The Evolution of Market Profile Theory

[Frank 10]

Urma has stated his HUD indicator is derived from things that are happening sub-second. You have to decide is this really an area you want to compete with machines on?

 

The more important issue for me is the part about an indicator marking a good lasting high or low for a material amount of time, if not for the day. Now this is an area that is right from market profile --- finding good location vs intraday value --- I would argue that you should be thinking about where value WILL BE for the day, not where it currently is -- per the histogram -- but this is separate topic. To the extent a micro timeframe indicator helps see that -- hey, I am all for it -- but building an indicator like this as your main weapon of choice is only going to work for a select few.

 

I would just say that if I were Goldman Sachs and I heard retail traders were trying to build sub-second indicators to compete with our raw compute power, I would be jumping up and down for joy.

[MightyMouse 12]

Urma has stated his HUD indicator is derived from things that are happening sub-second. You have to decide is this really an area you want to compete with machines on?

 

The more important issue for me is the part about an indicator marking a good lasting high or low for a material amount of time, if not for the day. Now this is an area that is right from market profile --- finding good location vs intraday value --- I would argue that you should be thinking about where value WILL BE for the day, not where it currently is -- per the histogram -- but this is separate topic. To the extent a micro timeframe indicator helps see that -- hey, I am all for it -- but building an indicator like this as your main weapon of choice is only going to work for a select few.

 

I would just say that if I were Goldman Sachs and I heard retail traders were trying to build sub-second indicators to compete with our raw compute power, I would be jumping up and down for joy.

 

Goldman makes its trading profits from trading gray area inside information. Otherwise, their trading profits are mediocre.

[UrmaBlume 10]

Peter has always stressed the importance of the balance of order flow as an indication of future prices.

 

We use a harmonic of short term money flow to measure the extent of the imablance of trade.

 

In the chart below the extent of an imbalance is mesured by how far apart the lines are. If the top line is blue, orderflow is imbalanced to the buy side and vice versa.

 

If the +'s are blue, orderflow carries a buy bias and a sell bias if they are red. We use this indicator in conjunction with other money flow metrics.

 

[Frank 10]

HUD for mere mortals

 

[Frank 10]

I saw discussion about TPO's --- and while I have never used Market Profile as a trading method (I use it for higher timeframe context) -- it seems to me that TPO's are in one way superior to volume. I am not discounting the importance of volume -- just stating that 'imbalance' is best observed in range -- and range is same thing as TPOs. Thus, high TPO count may be a sign of imbalance where high volume may not be.

 

comments appreciated.

[davewolfs 10]

Peter has always stressed the importance of the balance of order flow as an indication of future prices.

 

We use a harmonic of short term money flow to measure the extent of the imablance of trade.

 

In the chart below the extent of an imbalance is mesured by how far apart the lines are. If the top line is blue, orderflow is imbalanced to the buy side and vice versa.

 

If the +'s are blue, orderflow carries a buy bias and a sell bias if they are red. We use this indicator in conjunction with other money flow metrics.

 

 

Thank you for the clue. I just need to work on the subtleties now.

 

[become 10]

Okay here it is... in a standard indicator in the Market Delta Software. Results directly mimic those of the one I was duplicating, with less noise. Although I have a feeling it will be attacked by the original creator (Which I can understand, after having attracted so much forum attention with it)

 

 

I'm not familiar with Market Delta, but isn't that just Total transactions / time elapsed on a per-bar basis?

[davewolfs 10]

Last night I got the smoothing worked out. I'd be curious to know what type of smoothing you are using UrmaBlume. As you said, this does lead price.

 

Edited September 29, 2009 by davewolfs

[clmacdougall 11]

UB, I'd like to talk about why any form of traditional MP does not work, because it believes that the market is static (stationary) and in fact it is more dynamic than what either a TPO or Volume based histogram can capture.

It's like believing that if you took a picture of someone at some point in time and walked away then returned a day later that you will find that person to be at or near the place where the picture was taken when you return.

When you trade MP you are believing too much in the static/stationary properties of the market which are simply just not there. The market seems to be somewhat like a person with little to no short term memory. Sometimes yesterdays value area means something to the market but most days it's not relevant.

I'm discovering the need to find a balance between a small lean on value and a greater lean on the "now" of the market. Your software seems to uncover this needed balance, but does your software have any initial lean based upon all you've learnt from the past concerning MP?

[UrmaBlume 10]

UB, I'd like to talk about why any form of traditional MP does not work, because it believes that the market is static (stationary) and in fact it is more dynamic than what either a TPO or Volume based histogram can capture.

It's like believing that if you took a picture of someone at some point in time and walked away then returned a day later that you will find that person to be at or near the place where the picture was taken when you return.

When you trade MP you are believing too much in the static/stationary properties of the market which are simply just not there. The market seems to be somewhat like a person with little to no short term memory. Sometimes yesterdays value area means something to the market but most days it's not relevant.

I'm discovering the need to find a balance between a small lean on value and a greater lean on the "now" of the market. Your software seems to uncover this needed balance, but does your software have any initial lean based upon all you've learnt from the past concerning MP?

 

When you start talking about "initial lean" you are starting to talk about psuedo science.

 

Whether the time frame is days, intra-session or sub-second most of our trading is of a single modality. The trades are based on bias and time/volume frame. The method is to fade a shorter term thrust into a stronger, longer term opposite bias. That is to sell a short term up-thrust into a longer term down bias.

[davewolfs 10]

When you start talking about "initial lean" you are starting to talk about psuedo science.

 

Whether the time frame is days, intra-session or sub-second most of our trading is of a single modality. The trades are based on bias and time/volume frame. The method is to fade a shorter term thrust into a stronger, longer term opposite bias. That is to sell a short term up-thrust into a longer term down bias.

 

You've been sharing quite the hints lately UrmaBlume.

[clmacdougall 11]

When you start talking about "initial lean" you are starting to talk about psuedo science.

 

Whether the time frame is days, intra-session or sub-second most of our trading is of a single modality. The trades are based on bias and time/volume frame. The method is to fade a shorter term thrust into a stronger, longer term opposite bias. That is to sell a short term up-thrust into a longer term down bias.

 

UB, please explain to me the difference between the "lean" I mentioned and the "bias" which you have mentioned. I'm not catching the sense of what you're trying to relay with that term. Much thanks.

[davewolfs 10]

UrmaBlume,

 

I understand that you look at balance on multiple time frames, i.e. 2,4,8 minutes etc...

 

May I ask if you ever look at balance of trade on volume bars, for example I have seen that you have posted charts using 8000 bars, but have you ever used say 16000, or 32000 bars? I'm finding that the results here can be equally useful unless my calculations are wrong.

 

The idea for looking at balance on these larger volumes frames is based on the same principals of MP and balance areas. When the market comes into Balance on these large volume levels the same idea holds where price is easier to be moved out of balance. As Dalton puts it "Imagine a large stone balanced on a mountain peak...the stone might come loose and tumble down the mountainside". The key is understanding where that mountain peak exists and looking for clues on smaller levels to determine how hard other time frames are attempting to move that stone or what is the path of least resistance to move that stone.

 

I would assume that your preference is time since this is probably what most of the world looks at before placing a trade but I am curious to know if the same principals can also be applied using another unit of measurement. I appreciate any input and thank you again.

[UrmaBlume 10]

UrmaBlume,I understand that you look at balance on multiple time frames, i.e. 2,4,8 minutes etc...May I ask if you ever look at balance of trade on volume bars, for example I have seen that you have posted charts using 8000 bars, but have you ever used say 16000, or 32000 bars? I'm finding that the results here can be equally useful unless my calculations are wrong. The idea for looking at balance on these larger volumes frames is based on the same principals of MP and balance areas. When the market comes into Balance on these large volume levels the same idea holds where price is easier to be moved out of balance. As Dalton puts it "Imagine a large stone balanced on a mountain peak...the stone might come loose and tumble down the mountainside". The key is understanding where that mountain peak exists and looking for clues on smaller levels to determine how hard other time frames are attempting to move that stone or what is the path of least resistance to move that stone.I would assume that your preference is time since this is probably what most of the world looks at before placing a trade but I am curious to know if the same principals can also be applied using another unit of measurement. I appreciate any input and thank you again.

 

Dave,

 

As to Bars and Time Frames:

 

While we do believe that it is important to stay in touch with higher time frames, how much higher depends on where you execute. At some point much higher just adds lag. For our human traders in ES the 16k bars are our highest and 250 the lowest.

 

We use NO time bars and we use NO tick bars. In the price time volume continuum time is only a marker - price is motivated by trade, propelled by an imbalance in volume and time has nothing to do with it. Of all the different bar structures we find the tick bars to be something worse than useless - they can be misleading.

 

Knowing that it is an imbalance that drives price, bars that treat each transaction equally regardless of size make it impossible to measure that imbalance with any degree of precision. Tick Bars give transactions of 1,000 contracts the same weight as a transaction of a 1 lot and that can easily be misleading.

 

As to your reference to the public/retail trader "is time since this is probably what most of the world looks at" we don't care or notice this trade. Public/Retail trade never leads the market, never turns the market and loses tons of money chasing the market. You may have noticed that much of our work is about identifying and classifying COMMERCIAL trade.

 

As to Jim Dalton, I found him to be a very nice guy but too linear to evolve much past what he was doing 30 years ago.

 

Market profile is invaluable as a theory but almost usless as a method. Almost every commercial trader I know has spent tens of thousands of dollars learning market profile but only 1 of them uses it as a method.

[davewolfs 10]

Dave,

 

As to Bars and Time Frames:

 

While we do believe that it is important to stay in touch with higher time frames, how much higher depends on where you execute. At some point much higher just adds lag. For our human traders in ES the 16k bars are our highest and 250 the lowest.

 

We use NO time bars and we use NO tick bars. In the price time volume continuum time is only a marker - price is motivated by trade, propelled by an imbalance in volume and time has nothing to do with it. Of all the different bar structures we find the tick bars to be something worse than useless - they can be misleading.

 

Knowing that it is an imbalance that drives price, bars that treat each transaction equally regardless of size make it impossible to measure that imbalance with any degree of precision. Tick Bars give transactions of 1,000 contracts the same weight as a transaction of a 1 lot and that can easily be misleading.

 

As to your reference to the public/retail trader "is time since this is probably what most of the world looks at" we don't care or notice this trade. Public/Retail trade never leads the market, never turns the market and loses tons of money chasing the market. You may have noticed that much of our work is about identifying and classifying COMMERCIAL trade.

 

As to Jim Dalton, I found him to be a very nice guy but too linear to evolve much past what he was doing 30 years ago.

 

Market profile is invaluable as a theory but almost usless as a method. Almost every commercial trader I know has spent tens of thousands of dollars learning market profile but only 1 of them uses it as a method.

 

Thank you for the response UrmaBlume. Just to touch on the point that no time or no tick bars are used. May I ask why Time is displayed in the HUD you have developed. Is the purpose of this simply to gauge the markets in direction in terms of how well trade is being facilitated? Regardless, wouldn't all of this information be detectable through the use of the other methods we have discussed i.e. price volume time continuum and if so, then why not just use that?

Edited September 30, 2009 by davewolfs

[Guest AI3000]

Last night I got the smoothing worked out. I'd be curious to know what type of smoothing you are using UrmaBlume. As you said, this does lead price.

 

 

Do you care to share?

 

Much appreciated

[clmacdougall 11]

UB, after you figured out that distributions / day types / initial lean meant nothing can you let me in on what was your next progression forward from the static MP concepts that you find in all books available?

[UrmaBlume 10]

UB, after you figured out that distributions / day types / initial lean meant nothing can you let me in on what was your next progression forward from the static MP concepts that you find in all books available?

 

Cory,

 

Slow down a bit mate.

 

I value everything I have learned about market profile and market profile theory. It is just that we have found it to be a better market theory than it is a method of trading.

 

We believe that today's data streams, data processing and certain intelligent agents now make it possible to more precisely define such values as the location of intense commercial beyond the day time-frame trade, the balance of order flow and the velocity of trade, rejection and everything else Peter taught was important to predicting price as far back as 1980.

 

Everybody here is familiar with the "public" version of our measure of the location and intensity of commercial trade. The balance of trade is demonstrated in our NetMoneyKG, Net New Commercial Trade, MoneyHarmonicKG and the pie charts on the HUD. The velocity of trade is demonstrated on the HUD and calculated with millisecond granularity.

 

The next step you ask about is to be better able to more precisely calculate the variables that Peter so rightly taught us are valuable in predicting future price and define a trading protocol based on those inputs that is operative in the widest possible range of markets, instruments and time frames.

 

cheers

 

cheers

[clmacdougall 11]

UB, would it be correct to say you identify the VAL and VAH in the now by indentifying commercial buying and selling as it happens? Is this what you use to identify your balance area extremes until they are broken through?

[justlurkin 10]

Okay here it is... in a standard indicator in the Market Delta Software. Results directly mimic those of the one I was duplicating, with less noise. Although I have a feeling it will be attacked by the original creator (Which I can understand, after having attracted so much forum attention with it)

 

 

Results speak for themselves.

 

 

become,

For non-I/RT users it seems that this would simply be ( (TotalTicks * TotalTicks) / Bar Time in Seconds ) according to Investor/RT Tour - Volume Breakdown (VB)

 

However, I don't see results anywhere near UBs indicator.

 

Tooker,

Can you post a pic of todays action with the indicator. I'm going to have to run replay over the weekend to post a pic.

 

ps: TotalTicks = Total Trades (not volume) for the current bar (1000 Volume in this case)

[clmacdougall 11]

UB, I'm guessing that the point of this thread is the discussion our own personal evolution outside of the previous written concepts concerning MP. You've made it clear that you're not going to reveal the concepts involved in your trading system so then I'd like to move the discussion away from your computer program and more towards your theoretical evolution outside of the old concepts/methods. Can you point out the flaws you discovered concerning the old concepts and point us towards what you have now discovered to be worthwhile.

I can only speak for myself when I say "thank you" for being the only one around who publicly is allowing us to see the inside of what might be the way that commercial funds trade on a daily basis and it's a bit humiliating to think of just how behind the times we might be with our application of MP principles.

Leaving your computer system behind, can you give us a little bit of the professional/commercial understanding of the application of MP today. Thanks.

[UrmaBlume 10]

UB, I'm guessing that the point of this thread is the discussion our own personal evolution outside of the previous written concepts concerning MP. You've made it clear that you're not going to reveal the concepts involved in your trading system so then I'd like to move the discussion away from your computer program and more towards your theoretical evolution outside of the old concepts/methods. Can you point out the flaws you discovered concerning the old concepts and point us towards what you have now discovered to be worthwhile.

I can only speak for myself when I say "thank you" for being the only one around who publicly is allowing us to see the inside of what might be the way that commercial funds trade on a daily basis and it's a bit humiliating to think of just how behind the times we might be with our application of MP principles.

Leaving your computer system behind, can you give us a little bit of the professional/commercial understanding of the application of MP today. Thanks.

 

 

Cory,

 

Thanks for the kind words.

 

When you mentioned "evolution outside of the old concepts/methods" you might be putting concepts and methods in the same bin. The concepts as taught by Peter in the early 80's were about order flow, locating commercial trade, responsive trade, initiating trade, rejection and most important of all the balance of trade. All of these I put in the concept bin.

 

TPO's, Value Area, VAH, VAL and WVAP all go in the method bin. Peter's early seminars were much more about understanding the concepts than implementing the trades.

 

A couple of his early students, namely Jim Dalton and Kevin Koy, declared themselves masters of the concepts and started putting out material more about their idea of methods than Peter's idea of concepts. Kevin and Peter were once partners, the stories about their breakup have caused some to hold Kevin in perpetual disrespect.

 

The whole basis of the market profile graphic is that time at price should approximate volume at price. The basis of TPO counts is responsive vs initiating trade. The idea was that if TPO's accumulate in the upper part of the value area it is the result of the responsive seller being in charge at that area of price.

 

Then there was no online volume not to mention any way to break that volume down into buying and selling volumes or into responsive or initiating trade.

 

The weaknesses of the profile graphic when it comes to practical application/METHOD today are:

 

1. No accurate way to measure commercail participation until after the close.

[UrmaBlume 10]

UB, I'm guessing that the point of this thread is the discussion our own personal evolution outside of the previous written concepts concerning MP. You've made it clear that you're not going to reveal the concepts involved in your trading system so then I'd like to move the discussion away from your computer program and more towards your theoretical evolution outside of the old concepts/methods. Can you point out the flaws you discovered concerning the old concepts and point us towards what you have now discovered to be worthwhile.

I can only speak for myself when I say "thank you" for being the only one around who publicly is allowing us to see the inside of what might be the way that commercial funds trade on a daily basis and it's a bit humiliating to think of just how behind the times we might be with our application of MP principles.

Leaving your computer system behind, can you give us a little bit of the professional/commercial understanding of the application of MP today. Thanks.

 

 

Cory,

 

Thanks for the kind words.

 

When you mentioned "evolution outside of the old concepts/methods" you might be putting concepts and methods in the same bin. The concepts as taught by Peter in the early 80's were about order flow, locating commercial trade, responsive trade, initiating trade, rejection and most important of all the balance of trade. All of these I put in the concept bin.

 

TPO's, Value Area, VAH, VAL and WVAP all go in the method bin. Peter's early seminars were much more about understanding the concepts than implementing the trades.

 

A couple of his early students, namely Jim Dalton and Kevin Koy, declared themselves masters of the concepts and started putting out material more about their idea of methods than Peter's idea of concepts.

 

The whole basis of the market profile graphic is that time at price should approximate volume at price. The basis of TPO counts is responsive vs initiating trade. The idea was that if TPO's accumulate in the upper part of the value area it is the result of the responsive seller being in charge at that area of price.

 

Then there was no online volume not to mention any way to break that volume down into buying and selling volumes or into responsive or initiating trade.

 

The weaknesses of the profile graphic when it comes to practical application/METHOD today are:

 

1. No accurate way to measure commercail participation until after the close.

 

2. No way to note/locate locally intense commercial activity within the value area.

 

3. No precision in the calculation/presentation of total volume, buying volume, selling volume, order flow or the balance of trade.

 

Progress/Evolution for market profile practioners is about tools, technologies and protocols that 1)Do a better job of more accurately measuring commercial activity in real time 2) Precisely locate locally intense commercial activity withing the value area and 3) more precisely calculate/present and locate total volume, buying volume, selling volume, order flow and the balance of trade than does the market profile graphic.

 

Our indicators of Commercial Intensity, Net New Commercial Trade, Harmonic of Trade, etc., are all based on our effort to take the practical application of market profile theory beyond the capabilities of the market profile graphic.

 

As an old man who has seen all of this evolve from the perspectives of a retail trader, commercial trader, brokerage managment, premium arbitrageur, high-tech trader and trade decision support technologies developer over the last 3 decades some observations are that:

 

The Strengths of today's beginning trader are that he is smarter, more well-educated and has access to better data and technologies than in the past. The passion is equal as w/o passion players are quickly eliminated both then and now.

 

One weakness of tody's up and comers is a rush to method without a firm foundation in the basic concepts of how markets work and what drives price.

 

Whether it is on this board, the markets or a poker table, when there is opportunity, fear, greed, risk, reward or personal face involved - everybody gets a little naked. Here my bet is on those who come here to learn, discuss, contribute and grow and not on those who are here as a path to quick and easy money. The tone of the posts is a great classifier - some are positive, responsive, inquisitive, offer help and make contributions and others are negative, personal, never contribute and leech code. Makes it easy to pick the winners.

 

Only a very few will succeed and what separates that very small percentage from the rest of them is not about financial resources it is about attitude and intellectual capital.

 

And Cory, from our phone calls and your notes and posts and while I think you have a bit of a journey in front of you I think you have the attitude and proper capital to be one of those who make the trip.

 

cheers

Edited October 10, 2009 by UrmaBlume

[clmacdougall 11]

UB, please let us in on when your new book/books are going to be released. I'm sure by the response to your threads that plenty of us are interested in what you'd have to say. Is there any kind of a preview or table of contents from your book that is presently available for viewing? Thanks again for graciously taking the time to reply to my questions!

[davewolfs 10]

Do you care to share?

 

Much appreciated

 

No point, I've discovered the harmonic is only good to look at after the fact.