Tata Elxsi sat with RSI above 70 for 47 consecutive trading sessions in 2021. Every textbook would have told you to sell. Every "overbought" alert would have fired. And if you'd listened, you'd have missed another 280% rally sitting right in front of you.
That's not a cherry-picked anomaly. It's a structural problem with how most retail traders use RSI — and once you see it clearly, you'll stop treating 70 like a stop sign.
RSI Was Never Designed to Be a Sell Signal
J. Welles Wilder introduced RSI in 1978 in a completely different market context — slower-moving commodities, thinner liquidity, fewer algorithmic participants. The 70/30 thresholds he suggested were guidelines, not laws.
RSI measures the speed and magnitude of price changes on a 0–100 scale. A reading above 70 simply means recent up-moves have been larger and faster than recent down-moves over the default 14-period lookback. That's it. It says nothing about whether the stock is expensive. It says nothing about whether buyers are about to disappear.
What it actually captures is momentum. And in a momentum market — which India frequently becomes during sector rotations and bull phases — momentum sustaining above 70 is often a feature, not a warning.
The Tata Elxsi Problem (And Why It Keeps Repeating)
In 2021, Tata Elxsi went from around ₹2,000 in January to over ₹9,500 by December. During that run, RSI crossed 70 early in the year and largely stayed there through multiple breakouts. Retail traders who sold the "overbought" signal in February, March, or even June missed the bulk of that move.
This isn't unique to Tata Elxsi. Laurus Labs in 2020 held RSI above 70 through a near-400% move. Tanla Platforms did the same. Dixon Technologies in 2020-21 printed RSI above 70 for weeks at a stretch while tripling in price.
The common thread: these were stocks with a genuine fundamental re-rating happening simultaneously — earnings acceleration, sector tailwinds, or a structural business shift. RSI staying elevated was confirmation that institutional money was continuously buying, not a sign of exhaustion.
Selling because RSI hit 70 in these names wasn't disciplined trading. It was leaving money on the table while citing a textbook.
When RSI 70 Actually Works As a Warning
Here's where it gets more useful. Adani Enterprises had RSI above 70 for over 100 days through much of 2022, and traders who pointed to that as proof "RSI doesn't work" missed the real signal — divergence.
By late 2022 and into early 2023, the stock was making higher price highs while RSI was printing lower highs. Classic bearish divergence. That divergence, not the raw 70 level, was the actual warning the indicator was offering. The Hindenburg report in January 2023 was the trigger, but the technical structure had already started showing cracks weeks before.
RSI divergence is a completely different tool than RSI levels. Most retail traders never graduate to using it.
The stocks where the 70-means-sell rule does hold reasonably well are range-bound, low-volatility compounders: HDFC Bank, ITC, Hindustan Unilever. These names tend to mean-revert efficiently because institutional ownership is so high that no single catalyst shifts the entire ownership base quickly. When HDFC Bank RSI hits 72–74, you've frequently had a 3–5% pullback follow within two to three weeks. The RSI works there because the underlying price behavior is mean-reverting by nature.
The mistake is applying that same logic to a mid-cap IT services stock or a specialty chemicals name in the middle of a sector re-rating.
How to Read RSI Differently From Tomorrow
The first thing to do is stop looking at RSI in isolation. RSI at 72 on a stock that broke out of a 52-week base on 3x average volume is a completely different signal than RSI at 72 on a stock that's been drifting up on thin volume for three weeks.
A few practical filters:
- Volume confirmation: If RSI is above 70 and daily volumes are 1.5x to 2x the 20-day average, the move has participation. That's not overbought, that's institutional accumulation.
- Sector context: If the broader sector index is also in an RSI uptrend, individual stocks holding above 70 often reflect a sector rotation cycle. Selling early means sitting in cash while the cycle plays out.
- Lookback period: The default 14-period RSI is designed for daily charts. On a weekly chart, RSI crossing 70 is a much bigger signal — fewer false positives, but also rarer. Many traders don't switch lookbacks between timeframes.
- Divergence over levels: Train yourself to draw trendlines on the RSI indicator the same way you draw them on price. If price is making new highs but RSI peaks are declining over three to four weeks, that's your genuine exit signal.
One more thing: reduce the RSI period to 9 if you're trading intraday or swing trading on the 15-minute or hourly chart. The 14-period RSI lags too much on shorter timeframes. On NSE's liquid large-caps — Reliance, HDFC Bank, Infosys — a 9-period RSI on the 15-minute chart gives cleaner divergence signals during the first and last hour of the session.
The Structural Reason Indian Markets Break This Rule More Often
India's equity market has a relatively smaller free float in many mid and small-cap names. When a stock with, say, ₹3,000 crore market cap starts getting attention from a single mid-cap fund, the buying pressure can be disproportionate relative to available supply.
This means momentum can sustain far longer than developed markets theory would predict. RSI staying above 70 for 30, 40, 60 sessions is genuinely more common on NSE/BSE than it would be on NYSE for equivalent-sized stocks. The market microstructure makes it so.
Add to that the relatively high participation of retail traders on platforms like Zerodha, Groww, and Angel One — many of whom are using basic RSI rules from YouTube tutorials — and you get a perverse outcome: the "overbought" signal triggers selling pressure from retail while institutions are still accumulating. That supply gets absorbed, the stock doesn't fall, and retail exits a winner too early.
SEBI's data consistently shows retail traders underperform in F&O and equities over rolling 3-year periods. Part of that underperformance is precisely this: rule-based exits triggered by indicators that weren't designed for the way these participants are using them.
What You Should Actually Do With RSI From Here
Use RSI as a momentum gauge, not a traffic light. Above 70 doesn't mean stop. Below 30 doesn't automatically mean buy. Both are just data points about recent price velocity.
Build the habit of checking RSI divergence on the weekly chart before making any exit decision on a momentum stock. If RSI is still making higher highs along with price, the trend has internal strength. If RSI starts lagging price — three to four weeks of price climbing while RSI stays flat or drops — that's when you tighten your stop or book partial profits.
For mean-reversion names like HDFC Bank or ITC, keep using the 70/30 levels. They're appropriate for stocks that behave that way.
For momentum names in sectors like capital goods, defence, specialty chemicals, or IT services mid-caps — stocks that can re-rate 200–400% in a cycle — treat RSI above 70 as confirmation that something real is happening. Let divergence tell you when it's ending.
Tata Elxsi didn't stop going up because RSI crossed 70. It eventually stopped because earnings growth started decelerating and the sector multiple compressed. By the time that happened, RSI divergence had been visible for weeks. The indicator was working perfectly. It just wasn't working the way the textbook said it should.