Building An Index of Market Profile Concepts

[Frank 10]

Thought would just try something out --- what if we were to share a few ideas that drive your thinking on the day. I am putting this in the market profile segment because Market Profile is a good theoretical place to start.

 

In quantitative finance, there are structured methods to translate qualitative views into quantitative scores -- for instance, let's say you have a method for trading the opening 30 mins of the day. You might see a really good long set-up X% of the time, a really good short set-up Y% -- a solid long or short set-up Z% of the time -- and you would have 'No View' the

remainder of the time.

 

So you could set up a column in a table like this:

 

View:

Very Positive

Positive

No View

Negative

Very Negative

 

A way to quantify these qualitative views would be to to assign scores in a standardized way.

 

An Example of such would be:

 

View: Probability Score

Very Positive 11% 1.73

Positive 22% 0.87

No View 33% 0.00

Negative 22% -0.87

Very Negative 11% -1.73

 

Note that the 'Score' Column has a Mean of Zero and a Standard Deviation of 1.0

 

Now that we have possible 'starting point' for a possibly well-structured trading process --- we can discuss the meat of this thread. That is, what goes into the 'Index' that creates a 'View'? And then ultimately, it will be up to the trader to decide the 'weightings' to put on any components to an index.

 

---------------

 

First Concept:

 

The opening 30-min bar

 

1) Is this bar showing directional conviction? Is the bar showing movement vs the previous days range or is the bar 'inside' relative to the previous days range? If the bar is not able to escape the previous days range --- you make have reason to have incremental bias to be in a day where price does not trend -- and instead 'brackets' / overlaps around a point of control. (don't box yourself in to traditional 'book method' market profile with these terms -- just use the theory).

 

Hence, a more detailed analysis of the opening 30-min trade might be one possible component to an 'weighted index' of qualitiative/quantitative biases.

 

ok, enough from me --- let's see if this takes or not.......

 

Frank

Edited August 14, 2009 by Frank

[traderlife1520887472 10]

Watching the first 30 min bar would be similar to keeping in mind the Initial Balance in relation to previous day's range, etc?

 

I have a feeling you got some good insights to keep in mind, so please do go on.

[Frank 10]

If you take all 30-min pit session bars historically and compare them to that days final median 30-min bar RANGE, what you find is that the 1st and 2nd 30-min bars are a lot wider (more range) than the rest --- except the last two.

 

That said, it is statistical fact that once you get out 4-7 bars into the day, the market will always spend time overlapping prices -- even on very strong directional days. But you will have serious drawdowns trying to fade a market that is showing extreme strength in the opening 30 mins that is leaving the prior days range. As a general rule, you want to save your fade plays for days when the market is not leaving the prior days range 'early in day' (this is one thing of many to consider)... This is a quick adjustment that you have to make in real-time but its crucial.

 

Note that in a strong trending type of day (I didn't say 'trend day' -- I just mean a good directional MORNING move) -- many times ANY price that you got during the opening 30-mins that 'goes with' the flow of B will be a profitable trade. I am not saying 'go with every opening thrust' -- not at all --- I am just saying to be cognizant of a market that is driving OUT of previous days range. Note that yesterday, the market did a wide bar DOWN in the opening 30 mins --- but it wasn't even close to exiting the prior days range. In that example, the market reversed and drove up.

 

Regarding the 'initial balance' --- the size of the opening bar and the amount of extension beyond the opening bar is the I.B. As a baseline --- expect this 'violation' to be 4-20 ticks. So a 6.5 pt opening bar and a 2 point extension beyond that makes for a 8.5 pt IB (note that if you get an 'outside 2nd bar' will be just the size of the 2nd bar).

 

I guess what I am getting at is this --- you are trying to figure out whether to 'go with' the opening momentum --- or fade it. At this time of day, you expect some good range in the bars (some of best in the day) -- but you don't know if market is going to trend for a bit --- or not. You know that in vast majority of cases, there is a 'deadline' -- where price WILL start to overlap. So you want to either 'go with' the opening momentum and exit as the overlapping prices (value creation) begins -- OR you want to fade the market -- expecting the strong tendency to overlap price to kick in and trap the momentum chasers and allow your 'location trade' to work as these trapped players are forced to unwind.

Edited August 14, 2009 by Frank

[traderlife1520887472 10]

Frank,

 

If I understand what you said correctly, most of the opening directional moves that happen with conviction, have a higher probability of being a solid move or trend if they move outside of the range of the previous day.

 

How do you personally go about positioning yourself for such a move?

 

Looking at ZB, it opened right at the previous day's close 118060, but beneath the yesterday's high of 118110, and proceeded to do what you described for a trend day. Mkt paused around 118110 before taking off again.

 

what I'm asking is where you conceivably would have made your move as the market was unfolding. You can use ES examples as well if you like - I trade ZB and ES

 

Best,

 

Ben

[Frank 10]

usually, I like to just find a little retracement during the opening bar and then perhaps even add on a stop order as this trade should work right away. But I am not attempting to hold on for a trend day -- just get in and get out for a few points. If something really special seems to be going on, ie the time & sales is showing unusually robust buying or selling --- or the market just blasts through a well-defined support/resistance price with momentum -- then its 'sometimes' (read rarely) a good idea to hold on and ride out the first correction for a continuation move ... but those are rarer cases. The default idea is just to trade out of the position --- the reasons for which I will continue with in another post. That being --- what % of time does the market 'bracket'? (ie, show overlap) hint: it does EVERY day, even trend days. (talking about ES now)

[rigel 10]

Frank,

Having read you posts on the Taylor thread, your insights are always appreciated.

 

1. When you talk about retracement on the opening bar, are you referrring to a 30min chart ie. 30min bar.

 

2. If so when you say market brackets, are you talking about overlap of 30min bars during the day.

 

Perhaps you could illustrate with a chart, it would be much better to understand on a visual basis.

[traderlife1520887472 10]

Frank,

 

Thanks for your missive on things to look for on the open. I appreciate the learning process and the wisdom you provided.

 

I'm not quite getting what you mentioned as IB... I thought it was the range of first hour of trade.. if small - easily turned over and usually moves out of small IBs go for a bit. wider IBs signal rotational tendencies.

 

I guess I'm asking for a clarification of how you see IB and if my 1 hr definition range is too simplistic...

 

I think maybe you're saying that first 30 mins direction and any range past that direction is the IB... then price inevitably will overlap back into the IB... and from there, one can gauge moves that again move outside of the IB based on volume and rate and size of aggressive orders and choose whether to go with that flow...

 

 

 

As for the second concept you had in mind, please go on...

 

 

Thanks,

 

Ben

[Frank 10]

hey traderlife,

 

I am out of town and away from my primary computer so I will likely delay posts for a bit -- but will be back in a bit.

 

Regarding initial balance, I don't really use it except as a measure of range during that period. Initial balance is not something that I am focused on.

[Frank 10]

Concept #2

 

In the book 'The Theory of Poker' by David Sklansky -- he writes;

 

["Beginning poker players sometimes ask "What do you do in this particular situation?" There is really no correct answer to that question because its the wrong question.... The right question is "What do you CONSIDER in this particular situation before determining what to do?"]

 

So, what I am saying here is something to 'consider' -- which takes into account the market profile concept of VALUE. This is a Sklansky type of thing -- don't take this too literally, its highly situational.

 

What is value? One way to define it is that value is where price trades more than any other price (the statistical 'mode'). I won't go into this topic any more than this because that is not the purpose here.

 

Instead, let me just present this idea --- that is --- price will show significant 'overlap' on the vast majority of days. Yes, there is the occassional outlier -- but nevertheless, overlapping price is inevitible on every day. This is not to say every move is a fade -- hardly -- I am just presenting a concept to CONSIDER.

 

Let's look at some statistics that support this idea --- over the past ~2 years, there has been a price on the 30-min chart that has shown 4 30-min bars of 'overlap' 99+% of the time (last 536 trading days). Not only that, the S&P futures have actually shown 4 bars of overlap on 96% of days by 1pm.

 

Thus a concept born from actual market generated statistics might be.... Expect the market to form some type of tentative 'value' (overlapping price) by lunchtime and take this market-generated information and put it into proper context of all the other existing factors that you are considering at the time to handicap if price is leading value (trending behavior) - or is price not showing momentum and therefore the market is likely to 'bracket'?

 

Here are some stats of 30-min bar overlap at the end of each bar.

 

[Frank 10]

Using this mornings action as example of this concept.

 

Consistent with just about every day, the market found short term value after a directional attempt -- 'value traders' can use this information to incorporate into overall mosaic. Clearly, the market was not as weak as it probably appeared to many -- the market found value within the opening 30-min range and did relatively narrow C and D bars (10am - 11am EST), which eventually led to a major bear trap.

 

[Frank 10]

Today a good example of a market that does NOT overlap in the morning session. Instead of the typical overlap --- a very large 'E bar' (11 -11:30 EST) shows a trending move. It is very important to not get caught repeatedly trading against a market showing momentum. You can tell this very clearly by keeping a watch on the 30-min chart.

 

Thus concept #4) Respect and seek trades 'with' a market showing big bar momentum. Today was tricky to actually find an entry -- despite being quite clear one-sided momentum based selling.

 

 

[Frank 10]

Just wanted to show a somewhat similar profile from July 13, 2009

 

Note the similarities:

 

Big Rinse washout/fakeout 1 direction in C --- then no overlap and big bar (this one in F vs todays that occured in E) as HARD auction the other way launches big move.

 

[ant 10]

Respect and seek trades 'with' a market showing big bar momentum. Today was tricky to actually find an entry -- despite being quite clear one-sided momentum based selling.

 

Hi Frank,

 

I agree with you here. I think the big bar has to be taken in context though (i.e., make sure it isn't traders throwing in the towel which might create a spike signaling the end of buying/selling), but today was a go-with the market. I think you had two opportunities today, the single prints out of balance in B period (also a big bar) and the D/E period as you pointed out. As you said, these are tricky trades because these are cases where you simply have to put the trade on and then monitor for continuation. With this type of urgency in the market, the market doesn't provide exact entry points. Although, once the market re-entered yesterday's range for the second time in D period, you knew the market would go for the lower gap next and that the potential for an outside day was high. Just put the trade on and monitor for continuation. At the very least, if a trader is not comfortable going with the market, the trader should have at least stayed away from fading the market. The next destination for this week should be the balance area low at 975.50. The market doesn't have to do that, but that's where the odds are pointing.

 

Regards,

Antonio

[Frank 10]

good point on the re-entry of the previous days range for 2nd time. fits a little bit with my 1st concept in an indirect way --- today, the market at first appeared to be showing strong movement away from previous days range in B today -- but this was a trap, just like it was back on July 13th as it rinsed one direction and trapping those who didn't exit out in C (outside the range of B).

 

note that there is a clear and obvious tip-off that the '4 overlapping bars' would be delayed relative to 'normal' when D closed on its low -- and then E showed further momentum beyond the low of B. this was also a first-hour breakout after an extremely narrow pit-session bar on Monday -- which is normally a 'go-with' in MP anyway.

 

but clearly when a bar closes on its low as it did in D -- then E elongates away from that weak close, the 'market generated information' is telling you something. A test up in F would have been another place to short -- but instead it went further lower --- and then you did get the 4 overlapping bars by i -- consistent with the stats I posted - if you don't get the 4 overlapping bars early on -- then you WILL get 4 overlapping bars by i (this is 99% over past 2+ years) -- so when F does not test up and does not elongate further down --- you know that the move is mature and needs to rest.

 

There is power in combining the statistical tendencies with a common sense read of the bars. If you back the chart up to the end of E (see below) -- given the low close (marked by the italic 'E') -- if you just think logically --- at which price would you now be expecting a 4 overlap price? its pretty clear that there is no price to overlap on anytime soon --- and that in real-time, you know that this is consistent with a trending market and not a 'bracketing' market --- that is, you KNOW there is going to be a 'delay' to the 'normal statistical tendency' and you are in a 'special' type of momentum day.

 

here is how the chart looked at the end of E -- just to point out the obvious...

 

Edited September 2, 2009 by Frank

[ant 10]

this was also a first-hour breakout after an extremely narrow pit-session bar on Monday -- which is normally a 'go-with' in MP anyway.

 

I don't pay attention to the IB, but I think that that would have been the second entry, if one didn't get in as we traded back into yesterday's range. The reason I like that area around the IB low is because that coincides with the re-entry back into the previous balance area around 1016 and near the 6-day balance area low at 1013.25. I also would have had an exit nearby.

 

A test up in F would have been another place to short

 

Personally, I don't like that entry. Price was too far down for me and I didn't have an exit nearby. Just not my style, it feels like I would be chasing the market.

[Frank 10]

I see your perspective and I won't chase in general -- but this is the situation, (where price is leading value and value is clearly in process of a migration) --- where I have noticed that there is generally an 'overshoot' --- that is, even if value does end up forming somewhere within the range of the momentum bar (in this case within the price range of E), there will STILL be a move below that price. I will get to the statistics that support this in the future -- but it is a strong tendency and definitely tradeable ---not as a hold, just for a 'get a short on ABOVE projected value' and 'get out at/below the projected value area.'

[Frank 10]

Follow up from earlier post for todays action.

 

So we know that the market has strong tendency to pick a price in the opening few 30-min bars range and 'overlap there' --- I have quantified this tendency in stats above. Unless there is strong reason to believe otherwise, this should be your baseline expectation. These stats are very strong.

 

So in real-time environment, some real-time logic can really add to this statistically based framework.

 

Today -- the market tried down --- is that consistent with higher timeframe selling? Didn't seem like it given that C had large overlap with B -- but whenever you have a big bar, you must start to consider the fact that market generated info may be telling you something. In this case -- you can watch how the volume builds out within the big bar, how the big bar closes --- and what happens in the subsequent bar. Any trade you do should be consistent with that real-time logic. You may have to scratch if conditions change -- but you should have a thesis and monitor that thesis as more information streams at you.

 

well, volume built up well off the low of the C bar and D pressed back up towards the low of B. This creates a possible '4 overlap price' at 96.25 -- hey, that fits what the statistics would tell you.... You know that '4 overlap price' is coming at somepoint -- it happens EVERY day. And it happens every day before i. That doesn't mean the market is always a FADE, it is just something to watch for and think about as you digest incoming real-time market generated information in the context of any higher timeframe structure...

 

As you get closer to the end of D --- it should be clear that this is headed back into the range of B (to the overlap price and probably a bit higher). Bottom line, the thesis of a 'bracket' is being supported increasingly by real-time information.

 

Edited September 3, 2009 by Frank

[ant 10]

Interesting stats Frank...

 

Your stats make sense since about 85% of the time the market is consolidating. On most days, your stats would stand correct, and I think that most traders trade with that expectation. However, the only caution I would add is that you have to be careful with the other 15% of the time when you get a trend day. These are the days that will really hurt short-term traders if they trade it the way they would a trading range. These are the days where traders give back a lot of profits and cause them to turn a profitable month into a losing one. Just being able to avoid fading a trend day would be extremely beneficial to a trader. Having said that, trend days do provide the most profit potential if a trader can identify them early on. I would be interested to see what stats you may have for determining a trend day.

Edited September 3, 2009 by ant
Corrected sentence.

[Frank 10]

I would be interested to see what stats you may have for determining a trend day.

 

That is really what this thread is all about -- taking concepts and putting them into context of the current day. The statistics I have presented relate to just one concept of many. It is the combination of all these concepts that will determine trend vs bracket, not any single statsistical set.

 

Those statistics I presented are not meant to say 'always fade the market' --- they are presented as historical fact. The bottom line is, the market overlaps a lot. That is NOT saying to fade the market every day as you may (or may not) have interpreted that to be what I was implying. That is not what I was saying when I presented those stats -- I was just trying to say -- ok, here is a statistical analysis of the past -- lets apply it to this particular situation and see how we do against that benchmark from THIS particular angle (one of many) and then lets take that and create an overall 'mosaic' out of these concepts.

 

Thus, trend days are when there is NOT overlap and RF shows higher confidence. I have stats on RF as another guide.

 

Believe me, I learned the hard way years ago --- I did what so many others have done --- you learn about oscillator divergences and make some money fading the market -- and think you've discovered gold -- and then a trend day comes along and you lose 2 weeks of profits, go on tilt, and give away more.

 

But 'going with' the market has its own hazards of getting chopped up, as you know.

 

For me, I am constantly trying to think about where value WILL BE... If the concepts point to trending action, then value is migrating and current price might be advantageous as to were value IS GOING. Gretzky had some line about skating to where the puck will be, not where it is --- its a cliche but its also right. If the market is NOT trending, then gauging value is not actually that hard, the price will be determined relatively early in the day and you can narrow your choices down as the day develops..

 

Bottom line, I want to make money on days that AREN'T trend days but still very often play IN THE DIRECTION of that first impulse because if it IS a trend day (or if its just a bigger F.U. than you were thinking) -- you are on the right side of this. And if you do get in well, you do have that chance that you are in a 'big bar' momentum move and that while you should exit MOST of the time, you always have that option if the tape just completely ignites with extreme ticks, large orders etc... can stick a stop in and lock in the gain and still participate in a trend day.

 

But once the market shows signs that its NOT a trend day -- which it will do on majority of days -- now you know that you will be cannon fodder if you try to play for trend. Doesn't mean there is a fade set-up necessarily --- maybe there just isn't a set-up either way as you chop in a tight range towards that high odds 9+ overlap -- like yesterday.

[ant 10]

Frank, thanks for your reply and starting this thread. I have enjoyed reading it. Your quote below answered my question.

 

Thus, trend days are when there is NOT overlap and RF shows higher confidence. I have stats on RF as another guide.

[Frank 10]

concept 1 -- market shows little overlap to previous day, testing into the previous days range in B and then every other closing 30-min bar was outside of it. this is bullish

 

concept 2 -- expect the market to print 4 overlapping bars by F majority of days, by 'i' every day -- it achieved this objective in E.

 

concept 3 -- I haven't talked about this but if you look at the european cash session - the US session basically NEVER does an inside day to that session. I won't post the stats but they are compelling that as an objective, the S&P's should take out either the globex high or low and then the other side most of time holds for rest of day.

 

so putting these concepts together and combining with previous days --- where we saw much overlap (balance) -- given that price was being 'accepted' outside of the previous days range -- the odds favored taking out the globex HIGH. Once momentum devleops up -- you can see clear set-up for 'double distribution trend day' -- one distribution representing the 'high odds overlap tendency' (the morning session overlap) and the second distribution forming after 'big bar momentum' outside and away from the previous days range while in 'breakout mode'.

 

the complex part --- as price overlapped on intraday basis-- it got tough to figure where to enter. It became clear as price pushed up in F then shot up in G --- only play was to go to market or buy stop. I managed to miss this entry and then it didn't pull back quite enough to get a long scalp on in H. Once it extends past the G high --- you are now likely ABOVE the projected second distribution -- so now in tough spot -- buying ABOVE projected value (at time, current value is obviously MUCH lower, but this is not important info -- as you should be using a 'projection of value' given the structure).

 

 

Edited September 4, 2009 by Frank

[Frank 10]

This has kind of evolved into a journal --- will post todays final distribution so I can look back at the result in future.

 

[Frank 10]

Market had serious issue at 1025.00 today -- this was major center of gravity of Aug 21 - Sep 1 Volume distribution:

 

[Daniel Yashaev 10]

Hi Frank, as I know u use Open E Cry platform.... so I am just wondering what program do u use to see market profile?

[Frank 10]

hi Daniel,

 

Market profile is actually very simple stuff --- I like the customization offered by doing it yourself -- as you can customize a dashboard exactly as you like it. for example, I have added an overlapping bar indicator below my chart - then I put Rotation factor and range indicators in --- and I can then conditionally format any data to be bold or change colors or make a sound or whatever --- when they meet a criteria. I am using Open Ecry for things other than market profile --- and then just as a data source for my own customized dashboards. Since I am not a programmer, I do it in Excel. This takes some effort but its surprisingly simple once you have a feel for how to visually structure the incoming data, these are all simple cell calculations. Keep It Simple principle at work.

 

fwiw, I also like replaying the market day in my customized dashboard --- you can easily go back and see what a profile and its associated statistics were at a given time and day and imprint that on your brain for the future. For example, here is a snapshot of today.

 

Given this limited information in this screenshot you could say, while range is weak here at just 7pts, market is above the previous days high after testing into that range and there is no sellling. the lack of selling above previous days high is consistent with a day that has not made its final high yet. even assuming the likely 'P' formation occurs when buying is not sustained -- the lack of daily overlap combined with buying tail in B and lack of much selling in C or D --- all suggests market will auction higher from here. Now its up to you to find an entry off that top-down structure.