Statistics

[gassah 10]

Ray and James were recently discussing time and price windows, http://tradingsuccess.com/blog/cycle-high-for-sp-ii-873.html

 

Ray calculates the price and time changes of impulsive and corrective swings and uses these stats to:

1. estimate where a correction might end
   
2. determine how overbought or oversold a move is
   

 

I have a couple of questions on how to distinguish between impulsive and corrective swings. In the first attachment are Barros swings of the SP500. The blue lines are the 5d and and green ones are the next higher time frame, the 18d. If we are calculating the stats for the 5d then the impulsive waves are in the direction of the 18d line and the corrective waves are counter to it. There are a couple of exceptions. If the 5d line is totally covered by the 18d line (7-8) then that line is excluded because it is a higher time frame swing. The second chart has the second exception. If a swing in the direction of the higher time frame line is followed by acceptance back beyond the top or bottom 1/8th of the prior swing extreme, then the swing is counted as corrective, even though it's in the direction of the trend. Swing CD is corrective.

 

I'm not sure about about GH because it's a small swing that fits the above corrective criteria. Are the last swings always impulsive?

 

And in the 3rd attachment is AA-BB impulsive? There's a large retracement and it is part of a 5d range. Thanks.

[miriccar 10]

Hi Gassah,

i am quite new to barros method but it seems to me that AA-BB is impulsive of Subnormal Type (R0 move).

 

 

BR

 

Michele

[gassah 10]

it seems to me that AA-BB is impulsive of Subnormal Type (R0 move).

 

Michele

 

Thanks, Michele. I've had many questions answered by the second webinar and see it the way you do.

 

Rob

[jrob6519 10]

how can i learn this strategy?

[gassah 10]

Ray Barros Successful Trading Concept

[gassah 10]

NTES demonstrates the use of the time-price window (TPW). For those not familiar with the TPW it involves calculating the percentage change of prior swings and the number of bars for each of the swings. The goal is to estimate where a reaction will end and it utilizes the security's own history instead of some generic means such as the 50% retracement. It then narrows the window down to encompass approximately 70% of the values. In NTES's case the window ranges from 7-17% and 3-9 days.

 

It hits a MIDAS and then moves up with conviction allowing for an entry.

 

Rob

[gassah 10]

Attached are 5d TPWs for some of the stocks I'm following. It seems quite robust but it's a lot of work and very few will have the time or desire to do it.

 

Rob

[miriccar 10]

Hi gassah,

i would like to know your opinion in UUP chart.

Does it seem to you a Tree Drives to Bottom or a normal R0?

[gassah 10]

It doesn't look like a 3 Drives and it's premature to call it an R0 because there needs to be another reaction. If you zoom out there is a trading range that triggered a buy on 6/5.

 

Rob

[Tannism 10]

I can't seem to find "normalized volume" in the book. Can you tell me what it is or how it is calculated?

[gassah 10]

It's the volume as calculated here: Technical Analysis QQQ Trading Nasdaq 100 Stock chart S&P 500 market timing system

[Tannism 10]

Do you know what time period he uses to calculate ATR?

[gassah 10]

This post explains it: Measure of Contraction Blog for Trading Success

[gassah 10]

I can't seem to find "normalized volume" in the book. Can you tell me what it is or how it is calculated?

 

I purchased the normalised volume program from Kym Haines and he states it does the following:

 

"The program normalises volume based on:

- how a day's volume relates to the volume of the days before it

- what is the history of this volume relationship for similar days in the past (calendar day, day of the week, and many more)

- which type of similar day is the best predictor for the current day

- correcting the actual volume based on predicted behaviour (e.g. boosting volume on expected slow days)"

 

An example that Ray mentioned is July 2nd where volume was much higher after normalization.

[Tannism 10]

Ahh! Here a couple more that would have been nice to know.

 

Do you happen to know what formats the software uses?

 

Thank you. I am finding the book more and more helpful with your help with my questions and the link to the blog.

[gassah 10]

It's an interesting program. I'm not sure if manipulating volume in whatever manner he does is appropriate or more or less accurate. It costs $200.

 

In Vista I have to use a command prompt so it isn't windows friendly but it isn't difficult to use. You also need at least 10 years worth of volume for it to analyze.

[miriccar 10]

Hi gassah,

do you use mean+1stddev and mean -0.5stddev of average corrective swing for TPW you posted?

In NOT is not very clear

 

Thanks

Michele

[gassah 10]

When there are at least 30 data points I use the Steidlmayer distribution Ray teaches in his webinar. If you are serious about creating TPWs I can explain it.

[miriccar 10]

Yes please.

i would like to test it as i am doing on other Barros' tool.

As zone i like midas as well fibos anh harmonic but i am not convinced about TPW since i expect that a correction is function only of the strenght of the prior impulse swing in the same timeframe and the current movement in the higher timeframe so not on the very long term statistic.

 

Thanks for your help

 

Michele

[gassah 10]

Okay, I'll try to get to it this week. I'll do the Nasdaq Comp 18d that's attached. Don't mind the annotations as it was taken from another board.

Nasdaq.pdf

[gassah 10]

Hi Michele,

 

We'll be using the Frequency function in Excel.

Frequency function.pdf

[gassah 10]

The 1st spreadsheet displays all the 18d corrections since 1979.

 

The 2nd one adds the bins. The bins are evenly spaced numbers that range from near the low of the data and stop with a value above the highest data. In this example a "3" is entered in B7 and 2 is added down the column till 29 is reached, just above the 28.57. In B8 type "=", left click the "3" in B7, type "+ 2", enter, and 5 should appear in B8. Then click the B8 cell and go to the bottom right corner of the cell and drag it down till the 29 is entered in B20.

 

The bin column will take some experimentation. You could start off with "2" in B7 and stop at "30". Starting with 2 or 3 and adding 1 is another option. When you become familiar with a number of examples it'll become easier to decide.

 

Left click to the right of the first bins value in the empty C7 cell and drag down to the final value, C20. Type "=frequency". As frequency is typed you can hit tab to complete the word without typing the whole thing. Click the A7 cell and drag down to the last input A57. Type "," and drag from B7-B20. Hit and hold Ctrl/Shift together and press Enter. This will enter the number of data values that fall within each array. C7 equals 3 because there are 3 data values that are less than or equal to 3 (2.05, 2.75, 2.97). C8 equals 10 because there are 10 data values greater than 3 but less than or equal to 5, etc. See chart 3.

 

If C7-C20 is highlighted there will be the summed value (51) at the bottom right corner of Excel. If it isn't highlighted then you can do it at any time. This number will be used to calculate the percent each binned value represents of the total. In D7 type "=" and click on C7, "/51", enter. Click D7 and drag down from the right lower corner to the end of the values, D20. Highlight D7-D20 and right click in any shaded cell, choose Format Cells, Percentage, Decimal Places 0 and OK. Chart 4.

 

Now beginning with the highest percentage (22) move up and down choosing the highest value. Start at 22 and move to 20, all the time keeping track of the total. You want to stop when approximately 68% is reached as this represents the 1st standard deviation. After 20, 16 is larger than the 6 in D7 so include the 16. 10 is larger than the 6 in D7 so include the 10. The D8-D11 cells total 67% so stop there because more values will exceed 68. These correspond to the B8-B11 cells or an 18d correction of 5-11%. If there are more bins you can count two cells at a time and drop to one cell when 68 is approached. If there are equal initial high values start with the one nearest the center. If there are equal values as you move north and south from the center then include both.

 

Do the same for time which I've included in chart 6. Once you get the hang of it it's a breeze.

 

nic

$COMPX 1.xls

$COMPX 2.xls

$COMPX 3.xls

$COMPX 4.xls

$COMPX 5.xls

$COMPX 6.xls

[miriccar 10]

Thanks,

it is very clear. I am sorry for let you working so much for the explanation.

So you (and he) use the market profile VA method for determinating the limits of the statistic.

I see 2 problem for TPW application:

first as admitted by himself in a private mail divide the correction form impulse in programming is a problem

second: usually a correction is more a function of previous impulse (Action-reaction principle) and what type of elliott wave we are that on the long term padt stat

 

Thanks again

 

 

Michele

[Tannism 10]

Does auction bottom imply relatively low volume?

[gassah 10]

I believe so. He uses Market Delta when he comments on auction tops and bottoms.