DislikedHello all, Well, this really is great fun! Thanks a lot for this great idea. For many years we have seen folk trying to use Fourier and similar analysis on the markets but, as we have all observed, it isn't really very good. I am coming to the conclusion that this is because sinusoids are not suitable as basis functions for the underlying price data generation process. Price does not have rounded tops and bottoms like a sine wave. Price actually has quite spiky tops and bottoms, much more like a triangle wave function. I have written quite a powerful...Ignored
Thanks and thanks to remon78eg for a great idea, the results really look impressive. Could you please check how well does the method work when the next day's data step out significantly out of the scope of base functions for the previous day (for example, can one extrapolate a strong breakout within one timeframe)?
Cheers,
kprsa