Fibonacci 38.2% retracement: does it hold?
After a rise, price often pulls back part of the way before moving on. A bounce near 38.2% of the rise is a shallow pullback: buyers stepped in early, a sign of a strong trend.
When “Fibonacci 38.2% retracement” appeared, NIFTY 200 stocks were higher 12 trading days later 56.1% of the time, against 55.3% after any day: about the same as picking a random day.
| After | Higher | Average move | Median move | Higher after any day |
|---|---|---|---|---|
| 5 trading days | 53.0% | +0.6% | +0.4% | 53.9% |
| 12 trading days | 56.1% | +1.4% | +1.0% | 55.3% |
| 20 trading days | 57.3% | +2.4% | +1.6% | 56.8% |
Year by year
| Year | Signals | Higher after 12 days | Any day |
|---|---|---|---|
| 2021 | 182 | 56.0% | 54.4% |
| 2022 | 360 | 46.1% | 52.1% |
| 2023 | 417 | 60.9% | 61.3% |
| 2024 | 372 | 58.6% | 53.7% |
| 2025 | 167 | 59.9% | 53.1% |
| Large cap only | 751 | 57.0% | |
| Mid cap only | 747 | 55.2% |
What it is
After a rise, price often pulls back part of the way before moving on. A bounce near 38.2% of the rise is a shallow pullback: buyers stepped in early, a sign of a strong trend.
How we detected it
We find the last up-swing: the highest high from 45 to 3 days ago, and the lowest low in the 40 days before it (at least 5 days earlier). It must have risen at least 8%. The lowest low since the swing high is nearer the 38.2% level than the 50% or 61.8% levels, and within 4% of the swing’s size from it. The close is above the 38.2% level and more than 1% above that low. We count the first day this is true.
How traders read it
Fibonacci levels are zones, not exact prices. Traders want a bullish candle or a reclaimed moving average at the level, and place a stop below the low of the pullback.
A signal is a probability, not a promise
Even the most reliable signal in this study was followed by the opposite move 37% of the time. A signal tilts the odds; it doesn’t decide the outcome.
Signals work better when they agree with the trend and with each other. Whatever the signal, decide your stop-loss before you act.
Practise spotting it, free
The Fibonacci retracements chapter teaches this signal in five minutes, then you call real NSE charts and get graded on your reasoning.
Common questions
Does the fibonacci 38.2% retracement signal work on Indian stocks?
When “Fibonacci 38.2% retracement” appeared, NIFTY 200 stocks were higher 12 trading days later 56.1% of the time, against 55.3% after any day: about the same as picking a random day. No single signal is reliable on its own; traders combine it with the trend and manage risk with a stop-loss.
How often did stocks rise after this signal?
Across 1,498 signals on 184 NIFTY 200 stocks (2021–2025), the price was higher after 5, 12 and 20 trading days 53.0%, 56.1%, 57.3% of the time. After any day it was 53.9%, 55.3%, 56.8%.
How is “Fibonacci 38.2% retracement” defined here?
We find the last up-swing: the highest high from 45 to 3 days ago, and the lowest low in the 40 days before it (at least 5 days earlier). It must have risen at least 8%. The lowest low since the swing high is nearer the 38.2% level than the 50% or 61.8% levels, and within 4% of the swing’s size from it. The close is above the 38.2% level and more than 1% above that low. We count the first day this is true.
How these numbers were made
We ran the same signal detector MarketSeekho uses to grade players over every trading day of 187 NIFTY 200 stocks, Jan 2021 to Oct 2025, using NSE end-of-day prices (1,54,833 stock-days). We counted the first day a signal appeared, then ignored the same signal on the same stock for 12 trading days so results don’t overlap. Windows with splits, bonuses or trading halts were left out.
“Any day” is the baseline: the same measure after every day in the data, signal or not. It matters because 2021–2025 was mostly a rising market, so stocks were higher after 12 days more often than not anyway. Past results don’t predict future ones, and this is not investment advice.