Skip trade if last was winner

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
LeviF
Roundtable Knight
Roundtable Knight
Posts: 1428
Joined: Mon Dec 22, 2003 12:24 pm
Location: Des Moines, IA
Contact:

Skip trade if last was winner

Post by LeviF » Sat Dec 06, 2008 8:39 am

Does anyone use a rule like this? Just curious.

Paul King
Roundtable Knight
Roundtable Knight
Posts: 207
Joined: Mon Feb 23, 2004 9:13 am
Location: Vermont, USA
Contact:

Skippers don't win

Post by Paul King » Sat Dec 06, 2008 10:09 am

This depends on your beliefs (as in everything in trading really).

If you believe that each trade is an independent event, and that there is no way to determine winners and losers in advance and you have a positive expectation trading system then anything that skips trades will:

a Reduce your long-term overall return
b Randomly increase the short term volatility of your equity curve

Imaging a coin toss game where you paid me $1.20 for heads and I paid you $1 for tails - I'd be nuts to skip any chances to take that bet. On the other hand, the optimum number of bets for you to skip is all of them.

Hope this helps

Paul

Roger Rines
Roundtable Knight
Roundtable Knight
Posts: 1946
Joined: Wed Oct 06, 2004 10:52 am
Location: San Jose, CA

Post by Roger Rines » Sat Dec 06, 2008 12:14 pm

Yes - some people have tested this idea extensively. When thinking about this tactic, don't confuse skipping a trade with a rejection, and allowing a trade to work with zero size as the same thing. In one case the skip period is most often much shorter than the other, and it most often produces very different results. Also, notice which approach was used in one of the TBB Systems.

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Sat Dec 06, 2008 12:36 pm

When fooling around with the Turtle system that is presupplied with Blox, I usually get results similar to the figure below.

The top line (the one with more trades) takes all entry signals. The bottom line "skips"** entry signals if the previous trade was a winner. In my testing, the Turtle system usually gets just a tiny bit better when this rule is enabled (bottom line) than when it is disabled (top line).

Play around with the Blox Turtle system for yourself, see what you discover. Be fearless, your computer isn't going to catch fire and spray hot acid in your eyes if you make a mistake. Experiment! Explore, innovate, and learn.


**Read the Blox User's Guide description (.pdf file) of the Turtle system for all the hairy details. Its Turtle system skips entry signals if the previous trade was a winner, but it also has a "Don't be a schmuck" provision to notice whether a skipped trade seems to be turning into a monstrous winner. If so it shrugs its shoulders and enters (very late) into the trade that was skipped. Read the manual for all the hairy details.
Attachments
Turtle_LTL_filter.png
Blox presupplied Turtle system
Turtle_LTL_filter.png (7.13 KiB) Viewed 5161 times

LeviF
Roundtable Knight
Roundtable Knight
Posts: 1428
Joined: Mon Dec 22, 2003 12:24 pm
Location: Des Moines, IA
Contact:

Post by LeviF » Thu Oct 22, 2009 11:05 pm

I am getting interesting results by trading a fraction of normal position size if "last trade is winner".

I am not sure if the slight performance increase is worth the addition of another rule though...

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Thu Oct 22, 2009 11:31 pm

Good! It's often an excellent plan to replace a binary on/off switch, with a continuously variable "knob" that can assume all values between 0% ("all the way off") and 100% ("all the way on").

Another area to apply this is in Blox's "Trade Direction Portfolio Manager" which contains two binary on/off switches. You either take all Long trades (at full position size), or you skip them all (i.e. take them at zero position size). You either take all Short trades (at full position size), or you skip them all (i.e. take them at zero position size). How about replacing switches by continuously variable knobs? How about taking Long trades at position size X (0% <= X <= 100% of full size) and taking Short trades at position size Y (0% <= Y <= 100% of full size)? Those with a flair for algebra will notice that there is only one degree of freedom, namely the ratio of X to Y: (X/Y). Those without algebra, pay no attention to the man behind the curtain.

Advanced topic: If you follow Morpheus's advice to Neo ("Free Your Mind"), you can twist your knob starting from MINUS 100% (all the way on, IN THE OPPOSITE DIRECTION), rather than zero (all the way off). Liberating. In trading when you do the opposite of what is indicated, this is called "fading" the idea. You can twist the knob from 100% fade, to 0% (ignore), to 100% accept. And anywhere in between. Free your mind.

LeviF
Roundtable Knight
Roundtable Knight
Posts: 1428
Joined: Mon Dec 22, 2003 12:24 pm
Location: Des Moines, IA
Contact:

Post by LeviF » Fri Oct 23, 2009 12:27 am

Philosophically, I dont like the "last trade winner" filter. But, maybe that's what makes it a good trading rule... "do what you want to do the least"

I added a time filter to it that makes it perform a little bit better. Although this adds yet another parameter, it actually makes me feel a little bit better about the rule since a time filter makes sense. This rule is attempting to capitalize on whipsawed traders, who cares if the last winning trade was 2 years ago?...

danZman
Roundtable Knight
Roundtable Knight
Posts: 117
Joined: Thu Mar 10, 2005 10:07 am
Location: Newport Beach,CA

Post by danZman » Fri Oct 23, 2009 3:59 am

I've tested the idea on at least five different systems. It hasn't worked for me. I find it akin to looking at the past several outcomes on the roulette wheel. If there's an edge in your system, why wouldn't you take every trade? I suppose the idea is valid if you're trying to out-smart other system traders.

In my experience, the vast majority of traders use news and similar forms of information to make trade decisions.

rabidric
Roundtable Knight
Roundtable Knight
Posts: 243
Joined: Mon Jan 09, 2006 7:45 am

Post by rabidric » Fri Oct 23, 2009 7:15 am

of far more interest to me than the rule and its discussion...
...is the fact that sluggo got the out of the box turtle system to generate a MAR of over 2 !!!

even with highish leverage/risk i can't get anywhere near that with that system!

DeanoT
Roundtable Fellow
Roundtable Fellow
Posts: 52
Joined: Sat Apr 15, 2006 7:37 am
Location: Melbourne, Australia

Post by DeanoT » Fri Oct 23, 2009 8:45 am

rabidric wrote:of far more interest to me than the rule and its discussion...
...is the fact that sluggo got the out of the box turtle system to generate a MAR of over 2 !!!

even with highish leverage/risk i can't get anywhere near that with that system!
Take note of the number of trades, and Sluggo's preference for trading many, many markets. If you only run the Turtle system over the traditional 30 or so markets you likely get an MAR of < 1.0.

Post Reply