the intended use for those B-C features, how should
they be used?
Rik Rasmussen wrote:
There is no "optimum number of quotes for a stock."
Oh, there might be for a certain period of time in the past, but there is no
assurance that number will be optimum in the future.
What was optimum in an up market could kill you if the market goes into
consolidation, or turns down.
The developer can not associate such a number with a stock because there is
no such number.
The market is very dynamic. Analysis must also be dynamic.
Use numbers frozen in time and watch your funds creep away.
Rik Rasmussen
Best Charts Yahoo Group Moderator
http://groups.yahoo.com/group/bestcharts/
-----Original Message-----
From: Howard Larson [mailto:[EMAIL PROTECTED]
Sent: Tuesday, January 18, 2005 5:01 PM
To: [email protected]
Subject: Re: [bestcharts] Re: BC versions 4.32/4.321
Gents,
I agree with the statement on using multi-AIO too frequently. It has been
my understanding that one should use multi-AIO when first starting to follow
a stock in order to pick the optimum number of quotes for that stock. Once
the optimum number of quotes is established, one holds it constant over
time. That is why I have suggested the developer associate the optimum
number of quotes with the stock symbol by including it in opd.txt or opw.txt
as appropriate.
Regards,
hl
--
Howard Larson
mailto:[EMAIL PROTECTED]
Do you Yahoo!?
All your favorites on one personal page Try My Yahoo!
Do you Yahoo!?
Yahoo! Mail - Find what you need with new enhanced search. Learn more.
Yahoo! Groups Links
To visit your group on the web, go to:
http://groups.yahoo.com/group/bestcharts/
To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
Yahoo! Groups Links
Do you Yahoo!?
Yahoo! Mail - 250MB free storage. Do more. Manage less.
Yahoo! Groups Links
To visit your group on the web, go to:
http://groups.yahoo.com/group/bestcharts/
To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
------------------------ Yahoo! Groups Sponsor --------------------~-->
In low income neighborhoods, 84% do not own computers.
At Network for Good, help bridge the Digital Divide!
http://us.click.yahoo.com/EpW3eD/3MnJAA/cosFAA/GHeqlB/TM
--------------------------------------------------------------------~->
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/bestcharts/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection
around
http://mail.yahoo.com
Yahoo! Groups Links
http://groups.yahoo.com/group/bestcharts/
[EMAIL PROTECTED]
