Today's Action by Intelligent/Predictive Agents

[davewolfs 10]

Yes I know nothing. And I am not saying much

 

Folks the point is, neither is Urma and those who are try to replicate his ideas.

 

Find your niche, create your own ideas. There is no holy grail in this thread or any of the other threads that have been posted. There is no "rocket science" presented here. Price pays. The charts look great in hindsight, that is about it.

 

Happy trading and may you all be profitable.

[Kiwi 16]

Folks the point is, neither is Urma ...

 

Ain't that the truth

 

We had a guy like him over at T2W once ... self-aggrandizing ... putting everyone else's approaches down ... a legend in his own mind. Turned out he was living in a trailer park in Texas.

 

Isn't the internet a wonderful thing :haha:

[Tams 20]

Yes I know nothing. And I am not saying much

 

Folks the point is, neither is Urma and those who are try to replicate his ideas.

 

Find your niche, create your own ideas. There is no holy grail in this thread or any of the other threads that have been posted. There is no "rocket science" presented here. Price pays. The charts look great in hindsight, that is about it.

 

Happy trading and may you all be profitable.

 

you sound frustrated that Urma is not handing you a freebie.

 

well, neither are you giving a freebie... LOL

[TheDude 10]

Certainly size can show you that the commercial is present, i.e., the 2 lot trader never trades a 100 lot.

 

The issue with the efficacy of this approach is that you need to differentiate between commercial spec trades and commercial permium aribtrage.

 

The commercial who trades for premium capture doesn't care about price, he only cares about the ROI offered by the current permium level.

 

The commercial spec trader cares nothing about premium.

 

My head of technology has had 3 offers from 3 different hedge funds in the last 6 mos to build technology to disguise just such entries.

 

cheers

 

UB

 

Size alone is useless for this work.

 

PS Does anybody know what's wrong withthe search engine. I can search usernames but not keywords.

 

With all due respect, I think you are a little muddled here.

 

With all exchange traded products, be they futures or stocks, one of the selling points the/any exchange will promote is transparency. This means, in it's simplest fashion, that they do not publish who is trading what.

 

The CME/CBOT did produce the liquidity data bank (LDB) which is the closest thing available, but that will just show what category of market participant has traded what and at what price. This as I'm sure you are aware is of limited value to you given the data is and can only be published at the end of the day.

 

So, if we cant see who is trading what by label, we can only guess as you seem to be doing. You seem to be guessing by size of order. Correct me if I am wrong.

 

Anyone who has worked professionally in the industry should be aware of TCA (transaction Cost Analysis). This is essentially methods of breaking institutional (large) orders up so they have less of an impact on the market. (therefore, in your example, the 2 lot trader could actually be the 100 lot trader - sending orders via smart order routing technology to different brokers) If we take crude traded on ICE as an example, the DOM will typically show 2-4 lots bid offered all the way up and down, yet we all know that some of these 2 lots are retail, others are part of an iceberg for say 20 lots or more. You simply do not know. Millions are spent to make sure you do not, and con not detect. To take this to another level and to try to pretend you know if the trade is speculative or arb is for the land of La La. Nobody does. Whats more, nobody really cares.You take your view, you take your trade - based off YOUR objectivity. To try and follow someone elses coat tails is a hiding to nothing. If I buy ES, do I care what BlueCrest are doing? NO. They could be wrong anyway.

 

If you want to know which way the commercials are positioned, look at CoT reports, or easier yet, a daily price chart.

 

Of course, I could be very wrong. Morgan Stanley could indeed be trying to poach your buddy. In my experience though, such people try and keep a low profile. Attention is the last thing they seek. Blabbing over an internet chat room about some supposed holy grail that can detect the undetectable is the last thing they are doing!

 

Good luck,

 

Duderino.

[cunparis 10]

The CME/CBOT did produce the liquidity data bank (LDB) which is the closest thing available, but that will just show what category of market participant has traded what and at what price. This as I'm sure you are aware is of limited value to you given the data is and can only be published at the end of the day.

 

It's available every 15 minutes. But I can't figure out exactly who is each of the trader categories to make it useful.

 

So, if we cant see who is trading what by label, we can only guess as you seem to be doing. You seem to be guessing by size of order. Correct me if I am wrong.

 

Anyone who has worked professionally in the industry should be aware of TCA (transaction Cost Analysis). This is essentially methods of breaking institutional (large) orders up so they have less of an impact on the market. (therefore, in your example, the 2 lot trader could actually be the 100 lot trader - sending orders via smart order routing technology to different brokers) If we take crude traded on ICE as an example, the DOM

 

Have you read his other threads? The reason he uses intensity is precisely for the reason you describe, smart traders split up their orders. When one sees a series of orders in a very short timeframe (seconds for me, but milliseconds for him) then you can assume they come from the same trader or a similar group of traders.

 

In his post he said a 2 lot trader never trades 100 lots. In your reply you're saying a 100 lot trader can split his order into 50 2 lots. But he's still a 100 lot trader. What urma is saying is if we see a 100 lot order we know it's not a small retail trader so it's professional. From what I understand, if it's 100 lots on a trade it's more likely to be arb because a smart spec would have split his orders to hide them.

 

I hope that clears up some points. If I misunderstood anything please let me know.

[become 10]

you are saying a lot of nothing.

 

 

 

oops... more of nothing again.

 

- which makes your posts nothing to the power of two.

 

Really, all you're doing is stifling discussion. Right or wrong, davewolfs was making a contribution. Your metacomments aren't.

[TheDude 10]

Thanks Cunparis.

 

Perhaps it's just a difference of opinion (hey! we have a market!!)

 

1 - As for LBD, I wouldn't bother with it so much if it is available intraday/15mins. Here's why: Often a commercial will give large orders to brokers to execute, splitting it between brokers. Some they may even execute themselves on a different day or time of. Brokers like Citibank, MFG etc will come under one Cti, where as the commercials will have another Cti code. The broker probably won't allocate until End of Day. So we have one trader executing a position under different Cti codes. Therefore, it's limited in use intraday IMO. Please someone correct me if I'm wrong on this.

 

2 - No I haven't read the other threads. I've read a handful of other posts and the main core in my opinion is deluded and pied piper. Nothing more. Like you say. A lot of this is based on assumption. We all know the old saying about assumptions! If you see a spurt of prints, it could be an iceberg, it could be similar algo's generating orders, it could be a large bunch of stops being hit. Who knows? Who cares? You know the volume, you know the price. Thats all you need to know.

 

Saying a 2 lot trader never trades a 100 lots is again, a matter of opinion. Irrelevant opinion at that. Maybe my algo is telling me that today, I should build my position in 2 lots. Today, I'm a 2 lot trader. Next week, my algo tells me to put it all on. Hey today I'm a 100 lot trader.

 

If you cant tell whether the 2 lot is retail, or iceberg, whats the point? Whats the fascination with 'professional' anyway? They often get it wrong too! And why is a 100 lot likely to be an arb? Sorry, but this is ignorant. The problem with an iceberg is that it gives a higher degree of execution risk. When you get filled on your first 2 lots, the next 2 lots get sent to where - yep, the back of the queue. In the time it takes to receive a fill and send the next 2 lot in, the price could move and you end up with 98 unfilled. Great! This is rare, but it does happen.

 

My point is, is simply that if 100 lots trades, great. If 2 lots trade, great. Maybe both were part of the same order! Maybe the iceberg executes at random clip sizes. Who cares? Whats more important to me is that so far 102 lots have traded at a price. Who or why they traded isn't the concern. It's over complication. It has no value other than to try and appear that one is more cleverer than everyone else. It pleases the ego, but will not please the account balance.

 

Genius is to simplify the complicated

The foolish complicate the simple.

[MadMarketScientist 11]

Please just some help on the above - this is a great thread - very fascinating posts from several of you. However, as I read it there seems to be a tone of some personal cutdowns. If possible, please feel free to critique and debate the theories of course and ideas, but let's try to keep it on a different level and not take it personal if you do not agree. Keeps it more constructive and enjoyable for everyone that way. Thanks for the understanding.

[cunparis 10]

Thanks. I would like to see your spikes for 6:45 to 7:07 your time.

 

Urma, I'm not sure if you saw my request, but regarding this post could you please post your chart for the previous 15 minutes? Thanks.

[boomerangas 10]

Me too, really. Because it would be very interesting to see what happened on your indicators on 1158 level breakdown and what happened on the extreme. Urma, please, clear everything because it smells cheating. I could tell you the truth, that mine TI indicator sucked on these spike, it showed sell volume, which was just stops liquidating. And what helped me to avoid the loss - fully discretional trading and expirience. What happed later - just new trend, and if you have expirience you just are ridind such a trend till the first congestion. No predictive edge here, just trade what you see, and don't what you think. Strange, that you've just ignored mine and cunparis questions. Thanks!

 

Urma, I'm not sure if you saw my request, but regarding this post could you please post your chart for the previous 15 minutes? Thanks.

Edited May 23, 2010 by boomerangas

[cunparis 10]

Me too, really. Because it would be very interesting to see what happened on your indicators on 1158 level breakdown and what happened on the extreme. Urma, please, clear everything because it smells cheating. I could tell you the truth, that mine TI indicator sucked on these spike, it showed sell volume, which was just stops liquidating. And what helped me to avoid the loss - fully discretional trading and expirience. What happed later - just new trend, and if you have expirience you just are ridind such a trend till the first congestion. No predictive edge here, just trade what you see, and don't what you think. Strange, that you've just ignored mine and cunparis questions. Thanks!

 

I would not accuse him of cheating, I'm very glad he's shared his ideas & posts. I'm just curious if he got any spikes in that time leading up to the top because I did and I'm trying to figure out a way to make it more accurate.

 

I also found a few times that the intensity appeared to be from stop orders - for example stop loss for short sellers and buy stops for breakout traders. I noticed this because I saw spikes when I was stopped out on Friday. Here's two charts from Friday along with my comments. I had a problem with my DOM and my second target was closed with the first one so if it looks like I took a quick profit that's why..

 

 

 

I'd love to have some comments on these charts.

[UrmaBlume 10]

Me to, really. Because it would be very interesting to see what happened on your indicators on 1158 level breakdown and what happened on the extreme. Urma, please, clear everything because it smells cheating. You've just ignored mine and cunparis questions. Thanks!

 

boom,

 

Cheating?

 

You have pestered me with pm's and emails for weeks and I have responded to most of them. I have tried to explain that I get requests/demands like this everyday and don't have either the bandwidth or the inclination to respond to such, especailly at this low level of concept and granularity.

 

You are doing some good work. Thanks for what you have shown me and for the kind words.

 

I will not write your code for you and I will not test, prove or disprove every thought, suggestion or situation that comes to the mind of every 1 lot trader here.

 

Almost none of the posts on this board are about the poster's original concepts or technologies. They are almost always about what someone is saying about someone else's work. Almost all of my posts are about our own, original, in-house development.

 

I enjoy posting here and am grateful for the hundreds of thanks, pms and emials I have gotten thanking me for what I have shared. Some have taken the hints I have provided and gone on to develop their own tools and they report improvment in their trading results. Others don't get enough to finish on their own and still others don't get it at all.

 

Unlike those that are ill-equipped to handle a discussion of the topic at hand, I spend zero time stalking and bashing those I disrespect the most.

 

As I said I greatly enjoy the conversations I have on a daily basis with TL members, thank them for the thanks I have gotten and have even taken some TL members on as traders.

 

I post here for 2 reasons 1) To try and get the brightest here to think about outside the box and 2) to try and spot talent.

 

UrmaBlume

[UrmaBlume 10]

I started this thread and this thread to discuss the practical application of intelligent/predictive agents in trading.

 

While I was expecting the replies/discussion to be about neural networks, genetic optimizations and the like - No such posts, zero.

 

While I would believe that less than 1% of the members here understand and deploy such technologies in their trading, I thought there would be more interest by some in learning how they might add this higher level of processing to their work.

 

As to the bashing and the stalking - that's just what happens when you take people outside their comfort zone and what separates the ones we think about helping or hiring and the ones we don't.

 

UrmaBlume

[cunparis 10]

While I was expecting the replies/discussion to be about neural networks, genetic optimizations and the like - No such posts, zero.

 

While I would believe that less than 1% of the members here understand and deploy such technologies in their trading, I thought there would be more interest by some in learning how they might add this higher level of processing to their work.

 

As to the bashing and the stalking - that's just what happens when you take people

 

Maybe your posts are over our heads? Could you try breaking it down a bit and starting with simpler subtopics? I'm very interested in this but feel a bit out of it cause I don't understand a lot of it.

[boomerangas 10]

Don't get nervous, Urma, take it easy . Here is "nothing personal, just a business". You've done a great work, it works for you, also helped ( theoretically ) to a lot of members here. I've just discussed, that there is no ultimate edge in this method and asked to post just one screenshot several minutes earlier. It shouldn't be hard task for you and it's just very strange, when you are saying that you have no time to deal with a 1 lot traders and to post something for the community but post the posts with a critics and self-ego fulfilling.

 

"I post here for 2 reasons 1) To try and get the brightest here to think about outside the box and 2) to try and spot talent"

 

Myself is not a talent, but you even don't want to discuss something in more details with a talented people and to attract them because what you are providing is 10% of info you know. And this 10% of info are unneeded and empty info for the real quant. Just my subjective opinion...

 

p.s. and just observations. And once again - no real reply to mine and cunparis questions, just getting nervous and a lot of critics...

 

 

 

 

 

 

 

 

QUOTE=UrmaBlume;97022]boom,

 

Cheating?

 

You have pestered me with pm's and emails for weeks and I have responded to most of them. I have tried to explain that I get requests/demands like this everyday and don't have either the bandwidth or the inclination to respond to such, especailly at this low level of concept and granularity.

 

You are doing some good work. Thanks for what you have shown me and for the kind words.

 

I will not write your code for you and I will not test, prove or disprove every thought, suggestion or situation that comes to the mind of every 1 lot trader here.

 

Almost none of the posts on this board are about the poster's original concepts or technologies. They are almost always about what someone is saying about someone else's work. Almost all of my posts are about our own, original, in-house development.

 

I enjoy posting here and am grateful for the hundreds of thanks, pms and emials I have gotten thanking me for what I have shared. Some have taken the hints I have provided and gone on to develop their own tools and they report improvment in their trading results. Others don't get enough to finish on their own and still others don't get it at all.

 

Unlike those that are ill-equipped to handle a discussion of the topic at hand, I spend zero time stalking and bashing those I disrespect the most.

 

As I said I greatly enjoy the conversations I have on a daily basis with TL members, thank them for the thanks I have gotten and have even taken some TL members on as traders.

 

I post here for 2 reasons 1) To try and get the brightest here to think about outside the box and 2) to try and spot talent.

 

UrmaBlume

[ecubru 10]

Urma

 

Your ideas look very impressive at first glance. However, and don't get me wrong, to what extend you are just trying to look for rational justification for taking trades. Commercial buying, retail, whatever.

 

You are just trading simple MA crossover with an "oscillator" as a filter. If your oscillator includes volume or not seems to be irrelevant in your charts because both indicators in your bottom panel will show the exact same divergences than the original price chart. In addition your TPT intensity indicator its as random or good as the very old Elders Ray indicator. In other words, every time you have a new swing you will have some peeks or valleys and when the move is overextended it will overshoot in large spikes (nothing different than a smooth stochastic in oversold/overbought will show) .

 

i don't have a clue on how to program or setup your indicators but in 10 mins and using the most traditional indicators (MA and MACD) and with a little common sense i have exactly your same charts with exactly the same results.... and yes the reversal on your 05/21 charts was clear without any volume or big comercial retail, etc info available....

 

I think a picture will be worth more than thousand words in this post...

 

ecu

 

 

 

It is an absolute fact that some technical traders can indeed "actually determine who is buying or selling - then you are kidding yourself" and it is also an absolute fact that some others can not

 

The chart below is based entirely on our determination of the balance of trade by commercial traders.

 

You will notice net commercial trade is going straight up. You will notice only blue bars and no selling signals at all.

 

The + is a recommended entry point, the red dot a stop and the blue dot a recommended scalping point.

 

The text suggests the TradePoints for entry, stop and scalp for the current bar and is updated every tick. When there is no recommended trade as when the bars are gray then "NO TRADE" is posted.

 

This chart is from today's open at 0630

 

[UrmaBlume 10]

ecubru;

 

The first difference between your chart and the one displayed below is that the bottom 2 sub-graphs on ours have price as NO part of their calculation. By far the most heavily weighted inputs to our trade decision making process are those that have to do with the balance, velocity and acceleration/deceleration of those trades that we classify as comming from commercial speculative trade.

 

The process of the construction of the chart shown below is:

 

The signals on the chart specifically the text that recommends precise TradePoints or No Trade are the product of several layers of processing.

 

The first layer gathers data from seven different time frames, posts it to global variables, and uses some of it to form an index of weighted biases as described in this thread.

 

Other of these variables are then passed as inputs to a set of intelligent agents that range from genetically optimized neural networks, functions producted by MARS (Multivariate Adaptive Regression Splines) and CART which is "a robust, easy-to-use decision tree that automatically sifts large, complex databases, searching for and isolating significant patterns and relationships." Both MARS and CART are products of Salford systems and some of the easiest to use tools for the development of intelligent agents that we have ever found. We have developed a tool in house that can convert the functions from MARS to Easy Language and can put a newly developed model on line and producing the Trade Points you see on the charts in a very short time. This entire process is described in this thread..

 

While price is considered, the bulk of our decision making process is based on various measures of what we call commercial speculation. We first measure our calculation of time-of-day normalized commercial presence. This tells us that at this time of day the commercial precence is some percentage of what it normally is at that same time of day over the previous 72 market days. If that percentage is less than a certain number we do mean reversion trading, if that percentage is over a certain percentage then the trend following methods shown on the chart below are shown. We also measure the percentage of time-of-day normalized commercial speculative bias in those 7 same time frames.

 

We think in terms of portfolios of both methods and time frames for each instrument we trade. Some of our indicators indicate trade and some indicate method or time frame.

 

Some of these indicators of commercial presence for both method and time frame can also be used for instrument selection. For example if the S&P is only trading at 70% of normal commercial presence then it is for sure that smart money has found at least a couple of stocks in the S&P that are trading at 200% of normal commercial presence that is where the action will be on that day in that index.

 

There are many ways to evaluate the price volume relationship, implement VSA, follow Wyckhoff or implement the Market Profile. The notes above roughly describe how we input and process the concepts and information from all of the above to output the precise TradePoints on the chart below.

 

All of the indicators on the chart below are calculated in real time and the TradePoints and text recommendations are posted at the first tick of the bar and are updated every tick. The small red dots are stops, blue are scalp profit targes the + is a recommended entry price for that bar. The recommendations are for the last bar on the chart.

 

 

Cheers

 

 

 

 

Urma

 

Your ideas look very impressive at first glance. However, and don't get me wrong, to what extend you are just trying to look for rational justification for taking trades. Commercial buying, retail, whatever.

 

You are just trading simple MA crossover with an "oscillator" as a filter. If your oscillator includes volume or not seems to be irrelevant in your charts because both indicators in your bottom panel will show the exact same divergences than the original price chart. In addition your TPT intensity indicator its as random or good as the very old Elders Ray indicator. In other words, every time you have a new swing you will have some peeks or valleys and when the move is overextended it will overshoot in large spikes (nothing different than a smooth stochastic in oversold/overbought will show) .

 

i don't have a clue on how to program or setup your indicators but in 10 mins and using the most traditional indicators (MA and MACD) and with a little common sense i have exactly your same charts with exactly the same results.... and yes the reversal on your 05/21 charts was clear without any volume or big comercial retail, etc info available....

 

I think a picture will be worth more than thousand words in this post...

 

ecu

 

Edited May 28, 2010 by UrmaBlume

[natedredd10 10]

This is something that is really quite important. Most of the approaches that start with the premise that trade direction can be determined by looking at the last price with respect to best bid and ask neglect this phenomena. An advance can be stopped by aggressive buyers running out of steam or by passive sellers standing firm.

 

you are falling hopelessly behind old man...I would speculate your woman is just getting old..or at least what you see as prospective females too old to "light a fire"...Maybe its time to "trade up" to a newer model? Beyond that variable who cares...

You can't even start to have this argument that you tried to put up there without book data so why did you even bother to put it forth? the sad thing is I know you already know this.

[UrmaBlume 10]

Nate,

 

In answer to your first query "what tick database are you using" I assume your are talking about which data vendor, not the dbms itself. In any case I will answer both.

 

For our data feeds we use both E-Signal and TradeStation. Our analysis process puts this data into several of our own databases that make the data available to our trading platforms via dll. The database we use is one of the new open source dbms called SQLMaestro.

 

As to Swarm Intelligence, thank you for the suggestion but we are aware of the swarmed agent approach and find the cascade approach offers us a higher degree of efficacy.

 

We understand entrophy analysis and have used it to detect "changes" in both price and the balance of commercial trade. Thank you for the suggestion.

 

Also may I thank you for your most eloquent and well written expression of on-topic, non-personal constructive criticism? I would humbly ask that you describe and demonstrate any comprable work of your own creation in the same level of detail as is in the text below:

 

ecubru;

 

The first difference between your chart and the one displayed below is that the bottom 2 sub-graphs on ours have price as NO part of their calculation. By far the most heavily weighted inputs to our trade decision making process are those that have to do with the balance, velocity and acceleration/deceleration of those trades that we classify as comming from commercial speculative trade.

 

The process of the construction of the chart shown below is:

 

The signals on the chart specifically the text that recommends precise TradePoints or No Trade are the product of several layers of processing.

 

The first layer gathers data from seven different time frames, posts it to global variables, and uses some of it to form an index of weighted biases as described in this thread.

 

Other of these variables are then passed as inputs to a set of intelligent agents that range from genetically optimized neural networks, functions producted by MARS (Multivariate Adaptive Regression Splines) and CART which is "a robust, easy-to-use decision tree that automatically sifts large, complex databases, searching for and isolating significant patterns and relationships." Both MARS and CART are products of Salford systems and some of the easiest to use tools for the development of intelligent agents that we have ever found. We have developed a tool in house that can convert the functions from MARS to Easy Language and can put a newly developed model on line and producing the Trade Points you see on the charts in a very short time. This entire process is described in this thread..

 

While price is considered, the bulk of our decision making process is based on various measures of what we call commercial speculation. We first measure our calculation of time-of-day normalized commercial presence. This tells us that at this time of day the commercial precence is some percentage of what it normally is at that same time of day over the previous 72 market days. If that percentage is less than a certain number we do mean reversion trading, if that percentage is over a certain percentage then the trend following methods shown on the chart below are shown. We also measure the percentage of time-of-day normalized commercial speculative bias in those 7 same time frames.

 

We think in terms of portfolios of both methods and time frames for each instrument we trade. Some of our indicators indicate trade and some indicate method or time frame.

 

Some of these indicators of commercial presence for both method and time frame can also be used for instrument selection. For example if the S&P is only trading at 70% of normal commercial presence then it is for sure that smart money has found at least a couple of stocks in the S&P that are trading at 200% of normal commercial presence that is where the action will be on that day in that index.

 

There are many ways to evaluate the price volume relationship, implement VSA, follow Wyckhoff or implement the Market Profile. The notes above roughly describe how we input and process the concepts and information from all of the above to output the precise TradePoints on the chart below.

 

All of the indicators on the chart below are calculated in real time and the TradePoints and text recommendations are posted at the first tick of the bar and are updated every tick. The small red dots are stops, blue are scalp profit targes the + is a recommended entry price for that bar. The recommendations are for the last bar on the chart.

 

 

 

 

cheers

 

UrmaBlume

[MtnDog 10]

My

 

First, some Wyckoff:

 

"Whenever you study the tape or a chart, consider what you see there

as an expression of the forces that lift and depress prices. Study

your charts not with an eye to comparing the shapes of the

formations, but from the viewpoint of the behavior of the stock; the

motives of those who are dominant in it; and the successes and

failures of the buyers and sellers as they struggle for mastery on

every move."

 

Game Theory:

 

"Expressed in the specific terminology of advanced theorists,

poker (TRADING) can be defined as an asynchronous, non-cooperative,

constant-sum (zero-sum), dynamic game of mixed strategies.

While the game is played in an atmosphere of common

knowledge and no player possesses complete knowledge, some

players are better able to process this common knowledge into

a more complete knowledge than are their opponents"

-Pat Dittmar, "Practical Poker Math"

 

For those of you who have never traded in a live pit (before electronic execution became dominant), it used to be relatively simple to figure out trade "intensity" and who were the "dominant" players. If the pit was empty and everyone was reading the newspaper or talking about the cubs/whitesox, then there was no intensity. It was also straightforward to determine what the motives of the big players were as well as order flow, paper (literally), etc. As a local, you paid big money to either lease or purchase a seat in order to gain this advantage.

 

Now, electronic execution has made it much more difficult to determine who is doing what/when. That is why it is necessary to employ advanced methods (neural, genetic, millisecond resolution, etc.) in order to at least get a rough idea of what is really going on in the market (especially in the ES).

 

This is definitely not the only way to trade and make $$. But if I am making discretional trading decisions, I want as much "edge" on my side as I can get. At some point there may be overkill and too much info to process, but that is up to the individual trader's style and temperment ( I knew successful pit traders who barely graduated high school as well as genius/savant types who could count numbers like Rainman).

 

I'll be honest and say that I've learned a lot from UB's posts, but you have to do the work for yourself, and it isn't the holy grail. At least I have become more cognizant of the fact that there is a lot going on down to the tick/millisecond level that I was never aware of.

 

A lot of trading is about confidence in your methodology. Of course there are alternative methods to measure intensity and "commercial" participation. Maybe just a simple candlestick chart with volume is enough for some. I rather enjoy "forensic" trading analysis (ie picking apart trades, volume.bid-ask, up-down tick, etc. etc.) It is tantalizing that all of the information is seemingly right there in front of you, if you can only put it all together in a bigger picture. That is where I think advanced computational data mining/analysis, neural, genetic, etc. becomes necessary.

 

Have a nice Holiday weekend, and take a minute to remember why we have the day off on Monday.

[boomerangas 10]

I wouldn't complicate things so much because you can't take blindly long/short signal on the indications of trade intensity. And if you can't you just are followin price action. You can't predict anything in this business. Let's assume that you are observing trade intensity spike. It means that agressive buiyng was met by agressive selling. But is there any edge? 5 min's later another big order from the longer time frame player could drive market much-much higher or lower just to the point where he want's to see this market and this spike could be nothing. As a real example it is a mini flash-crash of May 6. It would be very, very interesting to see all those buying spikes on UB charts all the way down. But price was driven to the point where someone wanted to see it. No matter what all divergences, intensities, deltas are showing. Pure order flow and price action.

Another example - you see that commercials are accumulating longs overnight, interest is rising. But someone sells to them? It could be retail trader or another, bigger, as Dalton describes, seller from the larger time frame, who will remove liquidity as on May 6 and market will crash in a matter of seconds. So if it is useless, why should one use it? Trade just a pure order flow, as a locals do, what is the diffirence is this order flow initiated by retail traders or commercials. By complicating things so much you could get "paralysis from analysis", trading is really easy if you have a good understanding of things such a price, value, under/over valued price and order flow. Really electronic markets have changed nothing here.

 

 

My

 

First, some Wyckoff:

 

"Whenever you study the tape or a chart, consider what you see there

as an expression of the forces that lift and depress prices. Study

your charts not with an eye to comparing the shapes of the

formations, but from the viewpoint of the behavior of the stock; the

motives of those who are dominant in it; and the successes and

failures of the buyers and sellers as they struggle for mastery on

every move."

 

Game Theory:

 

"Expressed in the specific terminology of advanced theorists,

poker (TRADING) can be defined as an asynchronous, non-cooperative,

constant-sum (zero-sum), dynamic game of mixed strategies.

While the game is played in an atmosphere of common

knowledge and no player possesses complete knowledge, some

players are better able to process this common knowledge into

a more complete knowledge than are their opponents"

-Pat Dittmar, "Practical Poker Math"

 

For those of you who have never traded in a live pit (before electronic execution became dominant), it used to be relatively simple to figure out trade "intensity" and who were the "dominant" players. If the pit was empty and everyone was reading the newspaper or talking about the cubs/whitesox, then there was no intensity. It was also straightforward to determine what the motives of the big players were as well as order flow, paper (literally), etc. As a local, you paid big money to either lease or purchase a seat in order to gain this advantage.

 

Now, electronic execution has made it much more difficult to determine who is doing what/when. That is why it is necessary to employ advanced methods (neural, genetic, millisecond resolution, etc.) in order to at least get a rough idea of what is really going on in the market (especially in the ES).

 

This is definitely not the only way to trade and make $$. But if I am making discretional trading decisions, I want as much "edge" on my side as I can get. At some point there may be overkill and too much info to process, but that is up to the individual trader's style and temperment ( I knew successful pit traders who barely graduated high school as well as genius/savant types who could count numbers like Rainman).

 

I'll be honest and say that I've learned a lot from UB's posts, but you have to do the work for yourself, and it isn't the holy grail. At least I have become more cognizant of the fact that there is a lot going on down to the tick/millisecond level that I was never aware of.

 

A lot of trading is about confidence in your methodology. Of course there are alternative methods to measure intensity and "commercial" participation. Maybe just a simple candlestick chart with volume is enough for some. I rather enjoy "forensic" trading analysis (ie picking apart trades, volume.bid-ask, up-down tick, etc. etc.) It is tantalizing that all of the information is seemingly right there in front of you, if you can only put it all together in a bigger picture. That is where I think advanced computational data mining/analysis, neural, genetic, etc. becomes necessary.

 

Have a nice Holiday weekend, and take a minute to remember why we have the day off on Monday.

[UrmaBlume 10]

Dog,

 

Thanks so much for the very kind words. A real treat these days.

 

And thanks for buying my book and thanks for the quote. Every author loves a quote.

 

 

Game Theory:

 

"Expressed in the specific terminology of advanced theorists,

poker (TRADING) can be defined as an asynchronous, non-cooperative,

constant-sum (zero-sum), dynamic game of mixed strategies.

While the game is played in an atmosphere of common

knowledge and no player possesses complete knowledge, some

players are better able to process this common knowledge into

a more complete knowledge than are their opponents"

-Pat Dittmar, "Practical Poker Math"

 

cheers

 

pat

[UrmaBlume 10]

Boomerangas,

 

I have suggested before that you not over-focus on measures of intensity.

 

We would never take a trade based only on measures of intensity or acceleration/deceleration inputs alone.

 

Our process includes 28 inputs each confirming others and our measures of intensity are far from the most heavily weighted inputs.

 

Also we would never trade on price or intensity alone. Of those 28 inputs, price is only 2.

 

cheers

[davewolfs 10]

For those who plan to pursue trade intensity along with UB's other 27 inputs in a weighted index. You might want to experiment with a basic consensus system such as:

 

Formula 301 – #4 Basic Consensus System | NeoTicker Blog

 

The key is obviously finding the inputs and the appropriate weights and combination's that apply to current market conditions.

 

The above combined with a tool such as Neo's grid optimizer to vary the weights given to each input (or input settings) can allow for rapid testing of trade ideas. I myself have had good success being able to test ideas with Neo's GO.

 

UB, if/when you have the time, I'd like to see some of your charts which show false positives and why the signal was ignored.

[ecubru 10]

Good post and I read your reply and links in detail.

 

However, and without wanting to have a long academic discussion, as in any theory, what is the relevance in justifying your assumptions with all this very interesting and well detailed technical models (seven different time frames, index of weighted biases, intelligent agents that range from genetically optimized neural networks, functions producted by MARS [Multivariate Adaptive Regression Splines] and CART, etc) if the exactly same results can be achieved using simply tools.

 

IMO, I am more interested in the results of a system rather than if the assumptions are technically sound (which can sound stupid in academic terms but not when you are a trader). In other words, and going back to the chart I post, a modified MACD with MAs with dynamical adjusted ATR levels (dots) in a constant volume chart for stops and profits show 100% the same results as a model with all these complex technical interrelations. I am not implying that the chart i posted or your system is good or bad, I only have this aftertaste:

 

1. if either all these technical explanations are just a way to rationalize the assumptions on your system

2. or maybe they are not useful at all (because they show the same results of basic TA indicators)

3. or maybe they are not taking advantage of their potential

4. or maybe you are 100% right and you are just defined and analyzing the key reasons behind price analysis.

 

I may overseeing something in your charts but whatever the case, as long as they show the same results as a simple system, they are only relevant from a theoretical perspective because in practice the result seems to be the same as when using a simpler system.

 

ecubru

 

ecubru;

 

The first difference between your chart and the one displayed below is that the bottom 2 sub-graphs on ours have price as NO part of their calculation. By far the most heavily weighted inputs to our trade decision making process are those that have to do with the balance, velocity and acceleration/deceleration of those trades that we classify as comming from commercial speculative trade.

 

The process of the construction of the chart shown below is:

 

The signals on the chart specifically the text that recommends precise TradePoints or No Trade are the product of several layers of processing.

 

The first layer gathers data from seven different time frames, posts it to global variables, and uses some of it to form an index of weighted biases as described in this thread.

 

Other of these variables are then passed as inputs to a set of intelligent agents that range from genetically optimized neural networks, functions producted by MARS (Multivariate Adaptive Regression Splines) and CART which is "a robust, easy-to-use decision tree that automatically sifts large, complex databases, searching for and isolating significant patterns and relationships." Both MARS and CART are products of Salford systems and some of the easiest to use tools for the development of intelligent agents that we have ever found. We have developed a tool in house that can convert the functions from MARS to Easy Language and can put a newly developed model on line and producing the Trade Points you see on the charts in a very short time. This entire process is described in this thread..

 

While price is considered, the bulk of our decision making process is based on various measures of what we call commercial speculation. We first measure our calculation of time-of-day normalized commercial presence. This tells us that at this time of day the commercial precence is some percentage of what it normally is at that same time of day over the previous 72 market days. If that percentage is less than a certain number we do mean reversion trading, if that percentage is over a certain percentage then the trend following methods shown on the chart below are shown. We also measure the percentage of time-of-day normalized commercial speculative bias in those 7 same time frames.

 

We think in terms of portfolios of both methods and time frames for each instrument we trade. Some of our indicators indicate trade and some indicate method or time frame.

 

Some of these indicators of commercial presence for both method and time frame can also be used for instrument selection. For example if the S&P is only trading at 70% of normal commercial presence then it is for sure that smart money has found at least a couple of stocks in the S&P that are trading at 200% of normal commercial presence that is where the action will be on that day in that index.

 

There are many ways to evaluate the price volume relationship, implement VSA, follow Wyckhoff or implement the Market Profile. The notes above roughly describe how we input and process the concepts and information from all of the above to output the precise TradePoints on the chart below.

 

All of the indicators on the chart below are calculated in real time and the TradePoints and text recommendations are posted at the first tick of the bar and are updated every tick. The small red dots are stops, blue are scalp profit targes the + is a recommended entry price for that bar. The recommendations are for the last bar on the chart.

 

 

Cheers