|
PureBytes Links
Trading Reference Links
|
I would say that the following formula which was posted
some weeks ago to the ABtool group is a much better
measure for system efficiency:
pExpectancyRate =
pWinRate * pProfitRate / (pWinRate * pProfitRate
+ (1 - pWinRate) * (1 - pProfitRate))
where
pWinRate = (NbrOfWinTrades + NbrOfEvenTrades / 2) / NbrOfTotalTrades
pProfitRate = SumOfWins / (SumOfWins + SumOfLosses)
(The sums are absolute values (ie. positive numbers))
Both the WinRate (0..1) and the ProfitRate (0..1) of the
system is already counted in the formula.
The result is between 0...1 (multiply by 100 to get a percent value).
Of course the more trades the more accurate the results.
An ideal system or strategy would have the following characteristics:
ER = (1 * 1) / (1 * 1 + (1 - 1) * (1 - 1)) = 1.0
That is: an ideal system makes only winning trades, therefore has ER=1.0
ie. 100%
A real example:
The system xxSys005 has the following data:
pWinRate = 0.64
pProfitRate = 0.82
ER = (0.64 * 0.82) / (0.64 * 0.82 + (1 - 0.64) * (1 - 0.82)) = 0.89
That means: this system makes up 89% of an ideal system.
UM
----- Original Message -----
From: "quanttrader714" <quanttrader714@xxxxxxxxx>
To: <amibroker@xxxxxxxxxxxxxxx>
Sent: Monday, November 03, 2003 7:14 PM
Subject: [amibroker] Re: On Robustness, Post #2
> In the case of criterion #2, it's because I believe PF is the best of
> the metrics that come out of the standard AA backtest. In general, I
> really like PF and prefer blunt, simple metrics so I don't have to
> untangle factors or worry if some strange occurrence is giving an
> unintuitive result. Not saying MAR isn't a good metric or that it's
> complicated, lol, but personally prefer to look at profit and drawdown
> separately (as you'll see) because I understand them more intimately
> when I'm alone with them.
>
> --- In amibroker@xxxxxxxxxxxxxxx, "advenosa@xxxx" <advenosa@xxxx> wrote:
> > Mark:
> >
> > I may be jumping the gun somewhat with this question, and if I am,
> correct,
> > me, but I was wondering why you choose PF as your primary evaluation
> > criterion (as opposed to, for example, MAR, which takes into
> consideration
> > both profit and dd at the same time)?
> >
> > Al Venosa
> >
> > Original Message:
> > -----------------
> > From: quanttrader714 quanttrader714@xxxx
> > Date: Mon, 03 Nov 2003 15:52:47 +0000
> > To: amibroker@xxxxxxxxxxxxxxx
> > Subject: [amibroker] On Robustness, Post #2
> >
> > <html><body>
> >
> > <tt>
> > Here's Criterion #2 (from the condensed version of five total).<BR>
> > <BR>
> > Perform the following for each group of stocks (small, mid and large<BR>
> > cap) separately. With AA range setting on a range that contains a<BR>
> > bull, bear and sideways period (Mar 98- Mar03 is good enough),<BR>
> > commission setting at least .5% and in the reports tab, no trades,<BR>
> > backtest system on all stocks in the group. Now sort by # trades<BR>
> > (with most on top). Copy and paste into a spreadsheet. *Looking at<BR>
> > the top 20%* (ranked from greatest number of trades to fewest),<BR>
> > determine the % of stocks that have profit factors >1 (can copy down
> a<BR>
...
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
Printer at MyInks.com. Free s/h on orders $50 or more to the US & Canada.
http://www.c1tracking.com/l.asp?cid=5511
http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/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/
|