|
PureBytes Links
Trading Reference Links
|
Al, others:
you would not believe how easy walkforward testing is with the new Rx - Report
Extractor from Dale Wingo. It was annouced several days ago how Dale is
looking for additional beta testers (although I can say the program is a pretty
darn stable version right now). This is a tremendous time saver as you set
up some codes within your AFL file, make all your settings, code arrangements,
etc in the AA window. You press BackTest - Report, change the date range
(as in your example below), press Backtest again, then Report again. Very
easy to do quarterly or half-year (or more) rolling window
testing.
<SPAN
class=360451319-29092002>
Switchto Rx,
follow some screen commands, and presto, there are all of your results fromthe
two runs, in two (or more, if you did more than two tests) rows of Excel, with
all documentation of settings, formula, etc. You can even put some
"comment" code lines in the AFL code (and change it before each run) and
these comments will be captured as a field and put into the spreadsheet
also. Very easy to compare all of the fields of data and draw your
conclusions. You can do it probably faster than it has taken me to type this
note.
<SPAN
class=360451319-29092002>
Interestingly, if
you decide to vary a few more parameters, you make those new runs, andthen
click a different box in Rx and it appends the last data right behind the data
of the first runs. I keep several master excel files, one for each system
test I do, and the results of each test are appended right below the last run,
but with all settings, etc, including the formula, so I know just what each
result row refers to. It even can let me type in some conclusionary notes
in Column A, last empty row, and the next system test row is appended belowmy
notes. One of my master file has 35 rows of data, representing 35
different tests done over several weeks. Of course, it also captures the date of
the test.
<SPAN
class=360451319-29092002>
Folks,if you
have juggled html reports and tried copying certain fields from them to paste
into a spreadsheet in order to keep track of different system tests, and gotten
all mixed up about it, then I assure you the Rx program will literally amaze
you.
<SPAN
class=360451319-29092002>
Dale may still
need some more beta testers.
<SPAN
class=360451319-29092002>
Ken
<FONT face=Tahoma
size=2>-----Original Message-----From: Avcinci
[mailto:avcinci@xxxx]Sent: Sunday, September 29, 2002 2:01
PMTo: amibroker@xxxxxxxxxxxxxxxSubject: Re: [amibroker]
Suggestion for better backtesting/optimization
You can do walk-forward testing now with AB. Simply optimize from,
say, 1997 to 2000, then using the optimum parameter values from that
optimization, backtest on the same stocks from 2000 to present (out of sample
data). You can vary the time periods to do your optimizations and backtests. Of
course, this is 2 steps, but it's not that hard. TJ has said that, later this
year, he will be incorporating in his graphics engine 3-D graphics that enable
the user to view the robustness of an optimization in a manner similar to but
better than the Excel add-in that Herman developed and uploaded earlier. You
would choose parameter values from a flat area of the optimization graph where
the parameter values and equity curve would not change much from one set to
another.
Al V.
<BLOCKQUOTE
>
----- Original Message -----
<DIV
>From:
lel4866
To: <A
href=""
title=amibroker@xxxxxxxxxxxxxxx>amibroker@xxxxxxxxxxxxxxx
Sent: Sunday, September 29, 2002 12:29
PM
Subject: [amibroker] Suggestion for
better backtesting/optimization
My suggestion is for support for walk forward optimization.
This is what we all actually do in real life. Since we can't look into the
future, we test/optimize based on the past, then apply the results to
the following days trading. We then look at the results and make
additional changes.There are 2 simple variables: 1) The length of
time we look back, and the length of time (or other conditions) we wait
until we re-optimize.There's a 3rd, more complicated variable-
how we choose the optimium parameters from the last optimization run.My
suggestion is for a "score" formula that takes into account things like:
maximum % drawdown, best compound rate of return, highest minimum of the
running 1 year returns (or whatever period you like).I
particularly like the last one - I always look at a graph of the 1 year
returns (Equity() - Ref(Equity(), -253)). Ideally, it should be as flat as
possible (or maybe rising). I've been treating this a little like analysis
of variance - it should look like white noise with a given mean and
standard deviation - any periods that don't bear
investingation.Right now I use AmiBroker for experiments, but when I
find an idea I like, I must write my own program to do walk-forward
testing.Larry LewisPost 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.
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.
|