RSI Explained: Read Momentum Without Guessing the Turn
The Relative Strength Index compares recent upward and downward closing-price changes on a scale from 0 to 100. It describes momentum; it does not tell you whether a token is cheap, safe or about to reverse.
Your task: record an RSI reading with its settings, explain what it measures and identify what would make your interpretation wrong. Practise on a chart without placing a trade.
Three things an RSI number needs
A timeframe and a calculation
Wilder’s RSI commonly uses 14 periods of closing-price changes, with smoothed average gains and losses. On a one-hour chart those are hourly periods, not days. It is not the percentage return over 14 candles. Different settings or data feeds can produce different readings.
Thresholds, not instructions
Above 70 is conventionally called overbought; below 30, oversold. These labels describe recent momentum, not all-time highs or lows. A strong trend can keep RSI near an extreme while price continues in the same direction. A reading of 50 does not prove that price is stable.
Price context
Bullish divergence means price makes a lower low while RSI makes a higher low. Bearish divergence is a higher price high with a lower RSI high. Either can fail. Compare corresponding swings on the same timeframe; a divergence is an observation, not confirmation of a reversal.
Build a reproducible chart note
Record the exact setup
Choose one spot pair and venue. Note the timeframe, candle time, RSI length, price source and whether the candle has closed. For example: a fictional ABC/USDT chart, one-hour candles, length 14, close source. Keep these settings unchanged while comparing observations.
Describe before interpreting
Write down the price and RSI together. If RSI reads 76, record that it is above the conventional 70 threshold. Then inspect the preceding trend and nearby price levels. Do not rewrite that observation as “price must fall” or assume 76 means a 76% chance of anything.
Define a check you can repeat
In a paper journal, state what later price action would contradict your idea and when you will review it. Record failures as well as successes. Two indicators calculated from the same price history are not automatically independent evidence. Include fees and slippage when evaluating any simulated trade.
Worked example: why the reading is 75
Suppose the already-smoothed average gain is 3 price units and the average loss is 1. Relative strength is 3 ÷ 1 = 3. RSI = 100 − 100 ÷ (1 + 3) = 75. This is above 70, but it is neither a 75% price rise nor a 75% reversal probability. If those averages were both 2, RSI would be 50. The example supplies the smoothed averages; a live calculation also needs sufficient earlier candles for initialization.
Mistakes that change the conclusion
Buying just because RSI is low
An RSI of 22 can accompany a continuing collapse. Check the asset, liquidity and news; a low reading does not make a project sound. Choosing not to trade is a valid outcome.
Comparing different clocks
An hourly RSI of 28 and a daily RSI of 58 describe different observations. They are not a calculation error. Label every screenshot and use closed candles when testing a rule.
Selecting only the neat examples
Drawing divergence after seeing the rebound creates hindsight bias. Define the swing-selection rule before reviewing later candles. Keep a record of false signals, and avoid choosing settings solely because they fit one historical chart.
Try it: interpret before acting
The smoothed average gain is 1 and average loss is 3. Calculate RSI. A friend says this guarantees a rebound because it is below 30. What is wrong with that conclusion, and what settings must you record?
Reveal the reasoning
RS = 1 ÷ 3; RSI = 100 − 100 ÷ (1 + 1/3) = 25. That is conventionally oversold, but price may keep falling. Record the pair, venue, timeframe, candle time, length, price source and candle status. The number alone supplies neither a reversal guarantee nor a reason to use leverage.
