On this page
Z-Score Volatility Bands: Standardised Price Channels
Z-score volatility bands convert price into a standardised distance from a moving mean, expressed in units of rolling standard deviation. Plotted around price, they form an adaptive channel that widens with volatility and narrows when the market quiets, similar to Bollinger Bands but read as a pure z-score series.
Key Takeaways
- Z-score volatility bands plot price as (price minus mean) divided by rolling standard deviation, scaled by a chosen band width.
- A z-score of +2 means price is two standard deviations above its N-period mean, the same anchor as standard Bollinger Bands.
- Wide bands signal high realised volatility; narrow bands signal compression and often precede expansion.
- Use them for mean reversion entries inside ranges and for breakout filters when the bands themselves expand.
Key Takeaways
- Z-score volatility bands plot price as (price minus mean) divided by rolling standard deviation, scaled by a chosen band width.
- A z-score of +2 means price is two standard deviations above its N-period mean, the same anchor as standard Bollinger Bands.
- Wide bands signal high realised volatility; narrow bands signal compression and often precede expansion.
- Use them for mean reversion entries inside ranges and for breakout filters when the bands themselves expand.
What It Is
Z-score volatility bands are a charting tool with two outputs. The first is a z-score line that oscillates around zero, showing how stretched price is in standard deviation units. The second is a pair of bands on the price chart, set at fixed z-score levels such as plus and minus two.
The construction is closely related to Bollinger Bands. The difference is that the band width is expressed as a pure z-score, and the indicator typically plots that z-score directly in a sub-panel rather than only the bands.
The Intuition
Raw price moves are hard to compare across assets and time. A 5 dollar swing on a 50 dollar stock is different from a 5 dollar swing on a 500 dollar stock, and a 5 dollar swing in a calm week is different from one in a panic week.
A z-score solves both problems at once. By subtracting the recent mean and dividing by the recent standard deviation, you get a unit-free number that says how unusual the current price is given the last N bars. Plus or minus two is the textbook "outside the normal range" line for an approximately normal distribution.
How It Works
The rolling z-score of close price with window N is:
z_t = (Close_t - SMA_N(Close_t)) / StdDev_N(Close_t)
Where SMA_N is the simple moving average over the last N closes and StdDev_N is the sample standard deviation of those same closes.
The upper and lower z-score volatility bands are drawn back on the price chart by inverting the formula at chosen z-levels k:
UpperBand = SMA_N + k * StdDev_N
LowerBand = SMA_N - k * StdDev_N
For k = 2 and N = 20, the bands are exactly Bollinger Bands with the default parameters. Z-score implementations let you read both the bands and the underlying z-score time series at once.
A common refinement is to detrend price first by taking log returns or by subtracting a longer SMA, then z-scoring the residual. That removes long-term drift and leaves a stationary input, which is closer to the statistical assumption behind the z-score.
Worked Example
Take the last 20 daily closes of a stock with SMA = 100 and StdDev = 4. Today's close is 107.
z = (107 - 100) / 4 = 1.75
Today's price is 1.75 standard deviations above its 20-day mean. The +2 band sits at 100 + 2 * 4 = 108 and the -2 band sits at 100 - 2 * 4 = 92.
Now suppose volatility doubles and the 20-day StdDev rises to 8 while the SMA stays at 100. The same 107 close becomes:
z = (107 - 100) / 8 = 0.875
Price is now less than one standard deviation above the mean. The bands have widened to 116 and 84. Same price, very different message about how stretched the move is.
A trader using bands for mean reversion would treat the first reading (z = 1.75) as a fading candidate and the second (z = 0.875) as inside the normal range, even though the closes are identical.
Common Mistakes
-
Treating z = +2 as an automatic short. A trending market can ride above its +2 band for weeks. Bands are statistical descriptions, not signals. Always combine with a trend filter or a confirming pattern.
-
Using a window that is too short. With N = 10 the standard deviation is noisy and the bands flap. Most desks use 20 to 50 bars, matched to the trading horizon.
-
Ignoring the regime. During regime shifts, the rolling mean drifts and the standard deviation expands. A z-score against the new regime is fine; a z-score that mixes pre- and post-shock bars is misleading.
-
Confusing standard deviation with volatility units. The standard deviation in the formula is computed on prices, not on returns. To compare to annualised volatility, scale appropriately or build the z-score on returns instead.
-
Reading bandwidth in isolation. A narrow bandwidth often precedes a move, but the direction of that move is not in the bands. Use a breakout trigger or a momentum filter for direction.
Frequently Asked Questions
What are z-score volatility bands in simple terms? Z-score volatility bands show how far price is from its recent average, measured in standard deviations. The bands widen when the market is volatile and tighten when it is calm.
How do z-score volatility bands affect investment decisions? A z-score near +2 or -2 flags an unusually stretched move and often precedes a pullback or pause. Many systematic strategies use the level as a filter for mean reversion entries.
What is a real-world example of z-score volatility bands? The 20-day, two standard deviation Bollinger Band on SPY is the most familiar implementation. It widens during corrections and narrows during low-volatility ranges before a breakout.
How can investors use z-score volatility bands effectively? Combine band extremes with a directional filter such as a 200-day SMA, and only fade z-scores in clear ranges. Inside trends, use band reversions as continuation entries rather than fade signals.
How are z-score volatility bands different from Bollinger Bands? Mechanically they are very similar. The difference is presentation: z-score indicators show the standardised series directly, while Bollinger Bands display only the price-space envelope.
Sources
- Definedge Securities. "Z Score Indicator." https://www.definedgesecurities.com/library/z-score/
- Chan, S. "Trading Trends With The Bollinger Bands Z-Test." Technical Analysis of Stocks & Commodities, March 2006. http://traders.com/Documentation/FEEDbk_docs/2006/03/Abstracts_new/Chan/chan.html
- Investopedia. "Z-Score vs. Standard Deviation: Key Differences." https://www.investopedia.com/ask/answers/021115/what-difference-between-standard-deviation-and-z-score.asp
- QuantInsti. "Standard Deviation in Trading: Calculations, Use Cases." https://blog.quantinsti.com/standard-deviation/
Disclaimer
This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.