(standard deviation)LR-Channels Indicator finds all possible regression channels on the chart and drows channels with minimal standard deviation.
The channels with minimal standard deviation are more stable and represent the most likely boundaries of a trading range. Superposition of found channels and the support/resistance lines delineates reversal zones. Indicator also shows length (L) and width (W) of the each channel, Hurst criterion calculated for each channel by two ways (Criterion H1 and Criterion H2), the probability of breackdown (P) and parameters of the parabola describing the errors (A, B, C)
Hurst Criterion Explanation
H = Lg(R/S)/Lg(n/2)
R - Max range of sample
S - SD (standard deviation)
n - Number of observations
Hurst criterion is an empirical evaluation of how far we are from the distribution converges to normal or evaluation of fractality of the market. Any convergent distribution with increasing degrees of freedom converges to the normal, which means - the possibility of the applicability of the mathematical statistics apparatus on samples of approximately longer than 30 degrees of freedom (bar). It cuts off the sample and test - because the algorithm is obtained by iteration) - the error will tend to zero.
When this parameter is closer to 0.5 - indicates that the white noise dominates.
When approaching the 1 - indicates the presence of stable structures,
to 0 - unstable structures.
Further interpretation of the applicability of the stable - the trend,
unstable - kontrtrend.
Fractal, in this case (ie, self-similar structure) is a regression channel.
If Hurst Exponent is significantly different from 0.5 - reliable predictions is possible
If Hurst Exponent is around 0.5 - reliable predictions is impossible
We get parabola by evaluating the error with integral method. Parabola describes the most likely trajectory of trend reversal. The trajectory of the minimum energy.