Skip to main content
Track Price-indicator divergence
8 / 16
Search the archiveLibrary
Library

1998issue C031-11

Treat RSI as a testable filter rather than a trigger

Wilder's RSI is a 0-100 oscillator of a security's own up-close versus down-close momentum. This archive article pairs overbought and oversold warning zones with divergence and nearby support or resistance, so a reading is judged by whether the chart can still absorb a failed signal.

  • RSI is a 0-100 oscillator of a security's own up closes versus down closes, not a comparison of one security with another.
  • Readings below 30 and above 70 are oversold and overbought warning zones, not automatic proof that a trend has ended.
  • Shorter lookbacks produce more and noisier signals, while doubling a working window often reduces the signal count.
  • Weigh a reading against nearby support and resistance so the downside if it fails can be compared with the remaining upside.
Entries in this reading3 entries
Support and resistance5 readingsRelative Strength Index14 readingsCurrent trackPrice-indicator divergence19 readings

What the oscillator measures

Wilder's RSI is a 0-100 oscillator of a security's own up-close versus down-close momentum. It is not a comparison of one security with another.

In this archive usage, the relative strength index compares a security's recent up closes with its own down closes over a chosen lookback.

Lookbacks and warning zones

Common lookbacks include 9, 14, 16, 25 and 28 periods. Shorter windows produce more and noisier signals, while doubling a working window often reduces the signal count.

Readings below 30 and above 70 are treated as oversold and overbought warning zones, not as automatic proof that a trend has ended.

When price and the oscillator disagree

Divergence is a chart condition in which price and the oscillator fail to confirm each other's successive highs or lows.

A one-year, 285-stock check of 9-day RSI divergences counted 98 of 115 tops and 130 of 145 bottoms as correct, with W and M failure swings correct in 185 of 254 cases.

Room for a failed reading

Support and resistance are price zones where prior turning points can be used to size the downside or upside if an oscillator signal is wrong.

A signal should be weighed against nearby support and resistance so the downside if the reading fails can be compared with the remaining upside.

Gleason RSI crossing 30 and 70 too early

Gleason's RSI dropped to the 30 line in mid-October while the stock was still falling from the high 18s toward 14, then crossed 70 in early January a few sessions before the rebound stalled near 17.50. Those early warning-zone crosses are why a raw RSI signal is treated as a filter to check, not a trigger to obey. Oscillator values were read from the published daily pane, so they are approximate.
Gleason's RSI dropped to the 30 line in mid-October while the stock was still falling from the high 18s toward 14, then crossed 70 in early January a few sessions before the rebound stalled near 17.50. Those early warning-zone crosses are why a raw RSI signal is treated as a filter to check, not a trigger to obey. Oscillator values were read from the published daily pane, so they are approximate.Gleason Corp. · daily · 1996-09-30T00:00:00.000Z to 1997-01-21T00:00:00.000Z

The source pane uses the 30 and 70 warning lines. Its buy and sell marks sit on the raw crosses, which arrived before the October low and before the January high.

How the archive compared two rules

On a 16-day lookback, buying below 30 and selling above 70 was reported as 55 profitable trades out of 59, with an average gain of 17% and an average hold of 82 days.

Delaying the same 16-day 30/70 rule until RSI recrossed 30 or 70 reduced the average gain to 15% and produced worse average fill prices after the wait.

The evaluation used 242 stocks over roughly two years rather than a market index, covering more than 50 industrial sectors.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
8 of 16 in the Price-indicator divergence track
19991-7 pp.Next on Price-indicator divergencePrimary-cycle windows, then stochastic confirmationName a roughly 17-week primary cycle first and treat its four-week window as a calendar around expected troughs and crests, not as an automatic trade.
All readings on this track · 16 readings
  1. 1989Volume confirmation windows and exponential average construction
  2. 1990Constructing stochastic %K and %D from range position
  3. 1990Build a weekly leading sector composite from scaled transports and financials
  4. 1990Constructing stochastic K and D lines and divergence cues
  5. 1993Relative strength index events depend on the chosen input combination
  6. 1995Constructing a dual-horizon force index
  7. 1996Building a range-normalized divergence index from relative strength index
  8. 1998Treat RSI as a testable filter rather than a trigger
  9. 1999Primary-cycle windows, then stochastic confirmation
  10. 1999Stochastic rules versus buy and hold
  11. 2001Constructing confirmation filters for RSI overbought and oversold extremes
  12. 2003Constructing divergence-equivalent relative strength index and stochastic oscillators
  13. 2003Reverse-engineered RSI as a next-close projection
  14. 2003Scoring open versus resolved relative strength divergences
  15. 2003Bull-and-bear-balance from OHLC bar patterns
  16. 2003Constructing bull and bear balance from session paths
By fieldIndicators and filters56 readingsBy year19986 readings
All 19 readings tagged Price-indicator divergence
Also on Price-indicator divergence5 readings
2003Reverse-engineered RSI as a next-close projectionSupport and resistance2001Confirming a price-box break with on-balance volume and the accumulation-distribution lineSupport and resistance2015Mark the stop, the target, and the invalidation line before entrySupport and resistance2018Credit-spread risk budget beyond supportSupport and resistance1993Relative strength index events depend on the chosen input combinationRelative Strength Index