Continue reading...Limitations in usefulness of conventional standard deviation bands as overbought/oversold benchmarks

One of the popular ways of detecting a trend in financial market time series corrupted by random fluctuations (noise) is moving average smoothing. Actual prices fluctuate around the moving average. To measure an average distance of price values from the moving average the standard deviation (SD) can be used. The so-called in statistics ?range rule of thumb? determines the range of price variability around the moving average as approximately 4SD, providing confidence limit of about 95.5 %. According to the rule, the upper boundary of the range will be about 2SD above the moving average and the lower boundary of the range will be about 2SD below the moving average. In order to get higher confidence of the projected range, its half-bandwidth should be widened, for example to 3SD. The probability that the price data will be within plus or minus three standard deviations from...

Last edited by a moderator: