Indicators: FATL

[Automated-Trading]

FATL:

Fast Adaptive Trend Line (FATL) is based on the low frequency digital filter.

FLF-1 filter is necessary for high frequency noises suppression and reduction of market cycles with a very short volatility periods that can also be considered as a noise. The filter parameters (fc cutoff frequency and A attenuation at the rejection band) were calculated with the use of EUR/USD exchange rate spectral estimation.

Filters of low frequency FLF-1 and FLF-2 provide attenuation in the stop band with no less than 40 dB and absolutely don’t distort the amplitude and phase of entry discontinuous price series in the pass band (bandwidth). These properties of the digital filters provide significantly improved (in comparison with simple moving average) noise suppression that in its turn allows reducing sharply the probability of appearance "false" signals for buy and sell.

There are no analogues to FATL among widely known technical instruments. It is not a moving "average", but just the adaptive lines estimates of the short-term trends. Unlike moving "average", FATL has no any phase delay with regard to current prices.

Author: Nikolay Kositsin

[Shaharudin Ahmad]

hi all,

i already implement FATL in my testing ea and produce a good result on EURUSD. My question; it is suitable for other pairs? Please refer to article by Vladimir Kravchuk, "New adaptive method of following the tendency and market cycles".

[Peter Maxwell]

Jelam, take a look at the co-efficients used in the filter, they are what you get if you pass integers as the range for the sinc function.  When the sinc function is transformed into the frequency domain - in this case I think you use the Z-transfer, i.e. discrete Laplace transform - you end up with an "ideal" low-pass filter with frequency cut-off f, say.  The frequency cut-off is dependent on the width of the sinc function, hence why Nikolay has produced several very similar filters: the frequency cut-off, f, will be different in each.

Now, one must be aware that this filter design is "idealised" in that it would require an infinite number of co-efficients.  That is not possible, so the sinc function has been truncated which will produce artifacts in the frequency domain.  One can apply a windowing function to reduce such artifacts, but it is always a trade-off.

This is not to say the filter is bad, more that one should understand what the filter does, how to use it and its behaviour.  However, for the purposes for forex, the simple truncation of a sinc function is likely good enough.

So in answer to your question, it will suitable for any currency pair, however depending on what result you require you may have to regenerate the co-efficients yourself.