Pattern Noise Removal

One of the major challenges that any technical trader faces is the removal of noise from price charts as these obscure and camouflage the underlying trend to force trades that fail.

The current methods to handle noise vary from basic to sophisticated, but still do not help significantly especially in ranging periods where a smaller period noise interferes with the trend patterns. An example of smaller period noise is swing patterns within a larger trend as shown below, which is a very common occurence in markets:

In the above examples the noise patterns are not exaggerated as they are in real markets. Some of the techniques that are used to remove noise are averaging price, using trend identifying methods like Heiken-Ashi and other similar approaches. Other alternatives are to use time-less charting methods like Point and Figure, Renko and Kagi or other proprietary methods. However, these methods do not handle lower order noise as shown above effectively, without significantly increasing trading risk.

Another method to handle noise is to remove the spurious noisy patterns from the underlying data, before drawing the charts. This can lead to amazing clarity in the trends as is demonstrated in a noise removal technique used in a Kagi chart, as an example below in the two sets of images. The upper chart represents the same data but shows the underlying trend without any of the lower order noise. Another research innovation at TradeWithMe at work.