RSI Length Test: How Long Overbought and Oversold Last

RSI above 70 is called overbought, and RSI below 30 is called oversold. Those labels sound like a reversal is due. In a strong move, however, the line can remain beyond its threshold for several bars while price continues in the same direction.

On hourly USD/JPY with RSI 14, there were 312 complete runs that entered at or above 70 in 2023–2025. The next bar was still at or above 70 in 188 of them, or 60.3%. Below 30, 181 of 300 runs continued into the next bar, also 60.3%.

The two sides differed in their long tails. At least ten overbought bars followed in 19.2% of runs, against 6.3% for oversold runs. An overbought or oversold reading describes a lopsided recent move. It does not identify the bar on which that move must reverse.

Lengths 7, 9, 14, 21, 30, and 50 were also compared. Persistence did not move steadily as the length changed. The repeatable change was how rarely a longer RSI reached 70 or 30.

at or above 70 (312 runs)at or below 30 (300 runs)USD/JPY hourly, RSI 14, 2023–2025

What RSI overbought and oversold actually mean

RSI converts average gains and average losses over the last N bars into a value from 0 to 100.

RSI = 100 − 100 ÷ (1 + average gain ÷ average loss)

TradingView's RSI documentation describes readings above 70 as overbought and readings below 30 as oversold. The calculation itself only reports the balance of recent gains and losses. Crossing 70 or 30 does not put a reversal condition into the formula.

"Staying overbought" in this test does not mean RSI printed exactly 100. It means the line closed at or above 70 for consecutive bars. Likewise, an oversold run consists of consecutive closes at or below 30. With RSI 14, none of the three timeframes printed exactly 100 or 0 during the three years.

RSI settings and how a run was counted

The baseline uses the conventional RSI 14 with 70 and 30 as the two boundaries.

If RSI reads 68 → 72 → 76 → 71 → 66, the overbought run lasts three bars. The first 72 is bar one, 76 is bar two, 71 is bar three, and the run ends at 66. A one-bar run leaves the zone immediately. A two-bar run remains there on the next bar.

RSI is normally shown in a separate pane below price. Set the length to 14 and add horizontal levels at 70 and 30. The main RSI settings test covers platform setup, conventional 30/70 fades, and 50-line momentum systems. This article isolates the duration of each visit beyond 30 or 70.

How this was measured

One continuous visit to a zone counts as one run. Ten consecutive bars above 70 are one observation, not ten. A run that had not left the zone by year-end was excluded because its final length was unknown.

The trading comparison separates a fade entered as the run begins from one entered after RSI leaves the zone. Both use the next bar's open, after the signal bar is complete. Every run is treated as an independent event, so summed pips are not a compounded portfolio return and trades can overlap in time.

The next bar remained extreme about 60% of the time

The table pools 2023 through 2025 for RSI 14 at 70/30.

Across every timeframe, 58% to 71% of runs lasted into the next bar. Once RSI first entered a zone, remaining there was more common than leaving immediately.

Two four-hour bars represent eight hours. The apparent persistence is therefore meaningful in elapsed time, but the four-hour samples contain only 103 overbought and 53 oversold runs. Their percentages are less precise than the fifteen-minute and hourly estimates.

Long overbought runs were more common than long oversold runs

The asymmetry in long hourly runs appeared in all three years.

The one-bar continuation rates were similar. The difference appeared in long runs: roughly one overbought run in five lasted ten hourly bars, compared with about one oversold run in twenty.

This does not say that a high RSI predicts a decline. It says that, in these three USD/JPY years, strong upward moves kept RSI above 70 for longer than strong downward moves kept it below 30. Indicator persistence and the direction of the next price move are separate questions.

Even 80/20 did not force an immediate reversal

These are hourly RSI 14 runs pooled across all three years.

Moving the lines to 80 and 20 did not make the first extreme reading a turning point. More than half still lasted into the next bar. The stricter pair also cut the sample to 76 upper runs and 47 lower runs, so its percentages carry much more sampling uncertainty.

Wider zones naturally produce more and longer runs because values nearer 50 now count as extreme. Comparing the probabilities without the number of runs and the width of the zone would make 60/40 look more informative than it is.

Changing RSI length reduced the sample before it changed persistence

The 70/30 lines stay fixed while hourly RSI length changes. Upper and lower runs are combined in this table.

As length rises from 7 to 50, the share of bars beyond 70 or 30 falls from 27.91% to 0.66%. The number of complete runs falls from 1,557 to 23. Smoothing pulls a longer RSI toward 50, so it reaches either outer zone less often.

Next-bar persistence, by contrast, ranges from 56.5% to 64.7% without a steady trend. Five- and ten-bar persistence appear to rise at long settings, but length 50 contains only 23 runs, nine from the lower zone. Adding 2023 cuts its next-bar result from 71.4% in the two-year sample to 56.5% across three years. That swing is what a small sample looks like.

at least 2 barsat least 5 barsat least 10 barsupper and lower zones combined; USD/JPY hourly, 2023–2025

Separating the two sides does not create a clean relationship. Next-bar persistence is 60.0% versus 55.1% at length 7, 60.3% on both sides at length 14, and 59.7% versus 52.7% at length 30. The most reproducible effect of changing length is signal frequency, not run duration.

Only one of 144 period-based fades was positive in all three years

The trading scan combines six RSI lengths, two entry timings, and four holding periods on each timeframe. That is 48 conditions per timeframe and 144 across all three. Upper-zone shorts and lower-zone buys are combined.

Pips are average return per event after the recorded bid/ask spread. The only condition positive in all three years has just two events in 2025. That is too little evidence to call length 50 superior.

Changing length alone did not make the fade repeatable. The lone condition positive in all three years had the smallest 2025 sample in the scan, so its event count cannot be omitted.

Fading immediately and waiting for the exit both failed to repeat

If RSI can stay overbought, waiting for it to fall back below 70 before shorting sounds safer. The built-in TradingView RSI Strategy uses this type of exit from the zone: it buys a cross back above the oversold line and shorts a cross back below the overbought line.

The table combines upper-zone shorts and lower-zone buys on hourly RSI 14. Values are average pips per event after the recorded bid/ask spread.

The immediate one- through five-bar fades were positive in 2024 and negative at the same settings in 2023 and 2025. Waiting for RSI to leave the zone produced only one positive cell, the ten-bar hold in 2025.

The five-bar comparison reached the same conclusion across the other timeframes. Neither entry timing was profitable in all three years on the fifteen-minute, hourly, or four-hour chart. Leaving the zone can be a confirmation event, but it was not a complete trading edge here.

fade as the zone beginsfade after RSI leaves the zoneshort above 70, buy below 30; hold five bars

What this test supports

1. The next RSI bar stayed beyond 30 or 70 about 60% of the time. Hourly RSI 14 measured 60.3% on both sides.
2. Long overbought runs were more common. On the hourly chart, 19.2% of upper runs lasted ten bars, against 6.3% of lower runs.
3. Moving to 80/20 did not remove persistence. The next-bar rates were 57.9% and 55.3%.
4. Longer lengths reduced event counts before they changed persistence. Hourly runs fell from 1,557 at length 7 to 23 at length 50.
5. Only one of 144 period-based fades was positive in all three years. Its 2025 sample contained two events, too few to identify a better RSI length.

The 70 and 30 lines mark a state, not a reversal time. A practical test must distinguish entering the zone from leaving it, then include event counts, holding time, and the spread.

Related tests

The main RSI settings test compares 330 combinations of 30/70 fading and 50-line momentum. This follow-up isolates how long each overbought or oversold visit lasts.

Stochastic RSI settings apply another range calculation to RSI itself. The Bollinger Bands fade-versus-breakout test asks the same broader question on price: whether an extreme should be faded immediately or followed as a breakout.

Notes

• The sample contains only USD/JPY in 2023–2025. Other markets and regimes can have different run lengths.
• RSI uses closing prices. Intrabar touches of 70 or 30 that did not survive to the close are excluded.
• Runs already open at the start of a year and runs still open at year-end are excluded because their full duration is unknown within that year's sample.
• The 80/20 and RSI 50 samples are small, so their percentages are uncertain.
• Each run is treated as an independent event. Trades can overlap, and summed pips are not a portfolio return.
• Trend, volatility and session filters, stops, and profit targets were not tested.