RSI Divergence Indicators on TradingView: Compare Detection Methods
RSI divergence compares the direction of selected price swings with selected RSI swings. The important question for an automated detector is not just whether it draws an attractive line, but which points it compares and when that line becomes knowable. Pivot-based and regression-based detectors can give different answers on the same chart.
Define the pattern before choosing a tool
RSI = 100 − 100 / (1 + RS), where RS is smoothed average gains divided by smoothed average losses, not the ratio of up-day and down-day closing prices. TradingView's RSI documentation uses Wilder-style smoothing and explains the built-in divergence option.
| Pattern | Price comparison | RSI comparison | Hypothesis commonly investigated |
|---|---|---|---|
| Regular bullish | Lower low | Higher low | A possible upward reversal |
| Regular bearish | Higher high | Lower high | A possible downward reversal |
| Hidden bullish | Higher low | Lower low | Continuation of an existing uptrend |
| Hidden bearish | Lower high | Higher high | Continuation of an existing downtrend |
These are definitions of relationships, not statements that the proposed reversal or continuation will occur. A trend can continue through repeated divergences.
Three documented tools to compare
TradingView's built-in RSI
The built-in RSI offers Calculate Divergence. Enable it if you want its regular bullish/bearish divergence conditions to be calculated. TradingView's alert troubleshooting note confirms that these alerts depend on that setting. This is a useful baseline when comparing another script's added behavior.
RSI + Divergences + Alerts [MisterMoTA]
The publication currently displayed under NexusSignals describes regular and hidden divergence displays alongside RSI, a moving average, and configurable colors. Match the exact publication and current inputs; the title alone is insufficient to distinguish forks. Its feature description does not establish a lower false-signal rate than the built-in RSI.
RSI Divergence by faytterro
The author's publication describes divergence from differing signs of linear-regression slopes on price and RSI. That is a different definition from comparing two confirmed pivot lows or highs. Its update history also discusses alert behavior and labels identifying signal bars. Compare the current version's alert timing instead of assuming every line is a closed-bar signal.
These tools are examples of different detection approaches, not a tested ranking. A setting with more lines is not necessarily finding more useful events; it may simply apply a different definition.
Confirmation delay is part of the result
A pivot detector that requires right-side bars cannot confirm a candidate swing until those bars exist. If it draws the line back onto the swing candle, the historical chart can make the signal appear earlier than it was available. Record both the swing bar and confirmation bar. See TradingView's repainting guide.
For a hypothetical detector requiring five right-side bars, a candidate at bar 100 cannot be confirmed before bar 105. This is an illustration of timing, not a recommended lookback. Entering a simulated trade at bar 100 using that confirmation would introduce future information.
A settings experiment you can reproduce
Begin with the script's documented defaults as a baseline, not as an optimum. Change one question at a time:
| Input | What changes | What to record |
|---|---|---|
| RSI length and source | Momentum calculation | Exact source and length |
| Left/right pivot strength | Which extrema qualify and when | Both strengths and confirmation delay |
| Minimum/maximum separation | Which pairs can be compared | Accepted distance between points |
| Regression window | Which price and RSI slopes are fitted | Window and update behavior |
| Trend or volume filter | Which events remain eligible | Threshold and excluded observations |
Choose one symbol, exchange, interval, and period. Save every eligible event, including failures. Define the outcome window and any trade entry/exit rule before examining later prices. If adding a moving-average, candle, or volume condition, compare the complete rule with and without it on the same separate evaluation period.
A slower chart has different bar durations and usually fewer observations. It does not automatically produce a more trustworthy win rate. Report trade count, costs, average win/loss, and drawdown rather than accuracy alone.
Alerts and custom logic
Use TradingView's alert dialog and select a condition the installed script actually exposes. Inspect whether it runs intrabar or at close, then save a live alert log. Recreate an alert after changing the intended script inputs; see TradingView alerts.
Pineify can help draft custom Pine Script or configure supported conditions. Before generating a divergence strategy, specify the exact detector and confirmation timing. A visual editor does not imply that every community script's private or regression-based internals are available as a ready-made block.
Frequently asked questions
Which RSI divergence indicator is most reliable?
No comparative performance test is presented here. Compare definitions, confirmation timing, source access, and alert behavior first; measure trading outcomes under one explicit rule separately.
Does hidden divergence prove that a trend will continue?
No. It labels a relationship between price and RSI swings in a trend context. Continuation is a hypothesis to test, not part of the mathematical definition.
Why does a divergence appear on an earlier candle?
A detector may draw a confirmed pivot marker back onto its original swing bar. The confirmation bar is the earliest point at which a rule relying on that confirmation can act.
Why is my built-in RSI divergence alert silent?
Confirm that Calculate Divergence is enabled, that the intended alert condition is selected, and that the saved alert uses the current configuration. Then compare the alert log with the condition timing.