Trading Focal Points -- "A Leap Of Faith"... | Page 19 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Broker Challenge "NDD/STP" vs. True "ECN" and "Mkt Mkr" broker talk 46 replies

Leap of faith... 18 replies

The old "Control Points" vs. "Every Tick" testing question 2 replies

Leap of Faith 9 replies

Leap of Faith 6 replies

Attachments: Trading Focal Points -- "A Leap Of Faith"...
Exit Attachments

Trading Focal Points -- "A Leap Of Faith"...

  • #361
  • Apr 2, 2022 2:04am Apr 2, 2022 2:04am
Quoting robots4me
{quote} A new indicator -- R4M-Double-TopBottom-Inverse. I like it. I've marked major transitions above / below the zero baseline with vertical green lines. It aligns well with the R4M-Sinewaves-Return-To-Mean indicator (which would be expected), as well as with @Swingman's fractal indicator. In addition to providing direction I'm thinking it could also be used for Entries / Exits. {image} {file}
I've modified this indicator and changed the dots to lines...

Subtracting the SMA lines from Price action -- i.e. computing the differences or deltas -- has the effect of normalizing data. We are no longer distracted by the absolute price which will always fluctuate. Normalization is a type of filter that can have the effect of removing noise. When you see the lines move in the same direction that may provide another powerful clue.

I've also taken the liberty of adding the Double Top / Bottom crossings (16x32 and 16x32x64) -- the same markings as computed in the R4M-Double-TopBottom indicator.

Attached Image (click to enlarge)
Click to Enlarge Name: R4M-Double-TopBottom-Inverse-Line-Version.png Size: 96 KB

Attached File(s)
File Type: ex4 R4M-Double-TopBottom-Inverse.ex4   58 KB | 561 downloads
  • #362
  • Apr 2, 2022 4:57am Apr 2, 2022 4:57am
Hello R4m, Parisboy, and all others.

I will like to get some opinions from you...
See the gif i did. (Need yet a few little corrections)

https://gyazo.com/89a366da4301239d66b9152d483569d7

I used to track all crossing between all CMAs, and put an average in time and price from each swing giving this result.

I wanted to know, if i track all minors crossing between 2 focals (one confirmed, one potential), if i take from high and low (all crossing) until potential next high, i get the average midline, if i get from cross to cross (confirmed) to potential next crossing, i get the high and low average midline ? it makes sens to me

@Parisboy ?
  • #363
  • Apr 2, 2022 6:04am Apr 2, 2022 6:04am
Quoting js497
Hello R4m, Parisboy, and all others. I will like to get some opinions from you... See the gif i did. (Need yet a few little corrections) https://gyazo.com/89a366da4301239d66b9152d483569d7 I used to track all crossing between all CMAs, and put an average in time and price from each swing giving this result. I wanted to know, if i track all minors crossing between 2 focals (one confirmed, one potential), if i take from high and low (all crossing) until potential next high, i get the average midline, if i get from cross to cross (confirmed)...
JS497

nice replay of Price Action . agood way to analyze PA and lear.
What is the pair ? I could read 1 mn but not the pair name.
  • #364
  • Edited 6:27am Apr 2, 2022 6:16am | Edited 6:27am
BTCUSD 1 :n

Beautiful Chart to analyze
Attached Image(s) (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 188 KB
Click to Enlarge Name: screenshot.png Size: 135 KB
  • #365
  • Apr 2, 2022 6:18am Apr 2, 2022 6:18am
Quoting parisboy
{quote} JS497 nice replay of Price Action . agood way to analyze PA and lear. What is the pair ? I could read 1 mn but not the pair name.
Salut Alain,

its MNQ(micro nasdaq) 06-22 (current contract)

The replay of PA is cool yes, but not the heart of this tool . It's just to help to understand some situations, instead of waiting real market minute after minute

I have a little correction to make about the midline based on high/low (first point confirmed) to vertical line (potential second point) => i should to get average of all interactions for middle between high to low to high
I will add too from cross to cross (first point confirmed) to vertical line (potential second point) => i should to get average of all interaction for high or low between 2 crossing.

i hope that i clear ?
I will show result in 2h

a little example of result based on 3xCMA

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 89 KB
  • #366
  • Apr 2, 2022 6:52am Apr 2, 2022 6:52am
MNQ 1 mn building a Double Top
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 203 KB
  • #367
  • Apr 2, 2022 7:02am Apr 2, 2022 7:02am
Quoting parisboy
MNQ 1 mn building a Double Top {image}
I will compare when my correction of my tool will be done
  • #368
  • Apr 2, 2022 7:59am Apr 2, 2022 7:59am
Attachment g
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 100 KB
  • #369
  • Edited 8:31am Apr 2, 2022 8:02am | Edited 8:31am
I haven't yet received an answer to my earlier questions. That's okay -- I'm reading between the lines and charts...

I have a couple of points to make and I'm curious whether they resonate with anyone else...

(1) This regards the backshifting. @parisboy refers to CMA and I refer to MT4's SMA -- they are the same. And @parisboy is correct that the CMAs / SMAs of the individual lines displayed on the chart use all the data. However, the different SMAs are backshifted by different values. This means that SMA16 (which is backshifted by 8 bars) has to wait 8 bars before it can cross with an SMA32 (which is backhifted by 16 bars). And a 16x32 cross has to wait 16 bars before it can cross with SMA64 (which is backshifted by 32 bars) to form a 16x32x64 crossing. If I'm trading H1 the price action can change a lot while waiting 16 hours to see a 16x32x64 cross. If the market behavior doesn't change then all is good. But if it does then things could get ugly. So, what's the solution? Is there a solution? I think trading lower time frames addresses this problem. When trading M1 then waiting 16 bars isn't a big deal. The probability of the market changing within the next 16 minutes is much, much smaller than over the next 16 hours (in the case of H1). Is that why most of you guys are trading M1 and not the higher time frames? But, of course, there will always be a trade-off -- M1 bars are less accurate than H1 bars.

(2) This regards trading M1. Bars / candles represent accumulated ticks, where a tick is a change in price. The raw data -- in fact, the only data -- are ticks. Your trading software (e.g. MT4, Trading View, etc.) accumulates the incoming ticks to create OHLC bars for your trading pleasure. The more ticks that are accumulated per bar then the more accurate the resulting OHLC. In the case of an instrument that is not actively traded and for which there are few incoming ticks, then a single tick can skew the bar. As a result, M1 bars are much less accurate than H1 bars because their OHLC has many fewer ticks to work with. In the case of a less actively traded pair (e.g. AUDNZD) then this could be problematical. In the case of a highly traded pair (e.g. EURUSD) perhaps less problematical. So, what's the solution? Is there a solution? I'm wondering whether BTC is the way to go when trading M1 since I think it is highly traded and its M1 bars will represent lots of ticks, making its OHLC more accurate. But, of course, there will always be a trade-off -- BTC is highly volatile and may not lend itself to a trend-following type of strategy, which wave theory definitely is.

I'm not into trading M1 -- I'm not a screen-starer. And I won't trade BTC -- too volatile for the way I prefer to trade.
  • #370
  • Apr 2, 2022 8:03am Apr 2, 2022 8:03am
Quoting parisboy
{quote} Suppose we're computing the latest possible Moving Average data point for a 10-week Moving Average of the weekly closing price of a stock. The procedure used is to add up the last 10 weekly closing prices, then divide the sum by 10. Now on stock price charts, the resulting value of the 10-week Moving Average is always plotted at the time of the last price data point used in computing the average. This is incorrect !
Stated another way, the theoretical lag of a simple moving average is (n-1)/2.

Quote
The proper time relationship between the computed Moving Average and the stock price data it is smoothing is obtained by plotting the value computed above at a time half-way between the fifth and the sixth previous price data points. This falls in the middle of the week of a weekly chart, and hence is associated with no value for the weekly close of the stock.

So to center a 10 period SMA on the price the lag is: (10-1)/2 = 4.5 bars and the MA should be shifted back from the current bar this amount to center it on the chart. As you stated, this in between points due to the even number (10). To match discrete chart bars, the back shift could be either 4 or 5 bars with similar results.
Quote
For this reason, it is always better to use an odd number of data points in preparing a Moving Average, so that there is always a stock price to be associated with each average datum computed.
For an 11 period MA the lag is: (11-1)/2 = 5 bars back from the current bar will center the 11 period MA on the price.

If a different moving average is used other than the simple moving average, the theoretical lag will differ based on the weights used and may need to be computed analytically. The lag could be less than (n-1)/2 when the weights are more on the recent bars (i.e. WMA), and greater when the weights are more on the past bars.

All this you already know, but others might find my interjections useful.
  • #371
  • Apr 2, 2022 10:28am Apr 2, 2022 10:28am
FXEZ, R4M, 2 excellent posts. Thank you.

a) I am conscious that the serie 8,16,32,64,128,256 etc is not composed of odd numbers and should be according to the mathematics 9,17,33,65,129,257
It would be interesting to compare the results.

Anyway I have always choosen the coherence of the system based on Gann Time Price relationship (the Octave - division by 8)

b) About BTC R4M you are right about volatility.

But BTC is an excellent support to observe : very active, volatile and traded on week-end which is good for our technical reflexions

And as for 1 mn it is excellent for training because it minimize the time you need to observe the results of your hypothesis - ex: the time necessary to see the change of status of a potential Focal Point into a Confirmed Focal Point and their consequences on your Analysis and Trading
  • #372
  • Apr 2, 2022 10:42am Apr 2, 2022 10:42am
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 214 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 207 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 205 KB
  • #373
  • Apr 2, 2022 10:44am Apr 2, 2022 10:44am
Quoting robots4me
... the different SMAs are backshifted by different values. This means that SMA16 (which is backshifted by 8 bars) has to wait 8 bars before it can cross with an SMA32 (which is backhifted by 16 bars). And a 16x32 cross has to wait 16 bars before it can cross with SMA64 (which is backshifted by 32 bars) to form a 16x32x64 crossing. If I'm trading H1 the price action can change a lot while waiting 16 hours to see a 16x32x64 cross. If the market behavior doesn't change then all is good. But if it does then things could get ugly. So, what's the solution?...
I think you're asking the right questions - questions I've asked many times myself as I've worked on this the past couple of years. It's difficult to come up with real-time workable strategy based on these concepts as described by Hurst that fill in all the holes. I've spent a lot of time looking at and creating prediction algos to help in this but ultimately conclude that you need future data to avoid large prediction errors so a different approach is needed.

Tradingfives came up with an analytical way to apply the concept of focal points and created a pdf (see parisboy's thread and search the attachments) and a series of training videos. I don't have the software and don't know the exact algorithms used but it seems pretty clear from the videos what he is doing. See videos 1-3 for how he is using the "focal points" to make price and time projections.

He uses 3 main tools to do projections: (tools explained in video #3 and how to do projections in videos 1-2)
1) The E tool which gives extensions - calculated by the difference between two clicks on the y axis. He clicks on a swing high or low, then clicks on the MA crossover point after that high or low and it creates a red line the same distance below/above the MA cross. Hurst says allow for 10% error so there should really be 2 lines to give an idea of a range of values for a target.

2) The time tool gives time extensions. He clicks on the MA crossover before the high/low swing price, then the high/low swing price and then the crossover after the high/low swing price. This returns a yellow dot which gives an estimated time range. I think this just takes the absolute value of the two most recent MA cross bars - the swing high / low bar and adds them to the last MA cross to create a time range on the x axis.

Hurst basically did the same thing in this picture from chapter 4 of Profit Magic of Stock Transaction timing (figure IV-6) and turned it into a rectangle that takes into account both the high/low price estimates and the min/max time estimates and put it in context of the envelope channels.
Attached Image (click to enlarge)
Click to Enlarge Name: Hurst_Gruen_Figure_IV_6_pg_80.png Size: 43 KB


3) The R or regression tool. This isn't so much regression as it is slope extension. He clicks on two points near the end of a moving average and a straight line projects out based on the slope of the moving average at the time. The projected MA crossover (based on the straight line) is used to make projections on recent data of where the move may end and thus where a reversal position could be taken.

Anyway, there is more to it than this but I strongly suggest spending the 20 minutes or so to listen to the first 3 videos on the playlist (link below). I think it gives a simple analytical approach that uses many of the tools that are being discussed in this thread and that were introduced by Hurst.

https://www.youtube.com/playlist?lis...Sao_qnbj39ayLJ
  • #374
  • Apr 2, 2022 10:59am Apr 2, 2022 10:59am
Quoting FXEZ
{quote} I think you're asking the right questions - questions I've asked many times myself as I've worked on this the past couple of years. It's difficult to come up with real-time workable strategy based on these concepts as described by Hurst that fill in all the holes. I've spent a lot of time looking at and creating prediction algos to help in this but ultimately conclude that you need future data to avoid large prediction errors so a different approach is needed. Tradingfives came up with an analytical way to apply the concept of focal points...
Wow, wow -- there is a lot of meat here, thank you @FXEZ. Between you and @parisboy I have a lot of noodling to do...
  • #375
  • Apr 2, 2022 11:05am Apr 2, 2022 11:05am
Attached Image
Attached Image
Attached Image
  • #376
  • Edited 12:13pm Apr 2, 2022 11:06am | Edited 12:13pm
In short Hurst uses some "Mirror Image" (Symmetry) to build his estimated future Envelopes
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 263 KB
  • #377
  • Edited 11:45am Apr 2, 2022 11:22am | Edited 11:45am
Hurst labels " Nest of Lows " the grey zone where several Cycles of different periods converge within the same Time Zone. The reasoning is similar to the "Focal Points" where several Centered Moving Averages of different periods converge towards the same Price Zone

In this example Hurst uses another tool the VTL - Valid Trend Line -
A Descending Valid Trend Line links adjacent lower Peaks of successive Cycles of the same period. When a Descending VTL is broken on the upside , it means that the Cycle of the degree (period) higher than that of the VTL has made its Low. That creates a Candidate Action Signal to open a Long position.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 106 KB
Attached Image
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 26 KB
Attached Image
  • #378
  • Edited 1:20pm Apr 2, 2022 12:12pm | Edited 1:20pm
As you can see Hurst :

- defines a Time Zone where several Cycles converge to build a Nest of Lows
- defines a Price Zone - the Lower Mimit of the Envelope including "most of" Price Action

The conjonction or superposition of both Price and Time Zone generates a Potential Reversal Zone.

Then Hurst observe the building of 2 different kinds of Valid Trend Lines (VTL) by Price Action

Both VTL's are created by Cycles of different Cycles - one Cycle is shorter than the other. The First VTL falls within the Potential Reversal Zone, then the Second VTL too.

When Price Action breaks the First VTL on the upside , it means that the Cycle of the Longest Period at work within Price Action has generated its Low and that we have confirmation of a change in Trend.

Hurst always writes " use all the tools" . Depending of the conditions some tools are more efficient than others, but all used together it give us a powerful Edge.

As you can see it is a rational, logical and efficient system where it is difficult to separate each tool as everything is linked in a way or another.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 118 KB
  • #379
  • Edited 2:58pm Apr 2, 2022 2:30pm | Edited 2:58pm
Quoting robots4me
{quote} Regarding my comment about the wave theory providing direction, but not Entries / Exits. Sine waves have 3 properties -- frequency, length (i.e. period, span) and amplitude. Sine waves can be added and subtracted. Complex Sine waves can be broken down to their component waves using Fourier Transform. However, there are no straight lines in Sine waves. The problem that keeps popping up here is where to Enter / Exit, and the solutions so far have involved drawing a straight lines from "here" to "there". And the reason? Because it seems...
R4M "Wave Theory" is a vast subject. FXEZ, which is a serious guy with a lot of "theoretical" background, could tell us how he considers "wave theory" because he describes itself in his profile as a Wave Analyst.

I have very strong opinions (as always ) on "wave theory" (which theory ?), much more on wave theoreticians and much much more on propagandists and marketing men selling "waves" to the Dummies.

Obviously we will begin by R.N. Elliott. A nice guy who in his time was very valuable. THis means he focused investors on somme interesting concepts : Market Structure, Big Waves are made of Little Waves and little Waves summed create Big Waves. All that has been yet rationalized under the concepts of Fractals or of Autosimilarity.

Basic principle : If you do not take into account the scale, nothing is more similar to a Big Wave than a Little Wave. You can say the same thing for all Chart patterns (Triangles, Head and Shoulders and ..... Double Tops and Double Bottoms/

All that which is now obvious, falled in public domain, was when Elliott wrote "the Wave Principle" very new for the basic investor.
  • #380
  • Apr 2, 2022 3:21pm Apr 2, 2022 3:21pm
Half cycles from high/low presents yet some logic issues to work correctly.
However, i'm testing half cycles connecting focals from the baseline (cross-cross), almost done.
see both pictures
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 127 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 147 KB

based on 192/96/48/24/12 CMA
0 traders viewing now