Manually Inputted Context Zones

[SIUYA 11]

This one is for all the system gurus and programing whizzes.

(Also I am not asking if someone will do it for me, I am asking if anyone has seen it, and or what systems it might be easiest to implement on, multicharts, Ninja etc; and how hard and cumbersome would it be to do, and/or how hard and cumbersome it might be to actually use when trading

I am not a programmer but I do have some rudimentary knowledge of some of the issues)

 

This is an idea I have about auto trading with input zones (or values) being inputted by a trader throughout the day. This was the trader can determine the context of what to look for and then say if something happens between this zone, then do this on a continual basis swapping between strategies BUT without having to really focus and watch every tick. Ideally as a combo of context/discretionary trading and auto trading whereby once the zones or values are set, then the auto section kicks in.

 

example; rather than have an auto traded system that fires off all the time, this is turned on an off by the trader throughout the day, without having to focus on the charts/DOM the whole time. That way they can use the context of whats happening to determine what strategy to apply. How to establish the zones I guess can be determined by either a fixed amount eg; 15 ticks, Or the lowest low over 20 bars etc..... whatever they like. The point is this zone is more likely to be subjective. So the trader says there is a breakout, i want the zone to be between this price point and this price point. Now between these levels let the auto trader do its job.

 

I hope this makes sense, and I also hope this has already been done by some trading system that I dont know about that someone may recommend. OR if its not then this is an opportunity for those great programmers out there.thanks.

 

 

[SIUYA 11]

clearly not from the response.

I guess that means there is a business opportunity here.

[natedredd10 10]

Your trying to quantify "regime change" which happens on all time frames, forever..its probly the fundamental most misunderstood property of markets...

Even if I understood it, bet your ass I wouldn't tell you.. That kind of information is worth its weight in gold ^10

[SIUYA 11]

thanks for the reply nate...however

 

I am NOT trying to quantify regime change.

I believe as you clearly do, if you could program such measures then its worth gold. I dont believe you can quantify it, hence back test it, hence you cannot auto trade it.

 

I am looking to see if someone who has good programing knowledge is able to help direct some advice in terms of how hard/cumbersome it might be for the zones being inputted by a trader throughout the day...... that is the user needs to watch it but constantly adjust throughout the day. This way you only need to adjust zones for where you want certain auto trading strategies to occur, not specific price points or triggers. eg; you only want a cross of a MA within a certain area, and no where else.

 

I believe this would be far less work than just watching and waiting, its kind of like trying to get the best of both worlds. discretionary/context trading and auto trading

[natedredd10 10]

Well this is kind of what I'm working on. You should look into bayesian analysis ideas and the concept of priors. The computer does what its good at as far as crunching data but the initial parameters are set and variable based on my speculative talent.

Its hardly non trivial though, after a year I still don't have anything thats better than just looking at a chart.

I think you also have to keep in mind that the computer is good at looking at the vast interrelationships between markets and index components that would be impossible to chart and do analysis on yourself. I'm not really sure you can beat a well trained brain on straight out time series analysis for a single market given the problem though.

Manually marking off zones certainly cuts down on the variables but the variables are still massive for the computer to handle.

I also don't think you can really do this with candle sticks and not tick data. The more trades the system puts on backtesting the more random the results are from the information that is being tossed out with an ohlc summary of the tick data.

 

this is a pretty cool blog post from today though that i would think would be a piece of the puzzle for what you are trying to do.

http://intelligenttradingtech.blogspot.com/2010/06/quantitative-candlestick-pattern.html

Edited June 11, 2010 by natedredd10

[SIUYA 11]

thanks for the link nate.

I am not trying to do a fully automated system..... far from it. The thing I would like to attempt is a far simpler version of events of being able to switch on an off a collection of simple automatic trading systems (or even just Entry systems) if you like.

The tough part as I can see is manually, quickly and easily being able to manually tell a computer when to apply the system..... that is the human brain still provides the context. I also dont really care too much for statistics in the sense most people apply them and look at them, if anything its only being more interested in saying given these areas then what are the stats . This is more gut discretionary trading from experience of knowing (or at least thinking I know of when to apply things) This way you can quickly switch on and off the systems as the market moves.

 

 

That way I and the system dont need to understand the relationships between candle charts etc for all such past events. All I need to be able to say to the system is that between this point and this point, that the market has just fallen/risen too much, I wish to be able to apply this..... and then it might be something as simple as a MA cross over, whereby you as the human determine, I think the chop has been enough now, take the next (of the next few) MA cross overs.

(I use this as an example as I use neither candlestick patterns or moving avgs - the candle sticks I use just give me a visual map when I look at

it. Also tick data is not necessary as some of the systems are very simple limit placing orders.)

 

The picture I have added might show it better. Just say I determine as a trader my zone is based on the previous spike high. A couple of random lines later define my zone, and my stop. After that I can actually say within this zone, apply this system.... and only this system. That way I can just not watch the market so closely in the meantime, and I also know that the system is not going to go haywire in the meantime. Once I have the trade on I the system can mange it or I can exit it manually. I am wondering if there are any off the shelf systems that can easily apply this whereby the zones are quickly placed, and moved etc and varying systems/strategies are able to be used simultaneously.

 

 

There seems to be a few people like yourself that have worked on this.... maybe its a lost cause.

However I will persist and I am sure you will persist!!!!:crap:

Edited June 11, 2010 by SIUYA

[natedredd10 10]

The biggest problem with this type of stuff is in setting the thresholds for what tells the computer to not view things as a zone forming. Like in that above example at the start on the upper part of the zone, that one bar violates the zone by quite a lot and what is telling you to ignore that comes because of the forward bars after that.

If you look at the attached picture, programming the box on the left to capture the zone would be easy but would never really pick up what you want unless you line it up in the past. What you probably need is something like on the right, with smaller windows based off a few number of bars that set a larger window, that sets a larger window, that sets a larger window until you get what you want. Much much harder to do since you also have to keep track of something like if you picture 3 iterations

 

zone forming 1 = true

zone forming 2 = maybe

zone forming 3 = maybe

zone forming 4 = maybe

zone formed = maybe

 

zone forming 1 = true

zone forming 2 = true

zone forming 3 = maybe

zone forming 4 = maybe

zone formed = maybe

 

zone forming 1 = true

zone forming 2 = true

zone forming 3 = FALSE

zone forming 4 = null

zone formed = null

 

False can't just scrap everything and start over, although maybe you could have it backtrack one level to maybe again and keep going.

Edited June 25, 2010 by natedredd10

[Ingot54 13]

This one is for all the system gurus and programing whizzes.

(Also I am not asking if someone will do it for me, I am asking if anyone has seen it, and or what systems it might be easiest to implement on, multicharts, Ninja etc; and how hard and cumbersome would it be to do, and/or how hard and cumbersome it might be to actually use when trading

I am not a programmer but I do have some rudimentary knowledge of some of the issues)

 

This is an idea I have about auto trading with input zones (or values) being inputted by a trader throughout the day. This was the trader can determine the context of what to look for and then say if something happens between this zone, then do this on a continual basis swapping between strategies BUT without having to really focus and watch every tick. Ideally as a combo of context/discretionary trading and auto trading whereby once the zones or values are set, then the auto section kicks in.

 

example; rather than have an auto traded system that fires off all the time, this is turned on an off by the trader throughout the day, without having to focus on the charts/DOM the whole time. That way they can use the context of whats happening to determine what strategy to apply. How to establish the zones I guess can be determined by either a fixed amount eg; 15 ticks, Or the lowest low over 20 bars etc..... whatever they like. The point is this zone is more likely to be subjective. So the trader says there is a breakout, i want the zone to be between this price point and this price point. Now between these levels let the auto trader do its job.

 

I hope this makes sense, and I also hope this has already been done by some trading system that I dont know about that someone may recommend. OR if its not then this is an opportunity for those great programmers out there.thanks.

 

This is not in my kitchen at all, but after a third reading of this entire thread, it left me needing further breakdown of the steps in this puzzle.

 

Can I just ask if the use of say Donchian Channels might be of any help, in setting the range? The next part should simply be to switch on the bot.

 

This video (which I dug up with the assistance of Google) shows software that allows the periods to be set with the use of a slider.

 

MarketClub Videos: "Silly Season" behind us, it's time to get serious

 

It is just the idea - forget the commercial side of that.

 

Set me thinking that if the TF was set at any desired intra-day period, then this might be a giant leap forward in solving the puzzle.

 

Probably totally useless info, but I like to dig stuff up, and if I could "get it" more clearly, then I might chuck in 5cents worth where I can.

 

There are also price "boxes' based on H/L/range for set periods, but I think you are too far ahead of me for this not to have been considered as part of the solution.

 

The short answer is: No - I have not seen this done.

 

It seems to be a two-part idea you are probing for - a manual trigger for an auto-trader to get to work? You don't want the computer to set the zones - this is YOUR task ... is that correct? Then you want the computer to go to work and auto-trade, once you "tell" it the zone it is confined to?

 

If I am too uneducated in the tech requirements, or totally off-track, I am hard to offend. There might simply be too large a gap for me to be able to understand and help. I like the concept though - as far as I have understood.

[electroniclocal 10]

I do this now, manually. There are so many permutations and combinations that I have found no way to automate the selection of the "best" context.

 

What I do do, however, is let my Flobot trade some markets completely automatically with incomplete context and I also trade in what I call Hybrid Trading mode where I turn Flo on and let her make the entries and then I manage the trade either manually or by turning on another algo that I decide, again by eyeballing, is appropriate to the context and the trade.

 

The advantage of this methodology for me is that i can trade a number of markets simultaneously and not miss trades. The fully automatic algos are profitable monthly but have larger drawdowns and a lower win rate than the Hybrid Trading algos that I turn on and off.

 

EL

[Tams 20]

thanks for the link nate.

I am not trying to do a fully automated system..... far from it. The thing I would like to attempt is a far simpler version of events of being able to switch on an off a collection of simple automatic trading systems (or even just Entry systems) if you like.

The tough part as I can see is manually, quickly and easily being able to manually tell a computer when to apply the system..... that is the human brain still provides the context. I also dont really care too much for statistics in the sense most people apply them and look at them, if anything its only being more interested in saying given these areas then what are the stats . This is more gut discretionary trading from experience of knowing (or at least thinking I know of when to apply things) This way you can quickly switch on and off the systems as the market moves.

 

 

That way I and the system dont need to understand the relationships between candle charts etc for all such past events. All I need to be able to say to the system is that between this point and this point, that the market has just fallen/risen too much, I wish to be able to apply this..... and then it might be something as simple as a MA cross over, whereby you as the human determine, I think the chop has been enough now, take the next (of the next few) MA cross overs.

(I use this as an example as I use neither candlestick patterns or moving avgs - the candle sticks I use just give me a visual map when I look at

it. Also tick data is not necessary as some of the systems are very simple limit placing orders.)

 

The picture I have added might show it better. Just say I determine as a trader my zone is based on the previous spike high. A couple of random lines later define my zone, and my stop. After that I can actually say within this zone, apply this system.... and only this system. That way I can just not watch the market so closely in the meantime, and I also know that the system is not going to go haywire in the meantime. Once I have the trade on I the system can mange it or I can exit it manually. I am wondering if there are any off the shelf systems that can easily apply this whereby the zones are quickly placed, and moved etc and varying systems/strategies are able to be used simultaneously.

 

 

There seems to be a few people like yourself that have worked on this.... maybe its a lost cause.

However I will persist and I am sure you will persist!!!!:crap:

 

yes... this can be done with MultiCharts. Probably doable in TradeStation as well.

 

What you do is to set up a couple of trendlines (your zone lines),

you can manually move the trendlines to the zone you want.

then get the strategy to read the trendline positions,

and use the readings to calculate your strategy points.

[SIUYA 11]

Ingot 54 ""It seems to be a two-part idea you are probing for - a manual trigger for an auto-trader to get to work? You don't want the computer to set the zones - this is YOUR task ... is that correct? Then you want the computer to go to work and auto-trade, once you "tell" it the zone it is confined to?""

 

yes.....and thanks for generating interest in the thread again

and as EL says....... he does it in a similar way, and thats the way I have been heading also. I was more looking for the best system to do it in. NT I did not like (and it seems others seem to be steering away from it also), MC I liked but wanted manual trading from the charts as well.

I have switched to Sierra and love it...hard to learn but its stable etc; and I like it.

Currently I still find Excel the best system to manage a lot of instruments and the context, while I use the system to handle the data.

 

My real issue is having the time to learn stuff, and do all the things I am wanting to do..... plus make some money. I realize I have bitten off more than I can chew, but thats fine - you just chew harder.

Additionally it could be argued I trade too many instruments.....but I am not interested in making small amounts of money per day. I want it all

 

Tams - thanks for the MC feedback, I might look at it again at some stage, except I am liking Sierra more at this stage.

Edited February 6, 2011 by SIUYA

[Tams 20]

... MC I liked but wanted manual trading from the charts as well.....

 

Manual trading is available in the latest version (V7), which is in Alpha testing right now.

Chart trading is already done... should be announced soon.

Edited February 7, 2011 by Tams

[Tams 20]

...

I have switched to Sierra and love it...hard to learn but its stable etc; and I like it.

...

My real issue is having the time to learn stuff, and do all the things I am wanting to do..... ...

 

Tams - thanks for the MC feedback, I might look at it again at some stage, except I am liking Sierra more at this stage.

 

more than a few ppl have went against my advice,

just to come back to me later and tell me they have wasted months/years of precious time.

 

all that flash and glister are not gold,

if you are not a programmer,

but still want to have some control over your analysis,

the "English-like" EasyLanguage is your best possible friend.

Edited February 7, 2011 by Tams

[SIUYA 11]

I agree with you Tams, however as mentioned SC is working at present.

I also only needed to learn enough for about 100 lines of code, as I have pretty much done the programming in excel vba so understand the logic needed.....

plus thanks for the update re the MC newer versions and chart trading

[doubletop11 10]

Right now I'm in development of this Zone system for Ninjatrader! We hope to have the "indicator" up and running by this coming friday. The zones that this program draw are supply zone (retail price areas) and demand zone (wholesale price areas). Here's how they are traded. red line is the price to get in at, purple dot is where I got out at.

[doubletop11 10]

Zones are drawn at 4 hour chart, (swing trades, big view) 1 hour chart, 15 min, and 5 min. Reasoning for this is so you don't buy into a swing traders resistance area or sell into support. The programing for this is 500 lines of code so far and still growing as we automate signals.

[TraderWill 11]

Tams is correct, in Tradestation you can manually draw the trendlines and have your code take certain actions whenever it interacts with the trendlines. You can also specify different actions to take depending on the color or style (solid, dashed, etc) of the trendline. So definitely doable.

 

Good luck

[doubletop11 10]

So just out of curiosity how much do you think an "indicator" that drew your zones for you would be worth? $300,$600,$1000?

[Tams 20]

So just out of curiosity how much do you think an "indicator" that drew your zones for you would be worth? $300,$600,$1000?

 

it can cost very little,

it can worth a lot.

[SIUYA 11]

Drawing the zones is not a problem......

being able to easily change the zones, and tell the autotrader what to do while within the zones is the key.

In terms of cost - it depends on what offered...... if it just draws zones i have little control over -$0

if it has flexibility, draws zones and allows me to control what happens in those zones, AND is easy to use over a portfolio of instruments - not just a few.... who knows. $100pm, $100 once off, it depends on what system....if on NT for me $0.

If it can do what I want have flexibility and its all mine, and tailored for me..priceless

[doubletop11 10]

Drawing the zones is not a problem......

being able to easily change the zones, and tell the autotrader what to do while within the zones is the key.

We are drawing zones, there will be a signal generator that tells you your entry, 1st profit target (PT), 2nd PT, and stop loss at least 5-30 mins before the trade.

 

if it has flexibility, draws zones and allows me to control what happens in those zones, AND is easy to use over a portfolio of instruments - not just a few.... who knows.

We are making it for Ninjatrader and possible Tradestation. Zones on NT are completely customizable and movable. It will also be able to calc. multi-instruments at once, then give you an alert when one is touching or about to touch the zone.

 

So I really think you are on the right idea Siuya! We won't be making it for sierra, however, if you want to talk about programming logic or anything; PM me

Best of luck!

[Tams 20]

Drawing the zones is not a problem......

being able to easily change the zones, and tell the autotrader what to do while within the zones is the key....

 

this is already done/doable... (hint hint ... code is already floating around in the internet)

 

all you have to do is click and drag the trendline to a new position -- ie. your zone position.

when the trendline is hit, your code/logic/algorithm can tell the strategy what to do -- (Buy/Sell/Hold/Stand aside)

[SIUYA 11]

this is already done/doable... (hint hint ... code is already floating around in the internet)

 

all you have to do is click and drag the trendline to a new position -- ie. your zone position.

when the trendline is hit, your code/logic/algorithm can tell the strategy what to do -- (Buy/Sell/Hold/Stand aside)

 

hi Tams....a couple of questions if i may....

I had a quick look at MC again and their discretionary trader (the free version)

It seemed to me that it the full blown version still did not have the manual chart trading on it yet? is that correct or is the Discretionary trader also part of the main MC version?

 

From memeory, I think this is correct - I can hook MC up to various data feeds - eg; ESignal, and IB and look at both ASX stocks and CME futures as feed by Esignal and FX as fed by IB, and at the same time manually and auto trade via execution through IB as a broker?

thanks.

[Tams 20]

hi Tams....a couple of questions if i may....

I had a quick look at MC again and their discretionary trader (the free version)

It seemed to me that it the full blown version still did not have the manual chart trading on it yet? is that correct or is the Discretionary trader also part of the main MC version?

see post#12

 

MCDT is the same as full version MC, except it is non-programmable.

 

 

From memeory, I think this is correct - I can hook MC up to various data feeds - eg; ESignal, and IB and look at both ASX stocks and CME futures as feed by Esignal and FX as fed by IB, and at the same time manually and auto trade via execution through IB as a broker?

thanks.

datasource and broker are separate interfaces.

you can use ABC as datasource for your charts,

and autotrade executed on XYZ broker.

[SIUYA 11]

i just downloaded the free MC discetionary trader, and used it on a paper trading IB account setup for FX. ...it looks good and promising.

I might start to test it on some stocks on the ASX and see how it goes.

(one thing Sierra cant do well yet, is stocks here in Australia as it does not get a feed from someone like Esignal, and IB data is not good enough to rely on)

So lets retest MC and see how it goes..... luckily i have a lot of computers and screens