Last day rule

Discussions about Money Management and Risk Control.
Post Reply
ColdFact
Senior Member
Senior Member
Posts: 31
Joined: Wed Oct 27, 2010 2:07 pm

Last day rule

Post by ColdFact » Tue Aug 07, 2012 12:36 pm

What are people's thoughts on using the "Last Day Rule" - i.e. you set the stop loss on a breakout equal to the low the day of the breakout if the breakout is not a jump up on open - or the low of the previous day if it is.

Secondly, is anyone aware of a block having been created to test this?

Thanks!

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

Post by sluggo » Tue Aug 07, 2012 6:59 pm

Sounds like you're talking about a rule that only goes long, i.e., that only enters on breakouts to the upside.

I think it will be important to decide whether or not you'll enter on the bar where the breakout occurs, or on the next bar afterwards.

Similarly I think it will be important to decide when you want to place your stop order. During the same bar that you entry order was filled? After the close of the entry bar (protecting you during the Night Session)? At the open of the next bar after the entry bar? Lots of possibilities.

If you enter on the breakout bar itself, you won't know the Low of that bar until after the close, so you won't be able to place your stop at the Low (or at any price that's a function of the Low), until after the Close. Some days, the Low doesn't occur until the Close itself.

A good way to select among the possibilities is to code them all and simulate them all. Then choose the one that offers the best trade-off (in your view) between simulated performance, ease of trading, data requirements (daily data OK? need intraday data?), and perceived level of safety/risk.

ColdFact
Senior Member
Senior Member
Posts: 31
Joined: Wed Oct 27, 2010 2:07 pm

Thanks Sluggo

Post by ColdFact » Tue Aug 07, 2012 7:27 pm

If only my coding skills were up to it. A friend of mine has this on his list of things to add once he can. What I'm talking about works for up and down trends. If say a new high is breached in corn - you take it - i would want to get out if the low of the day was breached, even if it was minutes after the high was reached. Same for the low - if nat gas makes a new low for the previous month, for example - the high of that day would be used as a stop that would immediately be placed upon entry to nat gas. If I ever get this up and running I'll post the results. I became very intrigued when I noticed that over 90% of the profitable trades in my system never breached the breakout day low (or previous day if breakout was a gap) - it provides a much tighter stop.

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

Post by sluggo » Tue Aug 07, 2012 10:19 pm

Good luck with the idea, I hope it works well in testing and even better in live trading!

Versions of it that use intraday data, may perform differently than versions that use end-of-day data; you and your friend might see this as a cost/performance tradeoff.

LeapFrog
Roundtable Knight
Roundtable Knight
Posts: 695
Joined: Mon May 17, 2004 4:18 pm
Location: Boston, MA

Re: Thanks Sluggo

Post by LeapFrog » Wed Aug 08, 2012 10:23 am

ColdFact wrote:If only my coding skills were up to it. A friend of mine has this on his list of things to add once he can. What I'm talking about works for up and down trends. If say a new high is breached in corn - you take it - i would want to get out if the low of the day was breached, even if it was minutes after the high was reached. Same for the low - if nat gas makes a new low for the previous month, for example - the high of that day would be used as a stop that would immediately be placed upon entry to nat gas. If I ever get this up and running I'll post the results. I became very intrigued when I noticed that over 90% of the profitable trades in my system never breached the breakout day low (or previous day if breakout was a gap) - it provides a much tighter stop.
You don't specify what a "new Low/High" means - do you mean 'lower/higher than the previous day's Low', or the 'Low/High of the past x bars', etc.

In order to test entering a stop during the day as you describe you will need to test on intraday data. Getting a hold of that can be tricky if you want to test over say at least the past 5 years.

But assuming you can get reliable intraday data on the markets you want to test, and that you define your trading rules, Blox can be your friend to test it.

If you can't, or don't have the time but have the money, you could always retain one of the programing consultants on this site to build a test engine for you.

Me myself and I have not found reliable intraday data to make testing such an idea worth the effort of coding it up.

gunter
Roundtable Fellow
Roundtable Fellow
Posts: 83
Joined: Wed Aug 24, 2005 7:25 am
Location: George Town, Cayman Islands

Post by gunter » Wed Aug 08, 2012 10:50 am

What I did to get around the intra-day data limitation was use the prior day's high / low for stop placement and then "tighten" up the stop on the day after the breakout by moving my stop to the high / low of the beakout day.

That system was pretty aggressive and I am no longer trading it, but it was mostly due to poor exits. I did have a pretty low win-rate, but the winners made up for it. Only problem was that all my losing trades tended to cluster together and the account could go down 50% in a relative short period of time. This was due both to losing trades and open-equity drawdowns.

Good thing I only allocated a smallish portion of my money to it. :?

Post Reply