Skip to main content
Track Average Directional Index
54 / 56
Search the archiveLibrary
Library

2019issue C078-15

Exponential deviation bands with a moving average, RSI and ADX

A volatility envelope is assembled from a 20-period moving-average middle band and outer lines two exponential deviations away. Relative strength index and average directional index are confirmation overlays, so the combination is not a single-indicator reading.

  • Exponential deviation bands are a three-line envelope whose middle band is a 20-period simple or exponential moving average and whose outer lines sit two exponential deviations above and below that midline.
  • The middle band is the directional component. Exponential deviation sets band width from period deviations and a multiplier of 0.10 when the lookback is 20.
  • Band breakouts and changes in band direction are the chart events used to identify price trends and reversals under the standard settings.
  • A combination requires the envelope, a moving-average state, and at least one confirmation series to agree. The relative strength index is an overlay, and a flat moving average with a 14-day average directional index remaining below 20 marks a trading range.
Entries in this reading3 entries
Current trackAverage Directional Index71 readingsRelative Strength Index186 readingsMoving average449 readings

A three-line envelope

Exponential deviation bands are a three-line envelope whose midline is a moving average and whose outer lines sit a fixed multiple of an exponentially weighted mean deviation above and below that midline.

In the archive workflow the middle band is a 20-period simple or exponential moving average. The outer lines are placed two exponential deviations above and below that middle band.

How the middle band and band width are built

The moving-average middle band is presented as the directional component. Exponential deviation is presented as the component that sets band width.

Initial exponential deviation is the 20-period mean of absolute differences between each close and the chosen 20-period moving average. Each of those absolute differences is a period deviation. Later values update as period deviation times the multiplier plus prior exponential deviation times one minus the multiplier.

The smoothing multiplier equals 2 divided by the number of lookback periods plus one, which is 0.10 when the lookback is 20.

How far back the calculation begins

Exponential deviation and exponential moving-average values depend on how far back the calculation begins. A 20-period simple-average series is described as becoming more accurate after 20 days, while a 20-period exponential-average series needs an additional 20 days because the average itself is exponential.

The first completed spreadsheet row

A worked 20-day spreadsheet example reports, on 30 May 2018, a close of 12625.87, a 20-day simple average of 12643.49, an exponential deviation of 111.15, and matching upper and lower bands of 12865.80 and 12421.19 for both the simple-average and exponential-average columns on that first completed row.

NYSE Composite with 20-day exponential deviation bands

Daily NYSE Composite closes sit inside a 20-day simple-moving-average envelope whose outer lines are two exponential deviations away. Band width tightens through June, then the upper line lifts as the July advance stretches deviation. Every point is copied from the article’s worked spreadsheet, not traced off a plotted chart.
Daily NYSE Composite closes sit inside a 20-day simple-moving-average envelope whose outer lines are two exponential deviations away. Band width tightens through June, then the upper line lifts as the July advance stretches deviation. Every point is copied from the article’s worked spreadsheet, not traced off a plotted chart.NYSE Composite · Daily · 2018-05-30T00:00:00.000Z to 2018-07-31T00:00:00.000Z

The article computes both SMA-based and EMA-based bands; this chart uses the 20-day SMA columns. The first exponential deviation (30 May 2018) is the mean of the first 20 absolute deviations; later days use the exponential smoother with multiplier 2/(20+1). Exponential values depend on how far back the calculation is seeded.

Chart events under the standard settings

Band breakouts and changes in band direction are presented as the chart events used to identify price trends and reversals on a variety of securities with the standard settings.

Identifying a trading range

A trading range is identified by pairing a flat moving average with a 14-day average directional index remaining below 20, as illustrated on the FTSE 100 from October 2001 until a May 2002 breakout.

Relative strength index as confirmation

The relative strength index is specified as a confirmation overlay used with the envelope and other basic trend analysis rather than as a standalone replacement for the bands.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
54 of 56 in the Average Directional Index track
202026-29 pp.Next on Average Directional IndexA normalized-slope trend filter from linear regressionA raw least-squares slope inherits the price scale of the series, so a higher-priced path can post a larger number than a lower-priced path with the same relative shape.
All readings on this track · 56 readings
  1. 1986Cycle-aligned directional trend indicator
  2. 1987What crossover and directional entry rules actually compare
  3. 1988A directional-line cross needs a trend filter, an extreme-point rule, and a dollar stop
  4. 1988Constructing true range by offset addressing
  5. 1988Constructing directional movement from bar range
  6. 1988Average directional index construction: recursive smoothing and lookback offset
  7. 1988Staged Average Directional Index construction with Relative Strength Index confirmation and stop alerts
  8. 1988Average Directional Index construction with frozen true range and directional rules
  9. 1991Constructing the average directional index from range expansion and true range
  10. 1991Constructing five-session forecasts from stochastic, ADX, and MACD inputs
  11. 1993Constructing the average directional index from directional movement and true range
  12. 1993Confirming n-bar breakouts with ADX and DX filters
  13. 1994Constructing a Bollinger band-width trend filter
  14. 1994A pre-trade checklist that can refuse a long three ways
  15. 1997An ADX threshold and a moving average as a trend filter
  16. 1998Regime filters for mutated indicators
  17. 1999Building the average directional index from range extension and true range
  18. 2000Evaluating ADX, RSI, and moving averages in a multi-stock warehouse
  19. 2000Stochastic pop as a filtered continuation setup
  20. 2000Onset and exit from one average directional index
  21. 2002Joint ADX and MACD readout for trend strength and direction
  22. 2003Adaptive Donchian breakout with implied volatility and volume
  23. 2004The average directional index as a regime gate for the relative strength index and the stochastic oscillator
  24. 2004Constructing true-range-specified volume as a directional filter
  25. 2005Constructing a multi-filter penny stock breakout procedure
  26. 2005Construct one playbook that flips with session regime
  27. 2005Combining Bollinger Bands, the average directional index, and Fibonacci retracement on currency pairs
  28. 2006Assembling an adaptive price zone from double-smoothed averages
  29. 2006An ADX strength gate for MACD and the stochastic oscillator
  30. 2007Directional movement as a filter plus trigger
  31. 2007Constructing a veto-first trend permission stack
  32. 2007ADX gates for trend end, range, and reversal
  33. 2008Constructing a nine-cell directional-ratio grid
  34. 2008Average directional index and directional trend indicator lookbacks as trend-filter parameters
  35. 2008A holding-matched market lens from averages and directional-line crosses
  36. 2008A nine-cell directional scoreboard for multi-horizon entries
  37. 2010Building a Vortex Indicator from high-low distances
  38. 2010Constructing ADX, RSI, and MACD price filters
  39. 2011Constructing a volume zone oscillator with a moving-average and Average Directional Index regime filter
  40. 2011A volume zone oscillator conditioned by an Average Directional Index filter
  41. 2011Candlestick names need volume-price, ADX, and moving-average checks
  42. 2012Clustered average-directional-index traces as a trend-start filter
  43. 2012Average Directional Index cluster filters for trend-start signals
  44. 2012Confirming a trend start or turn with a triple ADX cluster
  45. 2013Constructing a late-entry stack from a signed DMI oscillator
  46. 2013A directional oscillator and its stochastic as a stacked timing filter
  47. 2013ADX cluster lookbacks are a locked specification, not a chart label
  48. 2013Combining moving averages, stochastics, and ADX in a daily scan
  49. 2015Assembling the Average Directional Index from directional movement
  50. 2016How an Average Directional Index filter and a breakout entry form one procedure
  51. 2016Score RSI and stochastic crossings only when ADX confirms the trend
  52. 2018Constructing an ADX filter for intraday breakouts
  53. 2018An ADX volatility gate for prior-day breakouts
  54. 2019Exponential deviation bands with a moving average, RSI and ADX
  55. 2020A normalized-slope trend filter from linear regression
  56. 2020Gating volatility-momentum divergences with a Trend filter
By fieldIndicators and filters682 readingsBy year201975 readings
All 71 readings tagged Average Directional Index
Also on Average Directional Index5 readings
2000Evaluating ADX, RSI, and moving averages in a multi-stock warehouse1987Constructing the commodity channel, average directional, and relative strength indexes on a shared cycle scale1988Staged Average Directional Index construction with Relative Strength Index confirmation and stop alerts1994A pre-trade checklist that can refuse a long three ways2004The average directional index as a regime gate for the relative strength index and the stochastic oscillator