Qudos Market Insight is a leader in product innovation. Our ethos ensures the continuous improvement (Kaizen) of your technical analysis and trading software so that you are always one step ahead. Please click on one of the buttons below for more information on our product range.
OOPS! You forgot to upload swfobject.js ! You must upload this file for your form to work.
A World 1st, Channalyze is our new technical analysis module for predicting market movement through the recombination of recently-stable cycles and their correlation with channels.
Tests show a predictability of around 90% (yes, that right, 90% !) for direction and extent of movement 10 days into the future on many stocks. The predictable stocks are readily separated from those that are passing through a more random phase in their history.
Channalyzeis the professionals' and enthusiasts choice alike: it's so simple to use, innovative and unique in the market place.
Channalyze Enables You To:
1) Determine which securities are passing through a phase when their major cycles are stable
2) Determine when these stable cycles are like to change direction
3) Add these cycles together to predict major market turning points
4) Categorize cycles into low, medium and high wavelengths
5) Add the stable cycles within each of these three wavelength categories to predict short, medium and long-term movements
6) Confirm these movements by Channel Analysis
NB: Product packaging is artistic impression (Qudos reserve the right to change)
Vista Compatible
Followers of Brian Millard's methods will understand that trends are represented by centered moving averages. The difficulty is that centered averages terminate half a span back in time. This of course means that the direction of the trend is not known either in this gap or into the immediate future. Channalyze makes the assumption that trends can be decomposed into individual cycles. It is possible, with some error, to predict the future path of individual cycles into the near future. Thus by focussing on individual cycles, each one can be predicted into the future, and then summed to provide a good estimation of future movement.
Take advantage of the strong $ - our prices are in £GBP.
Click on expandable panels below for more detail.
Other Features
You'll Also be Able to:
Flag securities which show imminent cycle tops and bottoms
Scan these flagged securities daily, as the turning points approach from the future
Confirm by Channel Analysis that the turn has arrived
Display and store the distribution data of stable cycles for each security
Display and store a plot of the sum of stable cycles
Show the distribution and cycle sum data as the chart of each security is plotted
Specify which folders/securities are to be scanned
The program will scan either a single stock, a consecutive block in the current folder, all stocks in that folder, and all stocks in designated folders, storing the stable cycles for each scan. Because millions of calculations are involved for each stock, scans of over 1000 stocks are best carried out overnight, so that the results are ready the next morning.
You'll also be able to perform a plethora of other useful calculations including:
Millard CyclesTM, Super Rises and Falls, Channel Analysis, Moving Averages, Moving Average Difference, Triangular Centred Averages, Bollinger Bands, Relative Strength Index, etc., etc.
Intro
The focus of the program is two-fold. One is on the analysis of cycles in market data, isolating those which are currently going through a stable phase. There are then several summation options - the addition of all stable cycles, or the addition of short-, medium- or long term categories. The boundaries between these categories can be selected by the user.
The other focus is Channel Analysis as described by Hurst and later by Millard. This features automatic or manual bending of channels. The proprietary Sigma-p which gives improved prediction of future channel movement is also included. The relationship between channels and cycles becomes obvious after a few sessions with the package.
The program can either be purchased directly from our store, in which case a boxed product including the CD and 150 page manual is shipped, or it can be downloaded along with the manual (in .pdf format). In this case the trial period lasts for 10 days, by which time it is necessary to purchase and enter an unlock code.
The program operates on end-of-day historical data in Metastock format, but can also read historical data held as .csv text files. There are many sources of historical and daily updates in this format, both free and subscribed. The program comes with sample data from which the many screen shots in the manual have been produced.
Millard's Approach
Brian Millard recognizes that individual cycles in market data fluctuate, with changes in wavelength, amplitude and phase as the cycle unfolds. This is why Channalyze focuses on stable cycles where the most recent peaks and troughs show only a limited drift in wavelength. The amount of drift can be specified so that cycles exceeding this drift are automatically excluded from consideration.
An important fact is that the stability of such cycles tends to last for only around three or four complete sweeps of trough to peak to trough. Any cycles which have appeared to be stable for longer than this have reached a point where their future behaviour is unpredictable, and they are likely to disappear as quickly as they arose.
The next stage is to examine the distribution of cycles and analyze only those stocks which have a limited number of stable cycles which are also widely separated in wavelength. The sum of these stable cycles is then an excellent predictor of the underlying trend into the near future.
Finally, channels are calculated which can be compared with the predicted cycle sum, and if necessary can be adjusted so that they are moving in the same direction as the cycle sum. Readers of Millard’s books will be aware that once a stock price approaches a channel boundary, then the probability is high that the price will rebound. Thus once the channel direction has been established, which is not always straightforward, short term turning points in the near future can be predicted with a high degree of confidence.
Putting it Into Practice
One very useful feature which helps to build understanding of cycles is the ability to experiment by generating up to three sine waves, choosing their wavelengths, amplitudes and phase. These are then added and the result displayed. The various analytical tools than then be applied. It is now understood that chart patterns are the result of various combinations of cycles, and some of these combinations can be replicated, for example the head and shoulders formation shown in Figure 1. This has been produced by the combination of a 41 day cycle of amplitude 50 with a 101 day cycle of amplitude 60.
The distribution of cycles when the cycle sum from Figure 1 is scanned is shown in Figure 2. The mathematical filter used allows a small range of cycles to be passed, so that essentially the cycles appear as small groups clustered around the most prominent cycle. In this particular case, as expected, the cycles are centered about values of 41 and 101. By selecting ‘Sum Blue’ in the distribution form, those cycles clustered about wavelength 101 can be summed. The resulting clean cycle isolated by this means is shown in Figure 3. Experiments such as these help the user to build up a thorough knowledge of how cycles combine and how they can be separated again.
FIGURE 1. A TEMPORARY HEAD AND SHOULDERS PATTERN FORMED BY TWO CYCLES. IN THIS CASE THE WAVELENGTHS ARE OF 41- AND 101-DAYS.
Cycles can be examined in two ways. Either a single wavelength can be analyzed, or wavelengths between two values can be scanned. In scanning mode, those wavelengths which exhibit stable cycles are stored for later examination. The Doppler Shift allows the user to set a maximum percentage on the drift in wavelength as the particular cycle traverses from the past to the present time.
In the automatic mode, securities held in one or more folders can be set to be scanned for stable cycles - a scan of one security takes around 15 seconds. The results of scans are held against each security name. Two items are held for a scan – the distribution of stable cycles, and the sum of all of the stable cycles. Either of these can be displayed later as the user steps through the security list.
FIGURE 2. THE DISTRIBUTION OF CYCLES ISOLATED FROM THE CYCLE SUM SHOWN IN FIGURE 1.
FIGURE 3. THE 40-DAY CYCLE HAS NOW BEEN ISOLATED (LOWER PANEL)
Applying Channalyze to the Markets
A chart of the Australian Dollar versus the Euro is shown in Figure 4. The cycle sum is plotted in the lower panel, and it can be seen quite clearly that the currency is passing through a phase where stable cycles are prominent.
The manual points out that it is important to understand that the mathematical method by which cycles are isolated means that they terminate some way back in time. The higher the wavelength of the cycle, the further back in time is the terminal point. The individual cycles then have to be extrapolated up to the present time and into the future. Thus the current behaviour of any cycle is only an estimate. To avoid confusion, the extrapolated portion of an individual cycle or the cycle sum is plotted in a different color. It might be an obvious point, but if, at the present time, the extrapolation does not reflect the actual movement to date, then no reliable prediction can be made.
In the case of the Australian Dollar/Euro chart, this extrapolation (shown in red) starts in mid-September 2007, and this is the latest point for which the cycle sum is totally accurate.
It can be seen clearly that the extrapolation mimics quite closely the actual movement of the currency up to June 2008. The anticipated fall which was expected to occur from June onwards has yet to occur. This implies either that the major cycles in this are now increasing in wavelength, postponing a fall from the peak price for another few weeks, or that a cycle of longer wavelength is now becoming influential.
In order to get a better picture of what is happening, the distribution plot is shown in Figure 5. It can be seen that the cycles fall into two distinct groups, which is exactly the type of distribution to look for. The sum of the second group of cycles (26 to 89, in the green sector) is shown in Figure 6. This shows a very clean cycle pattern which has remained stable for a number of peaks and troughs.
FIGURE 4. CYCLE SUM OF AUSTRALIAN DOLLAR VERSUS EURO
FIGURE 5. THE DISTRIBUTION OF STABLE CYCLES IN THE AUSTRALIAN DOLLAR VERSUS EURO.
Quite clearly, the actual currency movement at the present time is diverging from the extrapolated movement, and therefore no prediction can be made. In contrast the next section shows a situation with the same currency in which a decision can be taken.
Predicting Future Movement
Channalyze allows the user to cut off the data at any point in the past, so that predictions can be made at that point. The prediction can then be compared with the actual. The subsequent movement of the security or currency is plotted in a different color, so that the actual movement can be compared with the predicted movement.
This is shown for the Australian Dollar in Figure 6, where the data has been cut off on the 15 April 2008. The question is how well did Channalyze predict the subsequent movement? The answer is, as can be seen in Figure 8, very well. The prediction was for a rise followed by a period of sideways movement, which is what happened. Not only that, but the position of the first subsequent down turn was also predicted with uncanny accuracy. The predicted features which can be seen in the cycle sum which are replicated in the actual currency movement are: a peak at 8th May , a trough on the 23rd May and a peak on the 4th June.
Although detail of the cycle sum is shown for some four months ahead of the cut-off point, it is of course unrealistic to use such predictions for a period more than about two months into the future, since as stated earlier cycles are subject to variation in wavelength, amplitude and phase. The behaviour of the cycles of wavelength 26 to 89 is a case in point, since they suddenly increased in amplitude in May 2007.
FIGURE 6. THE SUM OF CYCLES WITH WAVELENGTH 26 TO 89.
FIGURE 7. THE DATA HAS BEEN CUT OFF AT 15 APRIL 2008 (VERTICAL LINE)
FIGURE 8. THE PREDICTED CYCLE SUM IS EXPANDED FOR EASE OF COMPARISON WITH THE ACTUAL CURRENCY MOVEMENT.
Cycles with Channels
The theory behind channel analysis is that as the price movement approaches a constant depth channel boundary, the probability is high that the price will rebound from that boundary back towards the center of the channel. If the future movement of the channel is known, then this information helps to pinpoint short term price reversals. The key to this analysis therefore lies in the determination of the exact future channel direction. A difficulty arises when the extrapolation (the orange part of the channel) does not tie in with the price movement, as shown in the chart of Dow Chemical in Figure 9. Logic dictates that the channel should have turned up somewhere over the last few months, but this needs to be confirmed.
To help with this determination a plot of the cycle sum is shown in Figure 10. This shows that the cycle sum is still valid at the latest point, and therefore the prediction of a further rise in this stock is to be expected. Armed with this knowledge, the channel can now be adjusted as shown in Figure 11. The price is now approaching the lower boundary, and therefore can be expected to rise towards the centre of the rising channel.
FIGURE 9. THE 101-DAY CHANNEL OF THE DOW CHEMICAL STOCK PRICE
FIGURE 10. THE 101-DAY CHANNEL HAS NOW BEEN ADJUSTED IN THE LIGHT OF THE CYCLE SUM EVIDENCE
Typical Screen Shot: Autoscan Cycles (point mouse over to enlarge)
Screen Show
Recommended System Requirements
Operating System
Windows 95, 95, 2000, NT, XP and Vista (NB: not x64 bit)
PC
Pentium III or faster
RAM Memory
512MB or more
Hard Disc Space
1GB
Optical Drive
CD-ROM or DVD-ROM
Monitor
1072x768 resolution or higher
Internet
Broadband connection: 128 kilobits per second (Kbps) or greater (for activation of product/registration)
