Elliott wave methods III

The Advanced Get

Perhaps the easiest approach to Elliot wave count is the e-signal proprietary software, Advanced Get. One of its main features is that it is able to search through a database of charts, including stocks, futures or a subset of these like the S&P500 or the Nasdaq components, to identify stocks at different stages of the Elliott wave cycle. One could select toppy stocks that have completed a 5th wave UP, or stocks at the bottom of a 4th wave, that are ready to have the last rally UP.

Advanced Get does not use the channeling or monowave approach, in fact the software seems to arrive at its count by proprietary computer subroutines. They appear to be based on the built in proprietary oscillator (see the chart below). If the Oscillator far exceeds previous highs, then it re-labels the wave as a wave 3. If the oscillator fails to exceed the previous high while the stock price does (divergence) then it labels the wave as wave 5. The inverse applies to down trends. Then, apparently based on Fibonacci retracements, it labels other waves ABC (corrective). Now, remember, in an impulse series, any of waves 1, 3 and 5 may extend, but wave 3 could never be the smallest. Advance Get seems to assume that wave 1 never extends and thus wave 5 is never the shortest. This misses out on truncated 5th waves, and thus brings into question the consistency of the type I trade.




I find it problematic with Advanced Get that it uses an arbitrary number of bars, set at 300 by default, to begin the count, essentially ignoring the preexisting sequences. To be exact about it, the Elliottician should start with historical data on each stock and find a major turning point, before commencing the evaluation. However, one could easily increase the number of bars to include a major turning point, before completely dismissing the counts.

In the example above, we are dealing with the same period in the Nasdaq composite in which the channeling method seemed to have problems. The same period using the Neely method, we concluded that the series was not an impulse series but rather a double zigzag (dz) which occurs frequently in the place of an impulse series, but that is corrective in nature, and that has serious implications about the stability of the rally in place. Here is the Neely count again:



A closer examination of the Advance Get count reveals that 2 monowaves were not labeled or included in the count, which brings into question the validity of subsequent waves. I also notice a problem, which is recurrent with the Advance Get counts, and it is the proportionality of waves of the same series. Take for example the first i-ii-iii-iv-v series: By Elliott rules, there have to be a proportionality between waves i, iii and v. They should keep a specific Fibonacci ratio. In the present case, wave i is very small compared to wave iii and wave ii is practically non-existent. The proportionality of waves is critical, in my mind.

Despite my objections, e-signal claims that their methodology has been tested extensively, and they are successful trading the type I and type II. No one can argue with success. Whether based on an orthodox Elliott wave count or a proprietary variant, the proof is in the results. The Type I and Type II trades merely use the Elliott count to alert you of stocks that after a big rally, have retraced sufficiently (according to Fibonacci ratios and oscillators), and are ready to mount a rally. If by their counts the rally coincides with a wave 5 that extends, then you have accomplished your goal to make money. Whether you end up calling the new wave an extended 5 or a continuation of wave 3 is only academic. The problem is that often enough, the 5th wave is shorter than wave 4, never to exceed the top of wave 3.

Advanced Get has a truck load of features that can be easily found in other web-based software. Take for example Stockcharts.com or iqc.com. OptionsXPress offers Prophet charting free of charge and that included real time charting. Stockcharts.com has many chart-searching features too, but they do not include Elliott wave searches. Platinum has also many search features that I find to be more efficient than Advanced Get, but that accomplish many of the same things. A combination of these programs could probably meet the expectations of Advanced Get alone. No one has the ability to search by Elliott wave features, or the ability to look at a historical chart as though you were at that moment in time: the training mode. I have come to depend on the training mode for my back testing of options strategies for decision making. For the Advanced Get searches to work, you have to download the entire database of stocks and futures every day, a process that lasts >30 min in my hands. Searches in Stockcharts.com are posted for you. Searches in Platinum are almost instantaneous. One could possibly display the stocks of interest without downloading any additional data, which is a major inconvenience. Why doesn't e-signal do the searches requested at the server level? I suppose they want to charge you for the huge amount of data they provide, and that you don't need.

Putting together the search and training modes and to do it at the server level would go a long way to alleviate the concerns about the validity of the type I and II trades and make it more user-friendly. Allowing the user to test their theories in the past to gain confidence in these systems would seem easy to accomplish and a great asset for the user. So far, they have refused to include this combined feature. Are they not confident that user will see that this type I and II trades are high probability set ups?