Position Score/Trade Ranking Ideas

[beachlife]

Yeah good points.

I'm just trying to make up some systems in Ninjatrader, but not sure where to start for a strategy, coding it up myself, have tried volume based stuff like if volume is more than the previous 3 bars and range is smaller and close is > open after a downtrend then go long, stop @ 10 ticks, profit @ 20, stop to B/E after 5 ticks in profit.....plus about a million different variations on that, but the equity curves didn't look too good on any of them.

Testing it on Crude Oil by the way. Doesn't seem to be any more of an edge in volume than there does in an MA crossover strategy

Probably not the right thread for this, but oh well, there's no "Coding your own strategies for fun and curiosity" thread

Here's a suggestion.

Open the CL daily chart, put a 20 period sma on it and zoom out to see about 3 years of data. For each decent move zoom in and have a look at what can be coded as an entry trigger, and have a look at ATR for initial stop and target placement. Then code it up, optimise and see what the best settings are. You want your system to catch the big moves and leave the smaller ones. For filters you can look at longer term averages, weekly data, oscillators whatever you think might work. Then run it on smaller time frames and see what happens. I find in general the smaller the time frame the worse the result. Also try it on something less choppy like an index or bank.

 

[rb250660]

So do you consider a rate of change greater than for example 5% over 5 bars or maybe an ATR multiple great than say 1.5?

That could work, however I use something more refined than that though.

 

[ThingyMajiggy]

Here's a suggestion.

Open the CL daily chart, put a 20 period sma on it and zoom out to see about 3 years of data. For each decent move zoom in and have a look at what can be coded as an entry trigger, and have a look at ATR for initial stop and target placement. Then code it up, optimise and see what the best settings are. You want your system to catch the big moves and leave the smaller ones. For filters you can look at longer term averages, weekly data, oscillators whatever you think might work. Then run it on smaller time frames and see what happens. I find in general the smaller the time frame the worse the result. Also try it on something less choppy like an index or bank.

What do you mean by ATR for stop and targets? Because if I'm looking at a SPI chart for example, and bring up an ATR set to 14(default), then it's currently at 52 on a daily chart, I'm not exactly going to set my stop 52 points away for a $25/tick instrument???

 

[beachlife]

What do you mean by ATR for stop and targets? Because if I'm looking at a SPI chart for example, and bring up an ATR set to 14(default), then it's currently at 52 on a daily chart, I'm not exactly going to set my stop 52 points away for a $25/tick instrument???

Yes thats what I mean. I use a stop of 1.5xATR(10) and a target of 2x stop. So you could test a range of say 0.5-2.5xATR with a target of 1 - 3x Stop for a target and see what it says is optimum.

THEN you decide if you can trade it or not. If a stop of around 52 points is too high then IMHO trying to use a tighter stop will see you stopped often and die a death of a thousand cuts, but test it and see.

If a wide stop is needed then you have 3 choices.
1. Use a CFD to position size.
2. Try a lower time frame where the ranges are smaller and see if the script works.
3. Trade something with less leverage.

Many years ago I tried to trade the spi with a $20k account and tight stops and blew it up real quick. IMHO $50k minimum to play with the spi futs, cfd's no problem at $1 per point, but that's a whole other thread.

 

[beachlife]

I just want to clarify so you dont charge off searching for the holy grail of code. I am not advocating a 100% coded system. The filters to stay out of trades can include things like trend lines, ranges, supp/resist that cant be coded. So start with the code to get the best settings that catch the good moves, but then manually review the chart to see what you could consider to keep you out of the losers. So the final step of your system testing has to be manual.

Then you end up with a hybrid system that can be used to create market scans. My scan today returned 6 possible trades, but my manual filters eliminated 5 of them. Some days my scans return nothing, some days my filters eliminate all results.

Of course you could look at a chart and decide to trade purley based on trend lines, in which case your testing must be 100% manual. The reason I use a SMA is because it approximates a trend line but can be coded and used in scans to flag possible trades.

Hope that helps.

 

[tech/a]

Many years ago I tried to trade the spi with a $20k account andtight stops and blew it up real quick. IMHO $50k minimum to play with the spi futs, cfd's no problem at $1 per point, but that's a whole other thread.

I'd say it was not the tight stops that blew you up.
Oscillator and Indicator trading are "p" plate technical methodsin my opinion.
If you want to be a competent technical trader these are tools which have little place in turning a profit.

Perhaps you'd like to open that whole other thread?

 

[ThingyMajiggy]

I'd say it was not the tight stops that blew you up.
Oscillator and Indicator trading are "p" plate technical methodsin my opinion.
If you want to be a competent technical trader these are tools which have little place in turning a profit.

Perhaps you'd like to open that whole other thread?

Do you believe nothing works except for VSA?

Cheers for the tips beachlife, certainly not going to look for the holy grail, just after some ideas to test to see what happens Will do some coding/testing!

 

[tech/a]

Do you believe nothing works except for VSA?

Cheers for the tips beachlife, certainly not going to look for the holy grail, just after some ideas to test to see what happens Will do some coding/testing!

Sam

I certainly believe that if you want to trade Futs Short term (in this case the SPI) you had better be able to read a chart. If your dying a death of 1000 cuts then your doing something seriously wrong. (Oscillators and Indicators are part of the seriously wrong scenario).

Lets take B/L's $20K lets say he blows up at $6000 left (Just as an example).
A loss of $14000 or 560 ticks Lets say a loss of 5 ticks a time because his stop was "TOO" Tight.(The reason he thinks he blew up!)---I think hes wrong

Thats 130 losses!!!

Do you reckon you'd be thinking that what your doing isn't working well before that??

Howard and others from time to time put up some Holy Grails---trouble is most cant believe its possible.
Of course it is! Go find your holy grail!!

Better------copy someone elses!

If your bleeding
(1) Find another way
(2) Learn about B/E stops

VSA is an important part of Chart reading. It falls far far short as a stand alone method in its very basic form (The form Tradeguider flog endlessly ).
If its not in your discretionary toolbox in MY opinion you are just one of the crowd---fodder.

 

[ThingyMajiggy]

Sam

I certainly believe that if you want to trade Futs Short term (in this case the SPI) you had better be able to read a chart. If your dying a death of 1000 cuts then your doing something seriously wrong. (Oscillators and Indicators are part of the seriously wrong scenario).

Lets take B/L's $20K lets say he blows up at $6000 left (Just as an example).
A loss of $14000 or 560 ticks Lets say a loss of 5 ticks a time because his stop was "TOO" Tight.(The reason he thinks he blew up!)---I think hes wrong

Thats 130 losses!!!

Do you reckon you'd be thinking that what your doing isn't working well before that??

Howard and others from time to time put up some Holy Grails---trouble is most cant believe its possible.
Of course it is! Go find your holy grail!!

Better------copy someone elses!

If your bleeding
(1) Find another way
(2) Learn about B/E stops

VSA is an important part of Chart reading. It falls far far short as a stand alone method in its very basic form (The form Tradeguider flog endlessly ).
If its not in your discretionary toolbox in MY opinion you are just one of the crowd---fodder.

Do you believe the edge is in VSA or in your management? I would like to see your management(of which I think is your strongest point, moving stops to B/E etc.) with another "method" or using indicators and oscillators, would be very interesting where in fact your edge lays.

I coded up a system yesterday that moved my stop to B/E after 5 ticks in profit on crude, but I just ended up having like 3x the amount of losses than winners, BUT my winners were bigger than my losses, not sure where to go next, I guess that boils down to method(VSA, indicators etc.).

Just a side note, I'm not having a go at you tech, I like VSA a lot myself, my questions are trying to point holes, so I hope you don't take offense, I value your input a lot.

 

[beachlife]

I'd say it was not the tight stops that blew you up.
Oscillator and Indicator trading are "p" plate technical methodsin my opinion.
If you want to be a competent technical trader these are tools which have little place in turning a profit.

Perhaps you'd like to open that whole other thread?

It wasnt being stopped out 130 times, it was using tight stops to just justify taking positions that were too large for my account and having those stops gapped past in the night market, and then running out of margin. No where near 130 trades required, your assumptions about how I was trading are incorrect.

 

[beachlife]

Sam I guess I should have asked this first, do you want to make money trading end of day data or do you want to intra day trade and for which ever you want why?

 

[tech/a]

Do you believe the edge is in VSA or in your management? I would like to see your management(of which I think is your strongest point, moving stops to B/E etc.) with another "method" or using indicators and oscillators, would be very interesting where in fact your edge lays.

I coded up a system yesterday that moved my stop to B/E after 5 ticks in profit on crude, but I just ended up having like 3x the amount of losses than winners, BUT my winners were bigger than my losses, not sure where to go next, I guess that boils down to method(VSA, indicators etc.).

Just a side note, I'm not having a go at you tech, I like VSA a lot myself, my questions are trying to point holes, so I hope you don't take offense, I value your input a lot.

Id say the ability to chart read in my discretionary trading--of which VSA is a part---but I do add---not the simplistic VSA touted by Tradeguider---I've been using it for more than 10 yrs!
So it encompasses Chart patterns/Elliott/VSA/and a large dose of experience.

Yes My management of a trade/s is a definite edge and coupled with the above --powerful.

To try and code up a discretionary trading method like mine would be impossible.

If you want results like a discretionary method you'll need to be a good discretionary trader.

Systems trading and their design are very different. I just don't think you can loop/reason in code as our brain can.
I can recognise something happening in seconds to code it would take months---then it changes in the next 3 min bar---
But PAV uses it and is making a second wage so 2 of us like the way it works (Pav has a very simplified version--I've only known him 18 mths.

In your case simply coding in a stop at B/E is always going to fail.
My B/E is moved for a particular reason and it could be one of many reasons---so unless I see it happening I wont move it.

That could be 1 to 20 bars!!!---Its just not that simple.

No offense Id be offended if there weren't discussion---if it gets personal then that's different.

 

[tech/a]

It wasnt being stopped out 130 times, it was using tight stops to just justify taking positions that were too large for my account and having those stops gapped past in the night market, and then running out of margin. No where near 130 trades required, your assumptions about how I was trading are incorrect.

So it wasn't tight stops it was position sizing
relative to trading capital + Poor trade planning---holding O/N
A tight stop to a 1 min trader and an EOD trader are 2 very different things.
Under capitalization appears to be a very common cause of failure in many traders. (You wouldn't need to have such a tight stop to trade size---money makes money)

 

[rb250660]

Great work destroying the thread topic guys. tech/a, you gotta stop arguing every point to death. Go outside and go for a bike ride or wet a line, leave the keyboard alone. Sadly, I expect you to retort however, don't expect me to bite.


Mike - do you enter using buy stops or limit orders? If this is the case and you get say 50 signals a day but only a few of them would typically fill you could place all 50 orders using conditional orders that trigger at a predetermined price. This way your capital is deployed progressively as price rises/falls on candidate trades. I use this technique with good success.

I've had a bit more of a think about the ranking and I reckon I would start by sorting the trades according to recent volatility. To me it would seem logical to set your volatility calculation look back period to the average hold period. For example, if you hold for 20 trading days, work out some volatility metric (maybe try using a smoothed 20 period ROC or 20 period standard deviation) and rank according to that.

rb

 

[ThingyMajiggy]

Sam I guess I should have asked this first, do you want to make money trading end of day data or do you want to intra day trade and for which ever you want why?

I'd love to do both really, I would like a system going long term, maybe some investments in shares for say a 5-10 year period, and I would like to trade intra-day on futures like I have before.

Id say the ability to chart read in my discretionary trading--of which VSA is a part---but I do add---not the simplistic VSA touted by Tradeguider---I've been using it for more than 10 yrs!
So it encompasses Chart patterns/Elliott/VSA/and a large dose of experience.

Yes My management of a trade/s is a definite edge and coupled with the above --powerful.

To try and code up a discretionary trading method like mine would be impossible.

If you want results like a discretionary method you'll need to be a good discretionary trader.

Systems trading and their design are very different. I just don't think you can loop/reason in code as our brain can.
I can recognise something happening in seconds to code it would take months---then it changes in the next 3 min bar---
But PAV uses it and is making a second wage so 2 of us like the way it works (Pav has a very simplified version--I've only known him 18 mths.

In your case simply coding in a stop at B/E is always going to fail.
My B/E is moved for a particular reason and it could be one of many reasons---so unless I see it happening I wont move it.

That could be 1 to 20 bars!!!---Its just not that simple.

No offense Id be offended if there weren't discussion---if it gets personal then that's different.

So how can one master VSA, or your version of VSA is what the Tradeguider guys say is bollocks, or at least over the top, they almost make me ill with how heavy they are on the sales and pitches, I want to learn your management ways before anything else, that's where I'm lacking most I think, but where to find good management skills is beyond me, can't exactly Google "trade management techniques" without getting the usually huffin puff of cut losers, hold winners blah blah.

Sounds like we should go to( or start ) another thread.

 

[tech/a]

Great work destroying the thread topic guys. tech/a, you gotta stop arguing every point to death

Thanks

You've reminded me why I left.
What a waste of time and effort it is
on forums.

 

[pavilion103]

Thanks

You've reminded me why I left.
What a waste of time and effort it is
on forums.

For every one of his 13 posts, you have constributed 1,000 valuable ones.

Then he takes a personal pot shot. Weird.

 

[beachlife]

appologies to the OP

Just quickly

I'd love to do both really, I would like a system going long term, maybe some investments in shares for say a 5-10 year period, and I would like to trade intra-day on futures like I have before.

My trades usually last around 2 weeks so I cant offer much with either of those, except to say that I am playing with 1min EA's on indexes that show promise in back testing but havent been proven live yet. But they dont get anywhere near the 80% win rate that you would like, and have really wide stops. Try your oil script with wider stops and see what happens.

 

[MikeRR]

Good discussion guys. It's no problem for me if the thread morphs and changes as we go. All of the topics being discussed are of interest to me.

The system that I have coded up is the weekend trend trader by Nick Radge. I've decided to take this system which essentially is a medium term trend following / momentum system that buys breakouts confirmed by an ROC indicator and the sets percentage stops that are tightened up when the index turns down.

The system tests out very well and I'm getting just about the same results as published in the book so I think it's tradeable as it is. However, to some degree I think there would be the opportunity to improve on it and since I managed to get AmiBroker to do what I wanted it to I thought I would go about implementing some of the changes to see how it performs.

After reading some of the suggestions here and some other materials I'm going to attempt to normalise various system elements based on volatility - for example, a 10% stop in a $20 stock is vastly different to a 10% stop in a $5 and this type of approach doesn't take into account the individual nature of how each stock behaves. I'm not saying this approach doesn't work, it clearly does but for my own purposes I need to investigate if there is a better way.

My first point of call (when time permits) is to look at implementing ATR based position sizing, stops and position score/ trade rank (lower volatility will be rewarded).

In addition to that I need to give some thought and research to the ROC filter that is used in the system. This filter is also percentage based which does the job of limiting signals that are presented but again this (to me in my small inexperienced world) doesn't give equal weighting to each stock that meets the breakout requirement because a 20% ROC(20) on one stock may be common whereas another it isn't.

Just a few thoughts that I have been working through along side a short term system.

Thanks again for everyone's input and conversation.