Free Narrow CPR Stock Scanner: Daily Shortlist for Breakout Trades
Narrow CPR · Stock Scanner
There are no items in your cart
Add More
Add More
| Item Details | Price | ||
|---|---|---|---|
How RSI is actually calculated, and where the "overbought = sell" idea quietly costs traders money.
RSI (Relative Strength Index) measures how strong recent price gains are compared to recent losses, on a scale of 0 to 100. It's calculated as RSI = 100 - (100 / (1 + RS)), where RS is the average gain divided by the average loss over 14 periods. A reading above 70 is called overbought and below 30 is called oversold, but neither level tells you a reversal is happening on its own โ it only tells you momentum has stretched in one direction.
Relative Strength Index (RSI) is a momentum oscillator that compares the size of average gains to average losses over a set lookback period โ traditionally 14 โ and expresses the result as a single number between 0 and 100.
You enter a NIFTY PUT the moment RSI crosses above 70 because a Telegram group calls it "overbought, reversal aa raha hai." The move keeps going for another 40-50 points before it actually turns, and your stop-loss is long gone by then. That's not a flaw in RSI โ it's a flaw in how it's being read. RSI doesn't predict reversals. It measures momentum. What you do with that measurement is where the trade is actually won or lost.
RSI is a momentum indicator, not a reversal signal generator. It looks at a fixed number of recent candles โ 14 by default โ and answers one question: over this window, were the up-moves bigger and more frequent than the down-moves, or the other way around? The answer gets scaled to a number between 0 and 100 so it's easy to compare across any stock, index, or option, regardless of price.
When RSI is high, it means buying pressure has dominated recently. When it's low, selling pressure has dominated. That's genuinely useful information about the current state of momentum. What it is not is a countdown timer to a reversal โ strong trends can keep RSI pinned above 70 or below 30 for long stretches, which is exactly where traders who treat 70 as an automatic sell signal get run over.
The standard RSI formula, using Welles Wilder's original 14-period setting, is:
RSI = 100 โ ( 100 / ( 1 + RS ) )
Where RS (Relative Strength) is:
RS = Average Gain / Average Loss
Here's what that looks like with real numbers. Say over the last 14 periods, the average gain was 2 points and the average loss was 1 point:
A reading of 66.67 tells you gains have been roughly twice the size of losses over that window โ strong momentum, but still short of the 70 mark that most traders label "overbought."
This is the part most explanations skip, and it's the part that actually matters if you want to understand why RSI moves the way it does. The first RSI reading uses a plain simple average:
Every RSI value after that first one doesn't recalculate from scratch โ it uses Wilder's smoothing, which blends in the new candle while still carrying forward the memory of every candle before it:
This is why RSI feels "sticky" rather than jumpy โ one big green candle doesn't send RSI to 90 in a single move, because the smoothing formula weights it against 13 periods of prior data. Once you know this, a lot of RSI behaviour that looks confusing on a live chart starts making sense: RSI lags slightly by design, because it's built to filter out one-candle noise rather than chase it.
Traditionally, RSI above 70 is labelled overbought and below 30 is labelled oversold. Here's where it gets interesting: those levels describe where momentum currently sits, not where price is going next. In a genuine trending move โ the kind you see in NIFTY or BANK NIFTY during a strong directional session โ RSI can stay above 70 for candle after candle while price keeps climbing. Selling the first touch of 70 in that kind of session means fighting a trend that has no reason to stop yet.
The setup looks good on paper: "RSI is overbought, so a reversal is due." The problem starts after entry, when the trend simply continues and the overbought reading becomes irrelevant to what price actually does next. RSI extremes are a prompt to pay closer attention, not an instruction to enter against the trend.
Now look at what the diagram above is actually showing. Price makes a new low โ the second low is below the first. If you only watched price, that looks like the downtrend is intact. But RSI, over the same two swings, makes a higher low instead of a lower one. That mismatch is a bullish divergence: the selling is still pushing price down, but it's doing so with less force each time.
This is the same pattern you'll see on a falling NIFTY option premium during a sharp intraday dip โ price keeps sliding to a new low on the chart, but the RSI panel underneath is already curling upward off its own low before price confirms anything. Divergence doesn't mean "enter now." It means momentum has quietly shifted underneath the surface, and it's worth watching the next few candles closely for an actual price confirmation โ a higher low on price itself, or a break of the last minor swing high โ before treating it as a trade.
A mistake I see repeatedly with new options traders is trading the RSI number in isolation โ no price structure, no support or resistance context, no volume, just "RSI crossed 70, sell" or "RSI crossed 30, buy." RSI was never designed to be used alone. It's a filter that adds confidence to a setup you've already identified on price, not a standalone entry trigger.
A second, quieter mistake: assuming every overbought or oversold reading is a divergence setup. A genuine divergence requires two comparable swing points on both price and RSI โ one isolated dip in RSI means very little on its own.
Before you think about the target, decide what would prove the idea wrong. RSI fits into that process as a confirmation layer, not the starting point. In live market practice, this is where I actually use it:
This is also where combining RSI with a level-based framework like CPR helps โ RSI tells you about momentum, CPR tells you where price actually respects levels. I walk through that combination in more depth in the CPR Brahmastra webinar.
| Indicator | What It Measures | Best Used For |
|---|---|---|
| RSI | Speed and size of recent gains vs losses (0-100) | Spotting momentum strength and divergences over 14 periods |
| Stochastic | Where price sits relative to its recent high-low range | Faster, more sensitive signals in range-bound markets |
| MACD | Relationship between two moving averages (trend + momentum) | Confirming trend direction and momentum shifts over a longer window |
None of these replace price action โ each one is a different lens on the same candles. Traders often layer RSI with volume or a moving average rather than stacking three momentum oscillators that tend to say the same thing in different ways.
There's no single "good" number. Many traders watch for RSI moving back above 30 from oversold territory, or holding above 50 during a pullback in an uptrend, as supporting signals โ but the actual entry should come from price structure, with RSI used as confirmation.
No. RSI above 70 means recent momentum has been strongly positive โ it does not mean a reversal is imminent. In strong trends, RSI can stay above 70 for extended periods while price keeps rising.
RSI divergence happens when price and RSI move in opposite directions at swing points โ for example, price makes a lower low while RSI makes a higher low (bullish divergence), suggesting momentum is fading even though price hasn't turned yet.
14 is the standard and most widely charted period, and it's what most other traders are also watching, which matters for how the market reacts around it. Shorter periods like 9 react faster but generate more false signals; longer periods like 21 are smoother but slower to confirm.
It's not recommended. RSI works best combined with price structure, support/resistance, or a level-based framework like CPR, since RSI alone doesn't account for where price actually is relative to key levels.
RSI standardises momentum into a 0-100 scale using Wilder's smoothing, which makes it comparable across different stocks, indices, and timeframes โ something that's hard to judge consistently by eye alone.
RSI tells you how one-sided recent price action has been โ nothing more, nothing less. Treat 70 and 30 as attention markers, not entry or exit signals. Confirm any RSI reading, divergence included, against actual price structure before it becomes part of a trade, and know your stop-loss before you know your target. That order never changes, regardless of what the indicator is showing.
If you want a structured, level-based way to combine momentum reads like RSI with entries, stop-losses, and targets, the CPR Brahmastra framework covers exactly that.
Explore the CPR Brahmastra WebinarFor more chart-reading breakdowns like this one, browse the Trading Direction blog, or see what other students found useful in the testimonials.
This article is for educational purposes only and does not constitute investment advice. Trading in equities, futures, and options involves substantial risk of loss and is not suitable for every investor. Past patterns, including indicator behaviour, do not guarantee future results. Please consult a registered financial advisor and consider your own risk appetite before trading.