Analysis and Trading based on Envelopes, Waves, Cycles | Page 339 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Planetary cycles, astrological theory of cycles of the markets 510 replies

Forecast Forex using Numeric Cycles and Astrology Cycles 38 replies

Planetary Cycles... Plotting Cycles into the Future 0 replies

Market Waves are NOT Cycles 8 replies

Waves/Trends/Cycles 0 replies

Attachments: Analysis and Trading based on Envelopes, Waves, Cycles
Exit Attachments

Analysis and Trading based on Envelopes, Waves, Cycles

  • #6,761
  • Feb 4, 2022 4:01pm Feb 4, 2022 4:01pm
Link always your charts and the theory whiich explains them.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 46 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 80 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 99 KB
  • #6,762
  • Feb 4, 2022 4:01pm Feb 4, 2022 4:01pm
Quoting parisboy
{quote} Fuseb My personal opinion conforted by experience and practice is that the system related to Hurst is the best one ....... because it is a complete system, rational and logical. Cycles for the Time Factor Centered Moving averages to determine the Trend(s) Centered Moving Averages intersections or crossovers to calculate targets and generate Focal Points Focal Points to determine the amplitude of the Sinewaves oscillating around the Trends Future Line of Demarcation (FLD) as Trend following tool and to calculate Targets and confirm Turning...
Thank you for the detailed explanation. I have had uncountable days of headaches in the past to make a complete system out of Jenkin's teachings, primarily based off the Secret Angle Method, and failed to get something consistent. But really enjoying taking up Hurst's system. It's refreshing how everything is quantifiable and tangible and has clear guidelines with no assumptions.
  • #6,763
  • Feb 4, 2022 5:09pm Feb 4, 2022 5:09pm
Quoting Swede
{quote} Your chart illustrates nicely the current trend starting to curve upward,,,,that is a key visual to either to take your profits from short positions and start looking for a retacement...anyway, it’s very late in the week so everything slowing down, not a good time to enter any more trades. I still have one running short from 47, it’s up about 5 pips atm and my target is the bottom of the Tillman fork, may have to close early if it pops up to break even... edit....hey Dave, got the final +10 pips sucked out of this session as price slowly...
Swede
At this rate you will soon need one of those
John Deere
ROCK TRUCKS
when you take your pips to the BANK
ps
Your doing Parisboy proud
  • #6,764
  • Edited 5:40pm Feb 4, 2022 5:30pm | Edited 5:40pm
Quoting OAPDave
{quote} Swede At this rate you will soon need one of those John Deere ROCK TRUCKS when you take your pips to the BANK ps Your doing Parisboy proud
I take my hat off to Alain for persisting with constantly posting even though there wasn’t much response at times. I have done a ton of watching, testing, analyzing the past couple of years to prove to myself how robust this method can be. Then I watched some huge moves that were clearly presented using the envelopes while I was scrambling with other methods not so robust.

The conclusion was to give him the honour he deserves by proving, with live trades, on the thread, for all to see, leaving little to argue. AND THAT WAS ONLY SCALPING WHICH IS THE HARDEST METHOD TO MAKE MONEY in this business.
Attached Image (click to enlarge)
Click to Enlarge Name: E83F8D58-CBE6-4549-9E6C-0C4DFB836306.jpeg Size: 573 KB
  • #6,765
  • Edited 3:46am Feb 5, 2022 3:31am | Edited 3:46am
Quoting Swede
{quote} I take my hat off to Alain for persisting with constantly posting even though there wasn’t much response at times. I have done a ton of watching, testing, analyzing the past couple of years to prove to myself how robust this method can be. Then I watched some huge moves that were clearly presented using the envelopes while I was scrambling with other methods not so robust. The conclusion was to give him the honour he deserves by proving, with live trades, on the thread, for all to see, leaving little to argue. AND THAT WAS ONLY SCALPING...
Something like this ?
Attached Image(s) (click to enlarge)
Click to Enlarge Name: GER30Dailyaa.png Size: 83 KB
Click to Enlarge Name: GER30Dailya.png Size: 78 KB
Click to Enlarge Name: GER30H4b.png Size: 107 KB
Click to Enlarge Name: GER30H4c.png Size: 111 KB
  • #6,766
  • Feb 5, 2022 8:03am Feb 5, 2022 8:03am
Quoting Swede
{quote} Parisboy made the comment one day that most traders do not know how to use moving averages, except him as he gave credit to Hurst. And He is right as I have experienced myself since going down this road and using Centered Moving Averages... example below.....1 Hour chart shows the 128 envelopes crossing north on the 256 , indicating we are in a strong uptrend atm...so the need to get on the right side of the market has been made clear....then adding in the focal points on the lower time frames gives tons of entries that can be done...
Swede I hope I was not so megalomanac !

Which is a sure fact is that 99,99 % of traders have no idea

- of the mathematical properties of Moving Averages
- and above all how to use these properies and their consequences for trading

see he common believings of MoeinYzz

Have to confess something and that is, I too have been one of those traders who never really understood moving averages and been mostly skeptical of it and calling it "the lagginggg indicator" without thoroughly understanding its mechanics.
  • #6,767
  • Edited 3:40pm Feb 5, 2022 8:32am | Edited 3:40pm
MOVING AVERAGES - - Channels and Cycles -Brian Millard - Chapter 7 - 1999

Let's go to the basics

The calculation of a moving average is quite straightforward, and the process for a 5-point average, is shown in Table 7.1.

A 5 point average is obtained by adding 5 successive values and then dividing by 5.

These could be daily prices, in which case the average is described as a 5-dayaverage, or weekly data, in which case the term 5-week average would apply.

It is taken for granted that the sampling interval, be it days, weeks, hours or years, is constant. It would be incorrect to use the data from week 1, week 2, week 4, week 7 and week 10, for example, adding these together and calling the result the 5—week average. Only by taking the data for weeks 1, 2, 3, 4 and 5 will we achieve a result that has any meaning when applied to investment data.

The 5 used in this case is known as the span of the average,

and in the stock market a variety of spans are used for averages.

We will see shortly that different spans have different effects on the data to which the average is being lied.
Attached Image
  • #6,768
  • Edited 3:40pm Feb 5, 2022 8:37am | Edited 3:40pm
A SIMPLE BUT KEY PROCESS TO UNDERSTAND HOW A MOVING AVERAGE IS BUILT FOR FUTURE USE IN TRADING - Channels and Cycles -Brian Millard - Chapter 7 - 1999

The adjective ‘moving’ is applied because the calculation moves through the data, taking the sum of the first 5 points and dividing it by 5, then the sum of the 5 points that start with the second data point and dividing this by 5, and then taking the sum of the 5 points that start with the third point, and so on until we have used up all the data.

Rather than starting each time from a point and adding up the 5 points, there is a short cut which simplifies the calculation, coming into its own when larger span averages are required.

We keep a running total, the first value is obtained by adding the first 5 points.

Dividing this by 5 gives us the first value for the average.

To compute the next value of the average, it is only necessary to add in the next value of the data and subtract the value the sixth point back (for 5-point averages).

This is the new running total, and when it is divided by 5 it gives us the second average point.

We proceed this way, updating the running total each time and dividing it by 5.

With say a 15 point average, we would first add up the first 15 points to give the running total, then divide it by 15 to get the average.

Then add in point number 16 and subtract the sixteenth point back, i. e. point number1 to give us a new running total that when divided by 15 gives us the second value of the average.

While it is obvious which point is to be added in to calculate the new running total, next one in the series, the point where the process could go wrong is if the wrong point is subtracted. In order to guard against this it is sensible to keep a note of the last point that has been subtracted by means of a tick or cross in a separate column, as shown in Table 7.1.
  • #6,769
  • Edited 3:40pm Feb 5, 2022 8:41am | Edited 3:40pm
- Channels and Cycles -Brian Millard - Chapter 7 - 1888

To the majority of readers who possess computers and software to carry out such calculations, this emphasis on the calculation of moving averages might seem misplaced. However, besides the fact that it is important not to treat computers as black boxes spitting out results, but to try to understand what is going on, there is another important point that is brought out by an examination of this method of computing an average.

The change in the running total, and hence the average, is brought about not only by the value of the next data point that is being added in, but by the data point, a whole span of the average in the past, that is being subtracted (the ‘drop point’).

This has an important implication when it comes to the turning points in averages and channels:

whether an average changes direction will depend upon the relationship of these 2 data points.

Besides the method of checking whether it is necessary to put a bend in a channel in order to allow peaks and troughs to settle comfortably at the boundaries, the investor can check the relative values of the next data point and the drop point as an additional aid in deciding if a channel has changed direction.

Moreover this essential process allows us to make hypothesis and generate even simulations. on a What if basis ?
  • #6,770
  • Edited 3:20pm Feb 5, 2022 10:12am | Edited 3:20pm
Swede I hope I was not so megalomanic !

Not at all....you gave credit to Hurst. And I agree with your statement about your understanding....im starting to also, as you witnessed this week.
  • #6,771
  • Edited 3:41pm Feb 5, 2022 11:39am | Edited 3:41pm
The Effect of Different Averages on Cyclic Data - Channels and Cycles -Brian Millard - Chapter 7 - 1999

We saw in Figure 7.1. that the effect of a moving average was to smooth the data to provide a ‘better’ estimate of the underlying trend.


Since we have defined trends previously as being part of a cycle, it is necessary to investigate the effect of applying averages of various spans to cyclic data.

The main properties of a moving average are:

1. A moving average will totally remove cy
Since we have defined trends previously as being part of a cycle, it is necessary to investigate the effect of applying averages of various spans to cyclic data.

The main properties of a moving average are:

1. A moving average will totally remove cycles of the same wavelength as the span of the average.

2. Cycles of longer wavelength than the span of the average will be reduced in amplitude. The smaller the span of the average relative to the cycle wavelength, the smaller the reduction in magnitude.

3. Cycles of shorter wavelength than the span of the average will be reduced in magnitude, and may sometimes be off set by half a wavelength from the original cycles, depending on the relationship between the span and the wave length. The greater the span of the average relative to the cycle wavelength, the greater the reduction in magnitude.

As an example, we will start off with a sine wave of wavelength 21 weeks and magnitude $10. The first 100 points of this wave are shown in Figure 7.2.


As expected, the peaks are 21 weeks apart and the troughs are also separated by the same amount. The wave swings from a value of zero up to $ 10.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 274 KB
Click to Enlarge Name: screenshot.png Size: 208 KB
Click to Enlarge Name: screenshot.png Size: 211 KB
Attached Image
  • #6,772
  • Feb 5, 2022 11:53am Feb 5, 2022 11:53am
Span equal to the cycle wavelength - Channels and Cycles -Brian Millard - Chapter 7

The first rule pointed out that cycles of wavelength equal to the span of the average would be completely eliminated.

This can be seen in Figure 7.3, where an average of span 21 weeks is applied to the cycle of wave-length 21 weeks. The original cycle is shown, while the result of applying the average is the horizontal line at the $5 level. Thus the cycle has been totally removed from the data.

Note the first point of the smoothed data appears at point 11. This is because we have a centered average and 10points are missing at the beginning of the average calculation, as explained earlier.

The reason for the constant value of $5 for the average is to be found in the running total calculation and the properties of sine waves. In a wave such as the present example, every point is repeated at exact distances of the wavelength apart in time. Since the Span is equal to the wavelength, the drop point in the calculation has exactly the same value as the next data point to be added. Consequently the running total and so the average, remain constant.

This effect (when the span is equal to the wavelength) is simply a special case of the more general one that removes a cycle completely when the span is an exact multiple of the wavelength: span = n x wavelength
Attached Image
  • #6,773
  • Edited 3:42pm Feb 5, 2022 11:56am | Edited 3:42pm
Spans less than the cycle wavelength - Channels and Cycles -Brian Millard - Chapter 7 - 1999

In the second rule above we pointed out that using a span less than the wavelength of the cycle should lead to a reduction in the magnitude. In order to check this point, averages with spans of 9, 13 and 17 were applied to the cyclic data shown in Figure 7.2. The result is shown in Figure 7.4.


The original sine wave is shown, and is the one with the greatest magnitude ($10) in the Figure. The other 3 waves of decreasing magnitude are the results of applying each of these averages. The wave with the smallest magnitude is the result of applying the 17 week average, and, as predicted, the magnitudes increase as the span decreases through 14 and 9 weeks. So, as long as the span is less than the cycle wavelength, the smaller the span used, the less is the attenuation of the data. The smoothed data sets are also aligned perfectly in time with the original data. The first plotted points of the 9, 13 and 17 weeks averages are at weeks 5, 7 and 9 respectively because they have been plotted as centered averages.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 203 KB
Click to Enlarge Name: screenshot.png Size: 231 KB
Attached Image
  • #6,774
  • Edited 3:41pm Feb 5, 2022 12:10pm | Edited 3:41pm
Spans greater than the cycle wavelength - Channels and Cycles -Brian Millard - Chapter 7 - 1999

In the third rule we pointed out that using a span greater than the wavelength of the cycle should lead to a reduction in the magnitude, but could also lead to an offset of the result so the original alignment was lost. In order to check this point, averages with spans of 35 and 45 were applied to the cyclic data shown in Figure 7.2. The result is shown in Figure 7.5.

The 35—week average starts at week 18 and the 45 week average at week 23 due to the fact they are plotted as centered averages. The interesting point is that while the 45 week average is aligned with the original data as


The effect of averages (with spans greater than the wavelength of the original data) suffers from the effect of the shifting of some averages. It also suffers from the fact that the magnitudes do not change in a straight-forward manner, as was the case with averages, that have spans less than
the wavelength of the data.

The shifting of some averages while others remain aligned with the original data is not a random effect, but does have a mathematical basis.

In the equation: span = n x wavelength
the averaged data remains shifted by one wavelength from the original data for spans which lie between an odd value of n and the next even value of n in the above equation. The first range of these spans would be142 from 22 to 41.

The averaged data remains aligned with the original data for spans that lie between an even value of n and the next odd value of n in the above equation. Thus the first range of these spans that are greater than the wavelength would be from 43 to 62.

In both of these cases, the magnitudes of the averaged data rise to a maximum and then falls back again. Therefore the least attenuation of the data occurs with a span of around 31 (shifted from original data) and 51 (aligned with original data.)

This shifting of the averaged data is something that the investor must giard against when it comes to the application of moving averages to real market data. We will see that averages will be used to highlight particular cycles, but that one has to be careful that a cycle is actually rising or falling, and not giving the impression simply because it has been shifted by half a wavelength because of the relationship between the span of the average and a cycle in the data. In such cases, the cycle will be moving in the opposite direction of that indicated on the plot. As long as the investor is aware of the possibility, it will be easy to check the exact status of the cycle by applying an average with a different (and very carefully chosen) span and viewing the result for consistency.

From these examples of the effect of different spans of average, the conclusion is we should use a moving average span less than the wavelength of the cycles we wish to observe. We will see later that we also have to take into account the presence of other cycles when deciding on the span most suited to our needs.

Sometimes we may wish to remove a particular cycle totally, in which case we can choose a span equal to the wavelength of this cycle.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 317 KB
Attached Images
  • #6,775
  • Feb 5, 2022 12:29pm Feb 5, 2022 12:29pm
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 165 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 276 KB
  • #6,776
  • Feb 5, 2022 12:33pm Feb 5, 2022 12:33pm
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 162 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 137 KB
Attached Image
  • #6,777
  • Feb 5, 2022 2:22pm Feb 5, 2022 2:22pm
Quoting Swede
Trading NFP is always a challenge......but here are a few tips using envelopes this morning... The trades shown are +10 +12 +11 +15 +15 so pretty good so far...now it has retraced back to the 50 area... Anyway.....NFP provided the direction For This Session....because overall Cable is moving north (see 1 hour chart)....but the NFP indicates there will be action on the short side for at least this session.... {image}
jolie en pleins NFP
  • #6,778
  • Edited 3:37pm Feb 5, 2022 3:24pm | Edited 3:37pm
AVERAGES AS TEMPLATES FOR CHANNELS - Channels and Cycles -Brian Millard - Chapter 7 - 1999

We mentioned at the start of this chapter that a trend could be represented by a centered average. We also came to the conclusion in chapter5 that channels had the same shape as the trends we drew through the center of the data, so it follows that we should be able to draw a channel by using the exact shape of a centered average as a template. The reason we have spent quite a bit of time in this chapter discussing the properties
of simple, smoothed and weighted averages is to make sure that when we analyze stock real data we apply logic to the selection of the averages from which channels can be produced. Since we also came to the conclusion that a smoothed average was superior in eliminating random fluctuations and cycles of shorter wavelength than the span of the average, we will use smoothed averages as the templates for drawing channels.

We will start with a consideration of what we called the inner channel impervious chapters. We will choose the lowest span of smoothed average that removes the fluctuations (whether random movement or cycles of shorter wavelength than this span). We will see that when we draw the channel and plot the original data, these fluctuations will be contained in the channel.

In the case of our test data from Figure 7.10, our inner channel is the channel that will contain the random movements. We drew such a channel graphically in chapter 5 (see Figure 5.10). The aim is to reproduce this channel by this numerical method as closely as possible. When we come to real data, this selection of the correct span of average for channel production will tend to be by trial and error.

In Figure 7.19, we show a plot of the 7-week smoothed average superimposed on the data. Note of course the average terminates before the last data point. It can also be viewed as a center line, since it runs through the center of the data. The channel is now constructed by duplicating the exact shape of this average twice, plotting one above the original and one below it. These duplicates must always be kept equidistant from the original average in order to preserve the rules of channel analysis. Since they are exact copies, this act of placing them above and below the original will preserve the essential feature of a constant depth. It is now only necessary to move the two duplicates closer to or further from the center line until the acceptable minimum of small violations of these boundaries is reached.
Attached Image
  • #6,779
  • Edited 3:42pm Feb 5, 2022 3:28pm | Edited 3:42pm
- Channels and Cycles -Brian Millard - Chapter 7 - 1999

This process is done automatically in a computer program such as Microvest 5.0TM by storing the distances between each data point and the position of the central average at that point. This gives a distribution of points that will get closer to the theoretical shown in chapter 6 (Figure 6.21) as the number of data points increases. From this distribution it then selects the position of the boundaries to enable a pre-determined number of points to lie outside of the boundaries. Although this number may need to be different in different sets of data, a good starting point is to allow about 3 to 4% of the points to lie outside of the boundaries.


The channel contains all of those fluctuations that are of wavelength less than the span used to calculate the average and so construct the channel.

As you can see from Figure 7.19, the channel depth is just over $8, which is in line with the actual magnitude of the random movement as running from +$2 to -$2.

Figure 7.19 demonstrates the problem in using the centered average as a template for the channel.
Attached Image
  • #6,780
  • Feb 5, 2022 3:30pm Feb 5, 2022 3:30pm
- Channels and Cycles -Brian Millard - Chapter 7 - 1999

We have a gap at the end due to the missing point, and the average and channel terminates 4 points back in time in this case. As we move into the future from the last data point, each new week will bring in a further calculated point for the average, and so the gap will be filled. At a point 4 weeks into the future we will know exactly how the average traversed the gap.

At the point we are in Figure 7.19, at week 190, we have to estimate the position of the average over these last 4 weeks. We are into the same problem addressed in chapter 6, where in the case of real data we will have to examine peaks and troughs in this gap closely in order to provide evidence of what has happened to the channel.

In the current example we take into account the large peak at week 162 this forces us to put an upward bend into the channel at the point shown. The estimated positions of the last 4 points of the average and hence of the channel boundaries are shown as dashed lines.
Attached Image
2 traders viewing now