|
this is jmo but if you have enough trades and you don't have too many Opt variables, by far the easiest way to optimize is to use visual optimizations.
We do this by running the system in an Indicator and plot the equity. Substitute the Optimize() with a Param() and simply drag the Param slider left and right. This is a hundreds times faster than optimizing and you'll automatically reject over-optimization because these conditions don't last over many consecutive opt values. You can also immediately see how optimum values for one ticker apply to other tickers simply by clicking on the tickers in your work space. This, btw, gives you a real good impression on how robust (market wise) your system is. You can also immediately pick up on any system bias towards stock price, sectors, markets, time periods, Long vs Short, etc. And all that is minutes...
I find the visual feedback is very effective, catches a lot more invalid/over-optimized results, and splits etc. Performance is immediately obvious from the equity chart. Not only that, if your equity jumps up/down you can immediately zoom in on the event and analyze that happened.
I like it simple  less surprise that way.
herman
Saturday, June 9, 2007, 11:48:15 AM, you wrote:
> The linearity of the equity curve is probably one of the best measures
> of predictable future performance, if you have enough data samples. I
> believe one way to check for the linearity of the equity curve is the
> standard error provided in the optimization table and backtest
> report. I'd like to know what the precise definition of this standard
> error criteria is, by the way.
> The data I am working with is 1-minute bars, but part of it was
> collected some time ago when IB was not providing after-market data
> (last fall?), so my bars per day is not uniform, and therefore
> the equity line will have a lower slope in recent months, so my
> standard error measure is higher than it would be with uniform data.
> Consequently, I've been experimenting with a variation that allows the
> trades to be less uniformly distributed over time, but I want the
> profit per trade average to be consistent. I average the profits from
> the last 50 trades, for example, and compute the linearity of that
> instead of the equity curve.
> For both of these, we don't just want a straight equity line - it
> should be a straight line that rises, so we really want a combination
> of the net profit and this linearity measure, and one way of computing
> that is netProfit / stdErr.
> There is a lot more to this. Some other factors to consider are how
> much of your equity you are risking with each trade, how long it is
> tied up not doing something else, and the magnitude of your potential
> loss.
> Daniel LaLiberte
> liberte@xxxxxxxxxxxxx
> On Monday 04 June 2007 03:00 pm, Dennis Brown wrote:
>> Alex,
>> What you might be looking for is how straight the equity curve is. I
>> have not tried this yet in automatic mode, but when I plot the equity
>> curve I look for the gain and how straight the curve is. As a single
>> number, that would likely be the correlation to a straight line
>> between the start and ending equity values. That way you are not
>> fooled by a single rare event.
>> Dennis
>> On Jun 4, 2007, at 2:50 PM, dralexchambers wrote:
>> > I am currently testing and optimising a trading system over 1 year of
>> > data, and sorting the results by Gross Profit Made.
>> >
>> > What I am finding is that by sorting the results by Gross Profit
>> > Made, the system has long periods of small losses then one big gain.
>> > Although over a year this provides a good return, drawdowns in the
>> > interim are bad - and I am looking for regular cashflow with lower
>> > drawdowns rather than the largest gain made over a year.
>> >
>> > Can anyone think of a way to optimise results for maximal cash-flow
>> > each month rather than Gross Profit Made in a year? Is there a
>> > mathematical formula I can use?
>> >
>> > I tried using a average of x bars, but this still doesn't solve the
>> > problem, eg:
>> >
>> > Week 1: -$40
>> > Week 2: -$40
>> > Week 3: $8000
>> >
>> > whereas I would like more:
>> >
>> > Week 1: $900
>> > Week 2: $1500
>> > Week 3: $2000
>> >
>> > (this is a very simplified example but illustrates what I am after).
>> >
>> > Many thanks,
>> > Alex
> 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
> <*> To visit your group on the web, go to:
> http://groups.yahoo.com/group/amibroker/
> <*> Your email settings:
> Individual Email | Traditional
> <*> To change settings online go to:
> http://groups.yahoo.com/group/amibroker/join
> (Yahoo! ID required)
> <*> To change settings via email:
> mailto:amibroker-digest@xxxxxxxxxxxxxxx
> mailto:amibroker-fullfeatured@xxxxxxxxxxxxxxx
> <*> To unsubscribe from this group, send an email to:
> amibroker-unsubscribe@xxxxxxxxxxxxxxx
> <*> Your use of Yahoo! Groups is subject to:
> http://docs.yahoo.com/info/terms/
>
__._,_.___
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
![]()
SPONSORED LINKS
__,_._,___
|