· 4 min read
RSI Overbought, Oversold, and Divergence: How to Check the Signal
Learn what RSI 70, RSI 30, and bullish or bearish divergence describe. Use a reproducible chart-review process instead of treating an extreme as an entry.
By TickerAI
RSI summarizes the balance of recent upward and downward price changes on a scale from zero to 100. An overbought or oversold reading describes momentum under the selected settings; it is not an instruction to sell or buy immediately.
A useful RSI review states the symbol, data source, timeframe, period, and completed candle. Without those inputs, two traders can disagree simply because they are measuring different series.
What RSI 70 and RSI 30 mean
Traders commonly inspect levels around 70 and 30, and a 14-period setting is widely used. MetaTrader's official RSI reference describes the oscillator and its calculation.
The ratio compares smoothed upward and downward changes. If the hypothetical average gain is 0.0006 and average loss is 0.0002, their ratio is three. Applying 100 − 100 ÷ (1 + 3) produces RSI 75.
That arithmetic does not include a target, an exit, trading costs, or a reversal probability. Strong positive changes can keep the reading high while price continues rising. Treating every high reading as a short setup adds an assumption that must be tested separately.
Define divergence before looking for it
A conventional bearish divergence compares a higher price high with a lower RSI high. Bullish divergence compares a lower price low with a higher RSI low. The comparison is meaningful only when the price and indicator points refer to the same relevant swings.
Consider this hypothetical bearish comparison:
| Observation | First swing | Later swing |
|---|---|---|
| Price high | 1.1050 | 1.1080 |
| RSI at corresponding swing | 74 | 68 |
The pattern describes weaker measured momentum at a higher price. It does not establish when a reversal begins or how far it will travel. You still need an entry rule, invalidation, and an evaluation method.
Avoid choosing swings after the outcome
Write a swing-selection rule before testing: for example, a high confirmed by a specified number of completed bars on each side. That is an illustrative research rule, not an optimized setting.
A confirmation requiring future bars becomes known only after those bars close. A historical chart can make the swing look obvious at its peak, but a live strategy could not use the later confirmation at that earlier time. Preserve the actual decision timestamp.
If you change which highs count whenever a signal fails, the divergence definition becomes impossible to evaluate consistently. Record rejected candidates as well as successful-looking examples.
Review price context and execution
Check whether the market is trending, ranging, or crossing a previously defined level. These labels also need observable definitions if you want to test them. Adding several indicators derived from the same closes does not necessarily add independent evidence.
Then ask what would cause an entry and where the idea fails. A trade with no defined invalidation cannot be sized from a meaningful stop. Use forex position sizing once an actual setup has been specified.
Inspect RSI with TickerAI
TickerAI can render an RSI pane alongside a price chart. Ask: “Chart EURUSD on the one-hour timeframe with RSI 14. State the broker source and the last completed candle. Describe the observations before suggesting any setup, and do not call an RSI extreme a guaranteed reversal.” Forex charts require a supported broker connection.
For crypto, ask for the exact Binance pair and avoid assuming that a broker CFD has identical candles. TickerAI's scanner uses particular RSI conditions as part of an explainable score; that score is not a tested probability of a winning trade.
Evaluate a rule rather than a screenshot
Keep a log of every qualifying signal under fixed settings, entry time, initial risk, costs, and realized result. Compare it against a simpler baseline on later data. A handful of attractive charts cannot show whether divergence improved outcomes.
The strategy-testing guide explains how to separate historical development from a later test. For broader context, use multiple timeframe analysis. This guide explains indicator interpretation and a research workflow; it does not claim that RSI creates a profitable strategy.