EuroFX (future) 60 min

Timing Solution Back Testing report provided by Ben Price

upgrade February 2, 2006

Click here to download the Excel spreadsheet for Back Testing report.

Any of the Back Tested solutions can by applied to any financial instrument with character/structure/movement similar to Back Tested instrument. For example, the Solution back tested for Dow Jones Industrial index can be applied to any security that is believed to be similar to Dow Jones Index (in other words, its changes coincide with DJ moves). The Solution will be adjusted to this chosen financial instrument, and the projection line generated by this Solution will provide the forecast for it. 

Information regarding upgrade February 2, 2002

Additional Back Testing revealed the new projection line with training interval 600 price bars. This Solution is available in the latest upgrade of Timing Silution software. 

Comments by Sergey Tarassov

This is the first Back Testing provided for intraday data. This is done for weekly intraday data, assuming 24 hour trades with weekend gaps:

As usual, you can run the ready solution provided by Timing Solution software. The program calculates the projection line for any chosen financial instrument. You can get the list of ready solutions clicking "Timing Solutions" button. Choose this one:

Actually I was surprised by the quality of projection lines generated by this solution. Here you can see randomly chosen 7 projection lines, without any selection. This is a good illustration how the "cycle mode"  works in practice (John F. Ehlers, MESA and Trading Market Cycles),.

First of all, for this data it is better to use "Price Bar" metrics while calculating Spectrum. In other words, we are looking for cycles that live not in Universal Time, but in Price Bars. During weekends, there is no time "flow", because we have no trades there:

This metrics is much more preferable for intraday data.

Another issue that surprises me a lot is the length of the training interval. The best solution in the term of Timing Solution Style looks like this:

The length of the training interval is 72 price bars only (120 price bars for another best model). It means we need 3-5 days history only. This fact has some contradictions with Neural Network science. But in reality the program uses more price bars to calculate the spectrum. The Neural Net just arranges the revealed cycles to fit to the available price history. Moreover, the usage of the big interval drastically corrupts the projection line.

One more note. For projection line based on fixed cycle (spectrum model), we face a very typical situation:

Sometimes the turning points are shifted. Looks like the cycles are "melting", their period is changing within the time.
I believe that this problem is related to the time metric problem: we do not really know in what time stock market exists (and fixed cycles as well). Any university graduate will suggest to use Universal Time as the most commonly used time since the great Newton revolution in physics. The real Back Testing shows that the time measured by trade bars provides better results. And I am not sure that it is the best variant as well. The problem of the nature of time still persists.

As one experimental approach, I can suggest the volatility time. In other words, we assume that the time is not even. When the volatility is high, the time flows fast; when it is low, the time flows very slow; when there are no trades, the time stops. Maybe, this understanding of time is more human than the physical time used by scientists (at least, this idea is very close to the idea of "real time" measured by events and their duration as suggested by Henry Bergson).

How does it look in practice? I would like to introduce the first step to realization of this technology. In Spectrum module, calculate the spectrum diagram using  True Range as a measure of time:

The Spectrum diagram looks like this:

This is the diagram where the time is measured in cumulative true range. The highest peak is 157%. In other words, when the summary price change reaches 157%, the stock market starts the new cycle. 

This is experimental technology, I need some time to do more research and incorporate these cycles into Neural Net module.