jperl

I didn't have much discussion about trading at Old PVP's except for the post in the HUP thread. There is a complete discussion of PVP trading at the[thread=2232] "Trading with Market Statistics VII: Breakout Trades at the PVP" [/thread]thread. In that discussion and in the HUP post, I point out that entering trades at the PVP is not a good idea. It doesn't matter whether the PVP is old or new, touched or untouched. The basic point is that if the skew is large, ANY PVP represents a dividing line between the high volume area and the low volume area. If you take a trade at the PVP in the direction of the high volume area and it turns out to be wrong, you can be wrong big time with a large breakout into the low volume area against your entry. If the skew is small (VWAP~= PVP), the volume is the same on both sides of the PVP. Then you might as well flip a coin. Bottom line, don't enter trades at the PVP, new or old. (This is in sharp contrast to the Enthios style of trading).

It's not fixed in stone unicorn. The point of having high volume is two fold, a)You want enough data to generate proper statistics b)you want enough volume to provide liquidity, so you don't get caught with your pants down. 36000 contracts traded/day would be about 5000 contracts/hour or 80 contracts/minute. That's not a lot of contracts.

I don't trade these, but I see no reason why market statistics data could not be used for these, as long as volume data is available.

Perhaps this is the place for me to comment on this since I have written 11 threads on this topic entitled [thread=1962]"Trading with Market Statistics"[/thread]. You should be aware that market profile was developed more than 40 years ago at a time when plotting distribution functions had to be done by hand. You can understand that it was not possible to include every trade in the distribution function. Instead prices in 30 minute increments were used to generate the distribution. The net result is you end up with a distribution that essentially is a 30 minute average of the real distribution. Even Ensigns computation of the distribution is approximate, but it is much closer to the real distribution than anything else. Using a distribution like Ensign's has many advantages, none the least of which the Peak Volume Price dynamic (which I call the PVP to distinguish it from Market Profile's POC) is easy to follow as it builds up at one price and then evolves into another price over time. In MP this tends to get washed out due to the averaging. As far as value area is concerned, again for Market Profile, this was an historical choice of 70% of the data to conform to something like 1 standard deviation observed in a Normal Distribution. Unfortunately, the distribution is rarely Normal, so that defining a "value area" in this way is purely heuristic and has little if anything to do with the true standard deviation for the distribution. If anything, computing the value area with respect to the PVP yields values which are usually too large. A complete discussion of this can be found in the [thread=1962]"Trading with Market Statistics" [/thread] threads.

Ensign displays price or volume histogram, not MP.

I have had a chance to take a look at the Ensign internal plot for VWAP. To be clear, I do not use this code, but wrote my own using Ensigns ESPL language. Here is what I found: For VWAP, Ensigns computation and mine are dead nuts on. No difference. For the SD of the VWAP, there appears to be minor differences between my code results and Ensigns for most charts that I have looked at, EXCEPT FOR ES DATA of JAN 10. The comparison is shown in the two charts below: In the first 1 min chart for the time period from the 9:30 EST open to about 11:30, the SD's are very close. HOWEVER, from about 11:30 on, there is considerable disagreement as shown in the second chart. I do not know the source of this disagreement, but it appears to be related to the large increase in volume occurring after 11:30. I have submitted this info to Howard Arrington to look at, but I have not heard back from him. If any of you use ensign, you might wish to nudge him about this to get a response.

Tasuki, You will find a (almost) complete discussion of market statistics beginning with the thread [thread=1962], "Trading with Market Statistics" I. [/thread]. There are 11 threads in this series, describing in detail the proper way to compute the standard deviation, SD, and how to use the statistical data for entry and exit. Included in the series, are a quantitative description of the volume weighted average price, VWAP, the peak volume price, PVP, the skew, and how, along with the SD, a trade position can be executed with reasonable likelihood of success. Methods for scalping, position trading, and more complex trades at hold up prices, HUP, are also described. Market statistics is a very complex subject, and I've tried to make it simple enough in the threads so that both new traders and expert traders would benefit from a proper description.

Can't comment at this time because I am out of town and won't be back until Jan. 15th. If someone reminds me, I will take a look at the data when I get back.

Other than putting the HUP lines on my chart, I don't do anything else to begin the day. I just follow the statisitics and what the price action tells me. I will be cautious around key economic events that usually come out around 10:00 EST, but other than that there is nothing else that I do premarket.

And for those of you who use TradeStation, according to epiktetos at post 12637 the following is the info about upticks vs down ticks: "TS uses the following rules when assigning a tick to the reserved word UpTicks:An uptick is a tick whose price is higher than that of the previous tick of a different price. Thus, if tick 1 occurs at a given price, tick 2 occurs at a higher price, and tick 3 occurs at the same price as tick 2, both tick 2 and tick 3 are counted as upticks."

For those of you who use Investor R/T software, you have a choice on how you break down volume, the Mister Ed version (bid/ask) or my version (uptick/downtick). Here is a quote from their site: http://www.linnsoft.com/new/index86.html The Volume Breakdown Indicator (VB) has been enhanced with an option to divide up volume by "Up Tick vs Down Tick Volume". In the past, volume was always broken down between Ask Traded and Bid Traded volume. Now, the user has a choice of dividing up Buy and Sell volume as either "Ask Traded vs Bid Traded Volume" or "Up Tick vs Down Tick Volume. When "Up Tick vs Down Tick Volume" is specified, each trade that occurs on or after an uptick is placed in the buy volume category, while any trade that occurs on or after a downtick is considered in the sell volume category.

Yes I agree. I've started a new thread entitled "What is DEMAND/SUPPLY volume" Well your example is somewhat incomplete. Both Trade 3 and Trade 6 was at 50 (the same price). So in my terms, you would need to know what price the previous trade occurred. If it occurred say at 25, then both trades 3 and 6 would be DEMAND trades whereas if the last trade was at 75, then trades 3 and 6 would be SUPPLY trades. It matters not what the bid/ask prices happen to be, only what the time and sales price/volume shows.

This thread is started as a result of a short discussion by Tasuki in the Volume Spread Analysis posts 982 and 991 to quantify the concepts of DEMAND and SUPPLY volume. The problem arises on how to define DEMAND volume and SUPPLY volume. Some traders define trades made at the ask as DEMAND volume, trades made at the bid as SUPPLY volume. Other traders would say volume should be considered DEMAND only if there is an uptick in price and SUPPLY only if there is a down tick in price. If price does not change, then that trade is taken to be DEMAND volume if the last time price changed was an up tick and SUPPLY volume if the last time price changed was a down tick. The former definition is used by Mister Ed in the following quote from the VSA thread In that series of trades, trade 3 would be DEMAND volume of 75 lots because it occurred at the ask and trade 6 would be SUPPLY volume of 30 lots because it occurred at the bid. For the alternative definition of DEMAND/SUPPLY (which I prefer) consider the following series of trades 1.Bid =25,Ask=50, trade at 25 for 25 lots 2 Bid=25,Ask=50, trade at 50 for 75 lots 3.Bid changes to 50, Ask changes to 75, trade at 50 for 30 lots In this scenario, trade 2 would be DEMAND volume of 75 lots because there was an uptick from trade 1 from 25 to 50. Trade 3 would also be DEMAND volume of 30 lots (even though it occurred at the bid at the same price as trade 2) because the last price change was an up tick(trade 2 after trade 1). This is not a small difference in definition and is crucial for quantifying many of the ideas in VSA. This thread is now open for discussion.

I've been sitting on the sidelines watching this discussion about volume. The amount of "Volume" in this thread is a clear indication that there is considerable confusion about how to interpret volume in relation to price movement. I've read "Master the Markets". There is a lot of interesting stuff in there that needs to be better quantified. So far on this thread, I haven't seen much of that. Perhaps we can begin somewhere by using Blowfish's comments about volume. Let's start here: This is an interesting statement. It is not what Tasuki has said. Volume at the bid/ask is NOT the same as up-tick/down-tick volume. Let's be clear about this. As Tasuki has pointed out volume of any bar on a chart can be divided into two parts, DEMAND or BUYING volume which makes price rise, and SUPPLY or SELLING volume which makes price fall. This would be the same as ask/bid volume traded if it weren't for a small but very important fact. If you've ever watched the tape, you will notice that many trades occur at the same price before there is a new up-tick or down-tick trade. So what are these trades? Are they DEMAND trades or SUPPLY trades. We need a definition. A trade that takes place at the same price as the previous trade is a DEMAND trade if and only if the previous trade was a DEMAND trade. Similarly for a SUPPLY trade. In practice what this means is that you have to look back a sufficient number of trades to determine whether the present trade is to be considered a DEMAND trade or Supply trade. Looking at whether the present trade occurs at the bid or the ask doesn't hack it. There is then no mystery about HOW MUCH SUPPLY trades there are in a WRB that closes near its high. You don't have to wait for the next bar to figure that out. Just dividing the volume bar up into its DEMAND component and SUPPLY component will tell you immediately. Until there is some consensus here about how to determine DEMAND/SUPPLY volume, then anything else we have to say about VSA will remain confusing. 1

Ok thanks for the explanation

Well, not quite Blu-Ray. I see in the code that down tick volume is subtracted from the uptick volume. What I don't understand is how this becomes an oscillator. Perhaps you can explain it for us. JERRY

James or Walter, for those of us who do not use TS, would you explain how this oscillator is computed? JERRY

Actually, I recommended it before you in "Trading with Market Statistics:I. [thread=1962]Check here[/thread] In any case it's a very useful site, although most of the stuff on Market Profile is only of historical interest.

Trvlwanderer, You need to make a distinction here between averaging down and scaling in. I would agree that arbitrary averaging down is a losing proposition, but scaling-in is not. The distinction has to do with the concept of risk tolerance used in managing a trade. If for example you have a strict risk tolerance say 2% of your account balance, then scaling in within the boundaries of your risk tolerance becomes a viable technique to limit loses. See the thread [thread=2189]"Trading with Market Statisitics VI: Scaling in and Risk Tolerance"[/thread] for a more complete discussion.

Even though we know that TL has the best threads around, there occasionally are some good threads at other forums that are worth knowing about. One of them at ET entitled "Time of Active Trading Before Becoming Profitable" is very good and well worth reading. While this has been discussed before in this forum (see threads by [thread=465]carter[/thread], [thread=258]kingston[/thread],[thread=578]Suzy[/thread] and [thread=622],Stockaddict[/thread], it is well worth repeating. Here is the link: http://www.elitetrader.com/vb/showthread.php?s=&threadid=110889&perpage=6&pagenumber=1ell

Here is my take on market internals JWhite. A number of years ago, I used to plot a chart of the weighted sum of market internals (consisting of the following indices: biotech, software,hardware,financials,semiconductors, Dow,S&P,Nasdaq,Russell). I was under the impression that if I watched the rest of the market, I could get a jump on whatever else I was trading. Net result was zip. Sometimes the internals would lead, sometimes they would lag, sometimes they would move in the opposite direction. I don't follow market internals any longer, but just use the market statistics of what I'm trading to decide entry and exits.

Yes it has "moved" to ICENYBOT, but will be available on CME until end of September 2008 JERRY

Ofer, what's presented here in these threads is not a "method", but a way of viewing the price and volume distribution. No matter how many traders looked at the volume distribution, the distribution would still be valid, since it represents the statistics of all trades that have taken place. There would still be a PVP, a VWAP and a SD. How you use that information is highly personal depending on your trading style and objectives

Walter, VWAP computations are of course intrinsically dynamic and their value depends only on the starting time. In the [thread=2322]scalping thread [/thread], I pointed out that you could start a VWAP computation at any time during the day and use it to scalp the market. This usually works well, but as usual you still have to be careful with breakouts. The other point to be made is the dynamic nature of VWAP and its standard deviations allow you to use them as if they were pivot points from previous days, weeks, months, years. I discussed this in the [thread=2735] HUPthread [/thread] The advantage of using these instead of the standard pivot values is a)They are not huristic, but based on the pure statistics of the market price and volume data and b)They update dynamically rather than remain static as more data is added to the computation during the day. Also thanks for the Paul Levine references. I was not aware of this work and will have to take a closer look at it.

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.