Log-Normal Regression and Standard-Deviation Bands for Price Analysis

Summary

This indicator overlays two kinds of bands calculated from log-transformed closing prices. The statistical bands use a rolling average and standard deviation, similar in spirit to Bollinger Bands. The regression bands use a rolling ordinary least squares fit, its time-series forecast value, and the standard error of the estimate. A multiplier controls the width of both sets of bands, and the outputs are exponentiated back into price units. The author says the shared lookback window lets the regression bands respond more quickly to moves and volatility, while the statistical bands can help confirm medium- to longer-term movement.

The post explains that the second version performs its calculations on a logarithmic price scale, which it presents as helpful for handling moves across low-priced securities, and replaces the raw regression endpoint with a forecast that adds the slope. It supplies indicator code but no trading rules, backtest, or evidence of predictive performance. Band touches or crossings therefore remain analytical observations, not validated entry or exit signals; results will also depend on the selected period and multiplier.

Key ideas

• The indicator calculates statistical and regression-based bands from logarithms of closing prices.
• The statistical bands use a rolling average and standard deviation, while regression bands use a fitted trend and its standard error.
• A shared lookback period aligns the two band calculations, and a multiplier sets their widths.
• The forecast line adds the regression slope to the endpoint, and the results are converted back to price units.
• The post provides no backtest or rules showing that band signals predict profitable trades.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.