The Relative Strength Index, or RSI, is probably the most widely displayed indicator on retail charts. It is also one of the most misread. Traders hear "above 70 is overbought, below 30 is oversold" and treat those lines as automatic reversal signals. They are not, and understanding why is the difference between using RSI as a measuring tool and using it as a source of repeated small losses.
What RSI actually measures
RSI compares the size of recent up-closes to recent down-closes over a lookback window, most commonly 14 bars. The result is scaled to a 0-100 range. A reading of 50 means gains and losses over the window have been roughly balanced. A reading of 75 means recent gains have dominated losses by a wide margin. That is all. RSI is a description of recent momentum, expressed as a ratio. It does not know about central banks, order flow or support levels, and it makes no forecast.
Because it is derived entirely from past closes, RSI is a lagging, smoothed summary. It can be useful for comparing the current push against recent pushes on the same chart, but it adds no information that is not already in price.
Why "overbought" is a misleading word
The labels imply that price has gone too far and must come back. In a strong trend, momentum can stay elevated for a long time. A currency pair in a sustained policy-driven move can hold an RSI above 70 for days or weeks while price keeps climbing. A trader who sells every time RSI crosses 70 will spend a trending market fighting the move, taking many small losses and occasionally catching a pullback.
The same oscillator behaves differently in a range. When price rotates between two levels, RSI tends to swing between roughly 30 and 70 and the extremes line up with the edges of the range. So the identical reading means different things depending on the market regime. The indicator cannot tell you which regime you are in; you have to determine that from the chart first.
Three honest uses
1. Regime hint
Some traders observe that in persistent uptrends RSI tends to oscillate in a higher band (for example 40-80), and in downtrends a lower band (20-60). If RSI repeatedly holds above 40 on pullbacks while price makes higher highs, that is consistent with a healthy uptrend. This is an observation to be tested, not a rule to be trusted, but it reframes the question from "is it overbought?" to "what band is momentum living in?"
2. Divergence as a warning, not a trigger
Divergence occurs when price makes a new high but RSI makes a lower high, suggesting each push has less force behind it. It can be a reason to tighten risk or avoid adding to a position. Divergence can persist through several additional highs before anything happens, and often nothing happens at all. Treat it as context that raises your attention level, not as an entry.
3. Comparing swings
RSI gives a normalised way to ask whether the latest leg is stronger or weaker than the previous one. That can help you describe a market in your journal: "third push up, momentum weaker than the second." It is a note-taking aid.
Settings: fewer decisions, not more
A shorter period such as 7 makes RSI jumpy and produces more extremes; a longer period such as 21 smooths it and produces fewer. There is no setting that is right in general. Fiddling with periods until past signals look perfect is curve fitting, the same trap described in backtesting honestly for retail traders. Pick the default, keep it, and judge it over a large sample.
An illustrative example
Imagine a pair that has climbed for six sessions after a hawkish central bank statement, and RSI on the daily chart reads 78. A trader treating 70 as a sell signal shorts and places a stop above the recent high. Price pauses for a day, then continues higher on the next data release and the stop is hit. A second trader reads the same chart differently: momentum is strong, the fundamental driver is intact, so they do not fight it, and they wait for a pullback that holds an RSI of 45-50 before considering anything. The numbers here are illustrative only, but the contrast is realistic: the indicator is identical and the decisions differ because of context.
Common mistakes
- Using RSI alone. Without trend, structure and event risk, a single oscillator reading is not a plan. See a pre-trade checklist.
- Ignoring the calendar. A central bank decision can push RSI to an extreme in minutes and keep it there. See trading around high-impact news.
- Stacking oscillators. RSI, stochastics and CCI are all built from the same closes. Three of them agreeing is one piece of information shown three times.
- Skipping the stop. An "oversold" label is not a stop-loss. Define risk first; where to place a stop covers how.
How to test it yourself
Write down a precise rule, for example "when RSI(14) closes below 30 inside a defined range, record what price does over the next 10 bars." Collect at least 100 instances across different market conditions, including trends, and record the average result, the worst case, and how the outcome differs between ranging and trending periods. If the edge disappears after realistic spreads and slippage, you have learned something valuable cheaply.
Takeaway
RSI is a ruler for momentum, not a crystal ball. Used to describe the market and to prompt better questions, it earns its place on the chart. Used as an automatic reversal signal, it tends to teach expensive lessons. This article is educational only and not a recommendation to trade any instrument.