Stochastic RSI Indicator: Reading RSI Against Its Own Range

Plain RSI has a habit that annoys short-term traders. It can drift between 40 and 60 for weeks, never touching 70 or 30, and so never giving you a clean signal. The stochastic RSI indicator was built for exactly that problem. It takes the RSI line itself and asks a simpler question: where is today’s RSI sitting compared with its own highs and lows of the last few sessions?

That small change makes the indicator far quicker. It also makes it far noisier. This guide walks through the formula, a worked example with numbers you can check by hand, the levels traders watch, and the ways it trips people up on NSE charts.

What the stochastic RSI indicator actually measures

StochRSI was developed by Tushar Chande and Stanley Kroll, who described it in their 1994 book The New Technical Trader, according to StockCharts’ ChartSchool. Their point was that RSI can stay between 80 and 20 for long stretches without reaching extremes, so a trader waiting for an overbought or oversold RSI reading can end up sitting on the sidelines for a long time. They wanted more sensitivity and more signals.

So they borrowed the stochastic oscillator’s formula, which normally compares a closing price with its recent high-low range, and fed it RSI values instead of prices. ChartSchool calls the result “an indicator of an indicator”. Price is turned into RSI, and RSI is then turned into a stochastic reading. That is why StochRSI is described as two steps removed from price, and why its line often looks nothing like the price chart above it.

The output runs from 0 to 1. Investopedia notes that some charting platforms show the same thing on a 0 to 100 scale, so a reading of 0.8 on one screen is 80 on another. Same number, different display.

The formula, without the jargon

Here is the calculation as both sources lay it out:

StochRSI = (Current RSI − Lowest RSI in the lookback) ÷ (Highest RSI in the lookback − Lowest RSI in the lookback)

The lookback is usually 14 periods. ChartSchool explains that the period you choose is used for both steps, so a 14-day StochRSI uses the current 14-day RSI and the high-low range of that 14-day RSI over the last 14 days. A few reference points help:

• StochRSI is 0 when today’s RSI is the lowest it has been in the lookback.
• It is 1 when today’s RSI is the highest in the lookback.
• It is 0.5 when RSI sits exactly in the middle of its recent range.

Investopedia’s step-by-step version is straightforward. Record 14 RSI readings. On the 14th, note the current, highest and lowest RSI and plug them in. On the 15th, do the same but only for the last 14 readings, dropping the oldest. Keep rolling the window forward one session at a time.

A worked example on a hypothetical stock

Let’s take a made-up NSE stock, XYZ, and its daily 14-day RSI over the last 14 sessions. These numbers are hypothetical, picked so the arithmetic is easy to follow:

On session 14 the highest RSI in the window is 66 (session 6) and the lowest is 38 (session 13). Today’s RSI is 45.

StochRSI = (45 − 38) ÷ (66 − 38) = 7 ÷ 28 = 0.25

Look at what just happened. RSI at 45 is completely ordinary. Nobody would call it oversold. But relative to its own last two weeks, it is close to the bottom of its range, and StochRSI says so with a reading of 0.25, just above the 0.20 line ChartSchool treats as oversold.

Now roll forward. On session 15, say RSI rises to 50. Session 1 (RSI 52) drops out of the window. The high is still 66 and the low still 38.

StochRSI = (50 − 38) ÷ 28 = 12 ÷ 28 = 0.43

Five points of RSI movement shifted StochRSI by 0.18. That is the sensitivity Chande and Kroll were after, and it cuts both ways.

The flat-RSI trap

Here is a second hypothetical case that catches beginners. Suppose RSI on another stock wanders between 42 and 48 for 14 sessions, and today it prints 48. RSI is dead neutral. StochRSI, though, reads (48 − 42) ÷ (48 − 42) = 1.0, its maximum.

Nothing is overbought here in any meaningful sense. The range is so narrow that any move to its edge throws StochRSI to an extreme. And if RSI is perfectly flat, the high and low are equal and the formula has nothing to divide by. Different charting tools handle that edge case differently, so a sudden blank or odd value on a very quiet stock isn’t necessarily a bug in your chart.

StochRSI doesn’t tell you whether momentum is strong. It tells you where momentum sits within its own recent range, and a narrow range can make an ordinary reading look extreme.

Overbought and oversold levels: the sources don’t fully agree

ChartSchool uses 0.80 and above as overbought and 0.20 and below as oversold. Investopedia’s key takeaways use the same 0.8 and 0.2 lines, but further down the same article it describes readings below 0.30 as oversold and above 0.70 as overbought.

We’d suggest picking one set and sticking with it while you test, rather than switching between them depending on which gives you the answer you want on a given day.

Trade with the bigger trend, not against it

The most useful idea in ChartSchool’s treatment is about direction. It says trend identification is the key to choosing between overbought and oversold signals. In a bigger uptrend, you look for oversold readings. In a bigger downtrend, you look for overbought ones. The readings that go against the main trend are mostly ignored.

ChartSchool’s examples define the medium-term trend with two simple moving averages: up when the 10-day SMA is above the 60-day SMA, down when it is below. It is also careful to say that oversold is not the same as bullish. An oversold reading is a warning to watch for a bounce, and you still need something on the price chart to confirm it, such as the stock closing back above its 10-day SMA or StochRSI climbing back above 0.50.

Put together, a disciplined long setup on a hypothetical stock might look like this:

1. The 10-day SMA is above the 60-day SMA, so the medium-term trend is up.
2. StochRSI(14) drops below 0.20 during a pullback.
3. You wait. No entry yet.
4. Price closes back above the 10-day SMA, or StochRSI crosses back above 0.50. That is the trigger.
5. The stop goes below the pullback low, so you know your risk before you enter.

For a downtrend, flip every step.

Using the 0.50 centerline for short-term trend

StochRSI is bounded between 0 and 1, so its midpoint is 0.50. ChartSchool says it reflects an uptrend when it stays consistently above 0.50 and a downtrend when it stays consistently below.

The raw line jumps around too much to use this cleanly, so smoothing helps. ChartSchool’s trend examples lengthen the setting to a 20-day StochRSI and add a 5-day SMA of the indicator, reading momentum as bullish when that average is above 0.50. Investopedia gives the example of a 10-day SMA of StochRSI producing a much smoother line. The general rule from ChartSchool is that momentum favours rising prices when the 10-day SMA of StochRSI is above 0.50 and falling prices when it is below.

Even smoothed, it isn’t clean. ChartSchool openly shows whipsaws, where a stock consolidates mid-trend and the 5-day average of StochRSI slips above and below 0.50 a couple of times before the trend carries on. If you trade every cross, you’ll get chopped up in sideways markets.

Speed comes at a price. More signals means more good ones and more bad ones, and the indicator can’t tell you which is which.

StochRSI versus plain RSI

The two are related but behave very differently on a chart. Investopedia’s comparison boils down to this:

Investopedia’s own line is fair: one isn’t better than the other. StochRSI just moves more, and more quickly. If you already use the Williams %R indicator, you’ll find StochRSI feels similar in temperament. Investopedia lists Williams %R, the money flow index and the stochastic oscillator as its closest relatives.

What to pair it with

Because StochRSI is a momentum oscillator built on another momentum oscillator, stacking it with more momentum tools mostly tells you the same thing twice. ChartSchool suggests volume-based indicators such as On Balance Volume or the Accumulation/Distribution Line, since they don’t overlap with momentum oscillators. Investopedia adds moving averages, MACD and Bollinger Bands as tools that can help filter false positives.

The examples in ChartSchool lean heavily on plain price action for confirmation: gaps, support and resistance breaks, and chart patterns. Honestly, that is where most of the value is. StochRSI flags the moment to pay attention. The chart tells you whether to act. If you’re building a broader toolkit, our technical analysis guide for Indian stocks and the explainer on the Commodity Channel Index cover other oscillators worth comparing it with.

A screening idea you can adapt

ChartSchool publishes two sample scans that show the trend-first logic clearly. The oversold scan looks for stocks in a medium-term uptrend (10-day SMA above 60-day SMA), trading below their 10-day SMA, with StochRSI(14) below 0.10. The overbought scan is the mirror: 10-day SMA below 60-day SMA, price above the 10-day SMA, StochRSI(14) above 0.90. Its scans also filter for minimum price and average volume, and ChartSchool warns they often return many stocks and may need further refinement.

On Indian stocks, the same idea works as a starting list, not a buy list. You’d still want to check liquidity, any results date coming up, and whether the stock has news behind the move. A pullback ahead of a bad quarter can stay “oversold” for a long time.

Where stochastic RSI goes wrong

• Too many signals. In ChartSchool’s own words, StochRSI dramatically increases the signal count compared with RSI. Without filters, you’ll overtrade.
• It lags price in its own way. Investopedia points out that being two steps removed from price means it can fall out of sync with what the stock is actually doing.
• Narrow ranges distort it. As the flat-RSI example showed, a sleepy stock can show extreme readings on tiny moves.
• Settings change everything. A 14-period StochRSI on a 15-minute chart and a 20-period one on daily candles are different tools. ChartSchool advises experimenting with settings and learning the indicator’s quirks before using it with real money.

The short version

The stochastic RSI indicator is RSI made more sensitive. It measures where RSI sits inside its own recent high-low range, on a 0 to 1 scale, with 0.80 and 0.20 as the common extremes and 0.50 as the trend line. It shines at flagging pullbacks inside a trend and is poor at telling you whether those pullbacks will hold. Use it in the direction of the bigger trend, smooth it if you’re reading the centerline, and let price confirm before you commit.

Disclaimer: This article is for education only and is not investment advice or a recommendation to buy or sell any security. The stocks and figures in the worked examples are hypothetical. Indicators can give false signals, so test any approach carefully and speak to a SEBI-registered adviser before trading.