RSI Explained: The 30/70 Rule Most Traders Get Wrong

The Relative Strength Index — RSI for short — is the most-used momentum indicator in retail trading. Most investors misuse it. The 30/70 rule that triggers their buys and sells is only half the story, and in trending markets it is the worst half.

What is RSI?

RSI is a bounded oscillator that converts recent price action into a single number between 0 and 100. It was introduced by J. Welles Wilder in his 1978 book "New Concepts in Technical Trading Systems" and has since become the most widely used momentum indicator on every major charting platform.

The intuition is straightforward. If a stock has been rising quickly, its RSI will be high — because recent gains are outweighing recent losses. If it has been falling, RSI will be low. The number tells you how lopsided the recent buying pressure has been relative to selling pressure.

Wilder himself suggested 70 and 30 as the default thresholds for overbought and oversold conditions. Those are conventions, not physical laws. Different traders use different thresholds — 80/20 for very volatile names, 65/35 for mean-reverting equities, 60/40 in strong trends. The thresholds should always be tuned to the instrument and the regime.

How is RSI actually calculated?

The formula is a three-step process. Pick a lookback — usually 14 trading days. Compute the average gain and the average loss across those 14 periods (using only positive or only negative price changes, respectively). Then apply:

RS = Average Gain ÷ Average Loss RSI = 100 − (100 ÷ (1 + RS))

The intermediate value RS is the "relative strength" — a ratio of how strong the up days were versus the down days. The transformation at the end compresses that ratio into a 0-100 scale.

After the first 14 periods, the calculation uses a smoothing step called Wilder smoothing — essentially an exponential moving average applied to the gain and loss streams. This keeps RSI from being wildly volatile day to day and is why values don't change dramatically from one close to the next unless the stock itself has moved sharply.

How do you use the 30/70 rule?

The short answer: carefully, and almost never in isolation. The conventional reading is simple — above 70 means overbought, below 30 means oversold. The implication is that you sell above 70 and buy below 30.

That reading is correct in sideways markets. It fails in trending ones. During a real rally, strong stocks can maintain an RSI of 70 or higher for weeks at a time. Selling every time they print 70 means exiting your winners early and chasing shorts that never arrive.

Here is a more disciplined framework:

• In a clear uptrend (price above rising 50-day moving average), 70 is not a sell signal — it is a trend-strength confirmation
• In a clear downtrend (price below declining 50-day moving average), 30 is not a buy signal — it is a trend-strength confirmation
• In sideways / range-bound markets, 70 and 30 are useful mean-reversion triggers
• In all regimes, 80/20 readings paired with exhaustion volume spikes are more reliable turning-point candidates

The practical rule: always check the trend before interpreting RSI. The trend is the primary context; RSI adds refinement.

Real examples: RSI on major names

Look at how RSI behaves on different names over 2024-2026. These are illustrative patterns, not forecasts.

NVDA in 2024 was the textbook example of how the 30/70 rule fails in a real trend. The stock printed RSI above 70 for long stretches; traders who sold at each print missed most of the multiyear run. The better play was to hold through the strength and use RSI divergence — not absolute level — as an exit flag.

META in late 2022 was the opposite lesson. RSI spent months below 30 as the stock collapsed. The eventual bottom was preceded by bullish divergence — price made a lower low, but RSI made a higher low. That divergence, not the "oversold" label, was the signal that the worst was priced in.

XOM in most macro regimes is a cleaner 30/70 name because energy tends to oscillate within a range. When RSI gets to 70 on XOM without a sustained Brent breakout, sellers usually appear. When it hits 30 without an actual supply shock, dip-buyers usually appear.

What mistakes do traders make with RSI?

The answer is: they treat a context-dependent tool as a context-free signal. Three specific mistakes dominate.

First, selling rallies too early. A stock in a strong uptrend can ride RSI 70+ for weeks. Traders who reflexively sell at the first 70 print get chopped up in the noise, pay extra in fees, and miss the meat of the move. This is the single most expensive RSI mistake.

Second, catching falling knives. A stock in a real downtrend can stay below 30 all the way to its lows. Buying just because RSI reads oversold — without confirming bullish divergence or a change in volume structure — is how traders amplify drawdowns in crashing names.

Third, ignoring divergence. Divergence is when price makes a new extreme (a higher high or lower low) but RSI fails to confirm. This is frequently a more reliable signal than the absolute level. Our technical analysis primer covers divergence patterns in detail, including hidden divergences that experienced traders watch for on the pullbacks of a larger trend.

When should you NOT use RSI?

There are three environments where RSI is the wrong tool.

Very low-liquidity stocks — thinly traded small caps, meme names with wild single-day moves — break the mathematical assumptions RSI was built on. The "average gain" and "average loss" inputs are distorted by single large trades, and the resulting reading does not reflect crowd behavior.

Event-driven situations — earnings week, FDA catalysts, takeover rumors — overwhelm the noise RSI is measuring. A stock can go from RSI 30 to RSI 80 in two sessions on a takeover leak. The indicator didn't fail; it simply is not what you want when fundamentals are re-pricing at high speed.

Passive long-term investing in diversified index funds. If your strategy is monthly contributions to a broad market ETF, RSI adds nothing. You are already not trying to time entries or exits. Adding an oscillator to that strategy is a distraction, not an edge.

Pro tips for using RSI

Three habits separate disciplined RSI users from the rest.

First, always pair RSI with a trend filter. The simplest is a 50- or 200-day moving average. Only take "oversold" signals when the trend is up, and only take "overbought" signals when the trend is down. This single rule eliminates most of the false signals retail traders chase.

Second, watch for divergence before reversal. Price lower low plus RSI higher low is the classic bullish divergence. Price higher high plus RSI lower high is the classic bearish divergence. Neither is guaranteed to trigger a reversal, but each raises the probability enough to justify a smaller starter position.

Third, size small on RSI-only trades. If RSI is your only input, assume the base rate of success is modest — probably not much better than flipping a coin. Use it as a confluence signal, not a primary one. Combine it with trend, volume, and at least one fundamental catalyst.

Counter-argument: RSI still works for many traders

Not every momentum trader agrees with the bearish take on the raw 30/70 rule. Systematic quant funds have published studies showing that — over long samples, across large baskets of equities — mean reversion after extreme RSI readings does generate a statistical edge. The key is the word "basket." Applied to a single stock, signal-to-noise is poor. Applied to 100 stocks with proper position sizing, a 30/70-style mean-reversion system can be profitable.

The takeaway for retail investors: the indicator is not broken. The single-stock discretionary version most people use is broken. If you want to trade mean reversion systematically, use it on a wide basket with disciplined position sizing and stop losses — not on whichever name happens to be on your watchlist today. For most long-term investors, it is simpler to lean on fundamental analysis and investor-based valuation frameworks than to chase momentum oscillators.

Ready to analyze these stocks yourself? Search any ticker on MainRatios to see valuations from 6 legendary investors - free.