Trading With Market Statistics.II The Volume Weighted Average Price (VWAP).

[Dogpile 10]

nelo, I understand all of that. I guess I just think of POC movement as secondary. I trade patterns while using the context of 'value' as a filter for my trades. I don't really care if you label POC 'abrupt' or not -- it is really not important what adjective you use to describe its movement. I understand how it moves -- I just think of it as a level of value and balance once it is established. Today you had a period of balance, a break away from balance and then an establishment of the second distribution. I couldn't care any less if everyone calls this movement 'abrupt' or not. Call it abrupt, don't call it abrupt, compare it to VWAP, don't compare it to VWAP -- whatever.

 

To me, what matters is that you understood that one price had been accepted -- and that price diverged away from that level which led to another balancing.

 

What is more interesting is -- look how NQ today went down and tagged Fridays closing VWAP. VWAP might be another powerful pivot for me -- but I will have to track it for a while before I use it like I do other key pivots (15-min EMA, last hour high & low, POC etc...). I use statistics for their tradeable tendencies. This helps give me a way to bias my trades. It is the pattern though that is important. Simple number-crunching of VWAP, POC, PVP or any other statistic is not going to leave you with an edge versus all the computer-powered algorithms of the world. You are not going to 'out-compute' the futures market. That said, I find the concept of Market Profile types of 'value determination' as very powerful ways to help give you confidence that the pattern you are seeing does have a statistical bias helping it to move in the direction you are trading.

[Dogpile 10]

nelo, please post how you used VWAP today to bias or enter trades. I would enjoy hearing how people are using it to trade.

[Dogpile 10]

here is my take for ES:

 

ES opened at Fridays POC/VWAP level. The market was thus opening 'in balance'. We had a push up away from this level and ES formed a balancing POC at 63.50 -- a level above the prevailing VWAP.

 

the market attempted and succeeded on a breakout away from the established point of balance near 1pm EST -- this had the effect of forming a morning high after 3 up days. This 'pattern' had me looking short. the market then offered 2 'shallow bear flags' on its push down. Price quickly fell and location now was FAR below VWAP and the earlier established POC at this point. Location was therefore difficult for any more shorting.

 

I did notice during this time that the PVP had now formed 'below' the VWAP (for ES). I was curious on this new concept that jperl outlined so I watched this with interest. jperl's advice would have been to bias trades long (if near the PVP price) since PVP was below the VWAP.

 

The problematic thing was that Dow, Rus & NQ all had PVP's above their VWAPs. So ES was kind of an outlier. So I would ask jperl and others what they do in this situation using their process?

 

I had my own opinion today -- but was curious what jperl thought of todays action.

[jperl 11]

 

Simple number-crunching of VWAP, POC, PVP or any other statistic is not going to leave you with an edge versus all the computer-powered algorithms of the world.

 

Dogpile, I think you are being somewhat unfair to those who want to determine if using statistical analysis will give them an edge in their trading. Those here who know how I trade, know that it is the only thing I use.

 

With regard to ES for today: I will only comment as if I were a NEWBIE trader since that's as far as I have taken the discussion in these threads. There are videos in [thread=2008]PART III[/thread] which show how NEWBIEs trade using market statistics, so I will look at today's ES only from that point of view. There were more trades that an advanced trader could have taken using only market statistics.

 

a)There were no long trades that NEWBIE could take

 

b)For short trades, NEWBIE takes a position only when the VWAP < PVP AND price action is below VWAP. So look at the first image. At 13:08 EST, VWAP (1563.12) < PVP (1564.25). Price action dropped below the VWAP, so NEWIE would have entered short. If he was aggressive enough he could have entered just below the VWAP. His stop loss would be just above the PVP at 1564.50. His profit target is undetermined. NEWBIE is still trading by the seat of his pants with regards to profit. Most likely he would have stayed in for 1 point (4 ticks).

 

c)Look at the second image. At 15:16 EST NEWBIE would have entered short again. Why? VWAP (1561.74) < PVP (still at 1564.25). He would have entered 1 tick below the VWAP and stayed in for 1 point. So his entry would have been 1561.50 and his exit would be 1560.50.

 

d)Note in the third image that at 15:54 the PVP changed abruptly to 1560.75. After that occurred, NEWBIE would only take long trades. But clearly this occurred to late in the day

 

[Dogpile 10]

Jerry, doesn't chart 2 have PVP < VWAP? I see a longer volume bar on the screenshot you posted creating a PVP < VWAP. If so, this would signal a long bias for Newbie... but only above the VWAP with a stop under 60.75? is that where Newbie is right now?

[jperl 11]

Jerry, doesn't chart 2 have PVP < VWAP? I see a longer volume bar on the screenshot you posted creating a PVP < VWAP. If so, this would signal a long bias for Newbie... but only above the VWAP with a stop under 60.75? is that where Newbie is right now?

 

The longest bar is where the red line is. It's hidden under the line, so it is difficult to see. (Make sure you maximize the chart size in order to see it ). You are correct in that volume is growing in at the lower price and in fact eventually the PVP will switch to a lower price as shown in chart 3. This doesn't occur until the end of the day.

[nickm001 10]

here is the pic of ES for July 16th that might help discussion... Developing POC(volume based) is white; VWAP is yellow/blue. You can see how the POC moved during the day

[nelo 10]

nickm001,

this is the kind of picture worth 1000 words, it is great to see developing poc and vwap together.

 

The concept certainly makes sense, however in practice I have many times gone into problems in situation as happened from 11:00 - shorting into rally to vwap. I guess this is the situation where third standard deviation comes into play but if Jerry could comment it I would highly appreciate it.

[nelo 10]

Dogpile,

 

till now I used vwap for some time just as strong level (together with HOD, LOD, opening range and yHOD, yLOD and yClose). For directional bias very simply - price over vwap = bullish, price below = bearish. Not in so sophisticated way as Jerry uses. For few weeks I also observed volume profile but it was much more detailed than 30-min MP and complex that it was simply too much to watch, so I dropped it.

 

I absolutely agree with you - person cannot compete with computer power used by institutions (which invested billions into algorithmic trading and creates every day more and more sophisticated programs). Ones advantage must be something computers cannot compete with next few decades, and that parallel processing of high-level patterns and feeling for market dynamics.

[jperl 11]

nickm001,

this is the kind of picture worth 1000 words, it is great to see developing poc and vwap together.

 

The concept certainly makes sense, however in practice I have many times gone into problems in situation as happened from 11:00 - shorting into rally to vwap. I guess this is the situation where third standard deviation comes into play but if Jerry could comment it I would highly appreciate it.

 

Nick-I know you are all salivating at the bit for more information, but you are going to have to wait until we discuss standard deviation of the vwap to get your answer. So stay tuned. In the mean time if you have not read [thread=2008]Part III[/thread], on the basics of VWAP trading, do so now.

[Dogpile 10]

I have spent a good deal of time thinking about the PVP last few days and I just don't think its something that can be considered significant. I will wait for jerrys forthcoming material but as of now, I think watching PVP is a total waste of time. That said, watching for 'price acceptance' via POC is very much a useful thing.

 

----

 

Here is an idea I am using VWAP for...

 

Note that I am combining a 'pattern' with the concept of 'statistical value' -- Personally, I think this is a powerful combination.

 

The concept of the ABC is that stops are gunned above the last swing high or swing low before the primary trend continues... This gives a nice clean 'pattern entry' for traders to find a low risk spot to enter a trade. But the trade becomes especially good if can combine the visual pattern recognition with a statistical tool -- enter the VWAP.

 

Here is todays NQ chart. Note that today the market auctioned down and then built a sideways congestion. This is generally difficult action to trade and can chop a lot of 'overtraders' up.

 

In my chart are the 2 ABC patterns of the day -- so just 2 entries for this set-up occured all day.

 

Also note that you can learn a lot by watching the action after an ABC pattern plays out. Does the market immediately go to another new low (ie, after the first ABC up) -- or does the ABC just lead to more congestion (the 2nd ABC up). Note that a market profiler would see that todays profile was not trending -- it was stalling. Lower prices were not attracting increased selling -- lower prices were beginning to shut-off activity. The failure of the 2nd 'ABC-up' to lead to a new low signalled that the market was stuck in a coil and a breakout one way or the other was next (a different set-up than an ABC is required for that).

 

http://bp0.blogger.com/_5h-SWVGx6Ms/Rp6CyUAHGPI/AAAAAAAAAUs/vo9P_qWvf7Q/s1600-h/July+18+ABC+Pattern.bmp

[jperl 11]

Dogpile- on 7/15 you said the following in [thread=2008]Part III[/thread] of the "Trading with Market Statistics Thread":

ok, thx for PVP explanation re. skewness -- that makes sense. I see what you are doing now with PVP. You are assuming that the PVP was an earlier point of 'balance' and that monitoring VWAP relative to PVP will keep you on the right side of a directional move as price diverges away from the earlier established 'balancing point.' This is an interesting specific strategy variation on classic market profile.

 

But today you said in the last post

 

 

I have spent a good deal of time thinking about the PVP last few days and I just don't think its something that can be considered significant. I will wait for jerrys forthcoming material but as of now, I think watching PVP is a total waste of time.

 

Sounds like you have done a 180 turnaround on this. Too bad, because the most exciting stuff is coming up soon.

 

You also made the following statement which is self contradictory

That said, watching for 'price acceptance' via POC is very much a useful thing.

 

If POC is a "very much useful thing", surely PVP must be also since market profile is a subset of the more general distribution function.

----

[Dogpile 10]

<<If POC is a "very much useful thing", surely PVP must be also since market profile is a subset of the more general distribution function.>>

 

Hey, I am open to anything so I am paying attention to what you are doing with PVP. But as of now, the very reason PVP isn't important to me is because it is like it is being treated like a POC --- but if you think this is the case, then I don't think you truly understand why POC is accepted price and PVP isn't.

 

PVP is just a number that has an outside chance of eventually become the POC. The vast majority of times, the PVP will not end up being the POC. Treating it as kind of a 'junior POC' and trading with that in mind is simply a concept that is not consistent with Market Profile logic. Do you see why?

 

To me, price is moving until it is 'accepted' -- through time -- and then it is a POC. Therefore, PVP is just another 'unaccepted' price. Trading with the belief that the PVP is 'like' a POC is a serious mistake in my opinion.

 

Again, I am studying what you are doing and hearing other peoples methods definitely helps me understand things more deeply because I am forced to try to understand what the logic is behind your and others approaches. I am all ears though re the PVP if you are going someplace with it that is NOT treating it like good market profilers treat the POC.

[nickm001 10]

For whatever it is worth, I have noticed that POC and PVP come withing few ticks in most cases. My thoughts are that they are very similar in shape and peaks are almost aligned. Exceptions are trend days and double distribution days when POC can pick one side and PVP the other.

If you were to plot globex and day session (24h) then definitely you wanna use volume distribution, or the MP would preference off hour session ( more time) and create unjustified peaks corresponding to low volume.

[Dogpile 10]

<<I have noticed that POC and PVP come withing few ticks in most cases. >>

 

I believe you mean in retrospect. Yes, at the end of the day they often might be quite close. But during the day, when you are watching price in real-time and while the market is building the distribution, the PVP will very often change multiple times. There is always a PVP -- there is always some price that is the highest volume price at that time. This price is just another price -- this has nothing to do with a market that is 'in balance'. Price is not thought of as being accepted because it hangs out there for an hour or two.

[BlowFish 10]

Dog,

 

Err I am not sure how to say this without appearing argumentative but I think you are just plain wrong. The modal volume point often as not is the same as the modal TPO point. Providing you sample the data in the same way of course (both in price and time). Actually the volume profile will generally be more accurate as commonly it is sampled every tick rather than every half hour.

 

Market Profile is simply a way to 'map' market structure. The stuff it is maping (supply and demand balances and imbalances etc.etc.) is best represented by volume as this actually is where traders met and agreed value. Order flow (volume) is the underlying stuff you are mapping.

 

When Steidlmayer 'invented' Market Profile he ingeniously came up with using time as a proxy and it works remarkably well. Even PS acknowledges that it is "volume that extends time" and "It (volume) is therefore the closet measurement possible to the start of non-randomness."

 

I guess I should I should read the second PS book. Most of the work he was doing between the books was more about detecting order flow and was an interesting departure from MP.

 

Cheers.

[BlowFish 10]

Jerry,

 

A couple of questions if I may. What if the gap between VWAP and PvP is 'large'. By large I mean placing the stop at the PVP would be too far away. I also wonder what newbie would do if the PvP jumps mid trade :o (I believe I am right in thinking you use today's developing PvP rather than yesterdays static one?)

 

Very much enjoying your posts, simple but elegant. Chomping at the bit for newbies next steps.

 

Cheers.

[Dogpile 10]

Blowfish, I think you miscontrued what I said. I never said volume was wrong or unimportant, I believe the opposite is true -- I just said that the PVP, at a given time of day, is not necessarily useful as a proxy for POC.

[jperl 11]

These are all good questions Blowfish:

 

What if the gap between VWAP and PvP is 'large'. By large I mean placing the stop at the PVP would be too far away.

 

In [thread=2008]Part III[/thread]you will notice in the videos that NEWBIES stops are larger than his profit targets. This is one of the dilemmas that NEWBIE is going to have to confront head on when we discuss the issue of what's wrong with the whole concept of a fixed stop loss. That's going to be a very interesting thread.

I also wonder what newbie would do if the PvP jumps mid trade :o (I believe I am right in thinking you use today's developing PvP rather than yesterdays static one?)

Yes NEWBIE for now only knows about todays data. We will leave to a later thread the significance of previous days,weeks,months PVP and VWAP. I will anticipate that discussion by telling you here that when NEWBIE is no longer wet behind the hears, he is going to want to know about previous PVP's, VWAP's, as they are going to tell him all about what price action to anticipate in the daily intraday time series.

As far as jumps in the PVP, that is another important event in the price action which we will discuss in a future thread. Again, I will anticipate that discussion by telling you that a jump in the PVP often will signal the onset of a reversal in the price action. For NEWBIE now that means exit the trade immediately and sit on the sidelines.

[nelo 10]

As I see the relation between PVP and POC, PVP is more precise and correct view on the market. It was difficult concept and complex to calculate so "time" proxy was chosen by P.S. for "value acceptance". 30-minutes MP and POC are just very simple (yet powerful) proxies for PVP and volume based profile in general.

 

The concept of PVP and HVL / LVL is independent on anything, it is just pure trade-based information, no time is required. I like time-based MP for simplicity, but for example little action and interest during lunch hours shifts the perceived value, it is not accepted in true sense, just nobody really wants to trade that time - 30-min based MP treats lunch time as great period of acceptance but nothing will fool volume profile.

 

Another advantage of volume profile is that it is much more detailed - the reality is that there is never (mostly) single accepted daily value as MP shows, but there are many levels with different strength. This might be considered as disadvantage since it makes it harder to read (to be honest I still struggle to use this advantage properly, right now mostly for timing profit targets).

 

Anyway, the concept of volume profile is logical, as elemental as it gets, chart independent and provides detailed fundamental view on market. These are the reasons it is useful for me.

[NAVEEVIa 10]

Hi Jerry,

i have benn following your threads for about a week now & have gained more insight than all tomes i have read earlier combined.

I was adviced to start with statistics & probability to learn about trading & its getting claerer now.

 

Some doubts

 

1. You said about VWAP that it being a average the volume traded above & below shoild be identical, i take the word identical as equal here.

So my question is , Is there a mathematical proof which says that for ANY distribution

P[X>=E(X)] = P[X<=E(X)]

i am thinking of VWAp as the expected value.

2. can this method be applied to stocks too.

Thnx

Naveen

[jperl 11]

1. You said about VWAP that it being a average the volume traded above & below should be identical, i take the word identical as equal here.

So my question is , Is there a mathematical proof which says that for ANY distribution

P[X>=E(X)] = P[X<=E(X)]

i am thinking of VWAp as the expected value.

 

There is no proof Naveen, because my statement about equality is only partially true. It's true for the median value by definition. You can see why it is true for a normal distribution and other well known distributions when the mean equals the median. For all other non analytic distributions what you can say and prove is that the

mean-median< 1 standard deviation

The proof is here

proof

 

can this method be applied to stocks too.

 

Short answer is yes. I qualify it with the statement, that there needs to be sufficient data to be statistically meaningful. So very low volume stocks would not work.

[Râlex 10]

Hello Jerry & day-traders,

 

For the programming of the VWAP (pro real time software):

 

Is it necessary to re-initialize (delay meters in zero)the VWAP ?

 

So yes, the first VWAP will be equal to the first price (first close for me, Pi=close(i)), since Vi (1) = Vtot.

 

Vtot= complete volume held concurrently there intraday, precise?

 

NB: Having grieved for my writing's errors, i hopes that you understand me.

 

Thanks.

 

Alexandre.

[jperl 11]

Hello Jerry & day-traders,

 

For the programming of the VWAP (pro real time software):

 

Is it necessary to re-initialize (delay meters in zero)the VWAP ?

 

So yes, the first VWAP will be equal to the first price (first close for me, Pi=close(i)), since Vi (1) = Vtot.

 

Vtot= complete volume held concurrently there intraday, precise?

 

NB: Having grieved for my writing's errors, i hopes that you understand me.

 

Thanks.

 

Alexandre.

 

If the first price and volume is p1,v1, second price and volume is p2,v2, third price and volume is p3,v3, then the VWAP after the third price is:

 

VWAP= (p1v1 + p2v2 + p3v3)/Vtotal

where Vtotal=v1+v2+v3

 

From this you can see how to extend the VWAP calculation to the next price and so forth.

[unicorn 10]

 

this is part of the POC's appeal, IMO. the market has an interesting habit of touching old POC's. .

 

 

Yes, I agree with you on this. Old POC's or PVP's do get touched. Again you will have to wait until we discuss this in a future thread. NEWBIE isn't ready for that yet. He only knows about today.

 

Hi Jerry;

I am going through your threads for the second time to absorb the nuances of your method.

I have not found the topic that refers to "Old POC's or PVP's do get touched. "

I suppose that the HUP in section XI, is the proper location, but I find that the information presented there is limited.

I would appreciate your advise with respect to this topic.

 

Thank you.

Unicorn.