The levels do not predict where a pullback will stop. What they give is a plan: a zone in which to look for a turn, a place to put the stop, and targets that follow from the size of the move. Used that way, with a written swing rule and a confirmation candle, the method can be tested and automated like any other. This guide explains how Fibonacci levels are drawn, rules you can test, worked examples on Nifty, what the trade costs, the limits of the method, and how to make it objective enough to backtest and run in QuantPower.
At a glance
| Fibonacci retracement strategy | |
|---|---|
| Type | Pullback entry in a trend |
| Tool | Retracement levels 23.6%, 38.2%, 50%, 61.8%, 78.6% |
| Markets | Nifty, Bank Nifty and stock futures, liquid stocks |
| Timeframes | 15-minute to daily |
| Long entry | Price pulls back to the 50% to 61.8% zone in an uptrend and turns up |
| Stop | Below the 78.6% level or the swing low |
| Target | The previous high, then the 127.2% or 161.8% extension |
A short history of Fibonacci levels in trading
The numbers come from a sequence described by Leonardo of Pisa, known as Fibonacci, in his *Liber Abaci* of 1202: each number is the sum of the two before it, 1, 1, 2, 3, 5, 8, 13, 21 and so on. The ratio between neighbouring numbers settles near 1.618, the golden ratio, and its inverse, 0.618, gives the 61.8% level; 0.382 is 0.618 squared.
Their use in markets came much later. Ralph Nelson Elliott, writing on market waves in the 1930s and 1940s, argued that price moves unfold in patterns whose proportions follow Fibonacci ratios. Robert Prechter and A. J. Frost's *Elliott Wave Principle* of 1978 brought those ideas to a wide audience of traders, and with them the habit of drawing 38.2%, 50% and 61.8% retracements on every chart.1 By the 1990s the levels were a standard drawing tool in charting software and a chapter in textbooks such as John Murphy's *Technical Analysis of the Financial Markets*.2
In India, Fibonacci levels are among the most drawn lines on Nifty and Bank Nifty charts, in trading rooms, broker research notes and television commentary, which is one reason traders keep watching them: a level many people watch can attract orders. Research has been less kind. Studies that tested whether prices turn at Fibonacci levels more often than at other levels have generally not found a reliable effect, and later textbooks, such as Kirkpatrick and Dahlquist's, treat the levels as a planning tool rather than a proven edge.3 That is the spirit of this guide: the levels give structure to a pullback trade, and the evidence for each trade has to come from somewhere else as well.
How Fibonacci retracements are drawn
Pick a completed swing: a clear low and the high that followed it in an uptrend. The levels are the high minus a share of the move:
The 38.2% and 61.8% ratios come from the Fibonacci number sequence. The 50% level is not a Fibonacci ratio, but traders watch it because a half-retracement is a common pause point. The 78.6% level is the square root of 0.618, and a pullback beyond it usually means the move has failed.
Extensions project targets beyond the high: the 127.2% extension is 25,400 + 0.272 × 1,200 = 25,726, and the 161.8% extension is 25,400 + 0.618 × 1,200 = 26,142.
Drawing them in a downtrend
In a downtrend the levels are measured up from the low. If Nifty falls from 25,400 to 24,200, the 38.2% retracement of the fall is 24,200 + 458 = 24,658, the 50% is 24,800, the 61.8% is 24,942 and the 78.6% is 25,143. A bounce into the 50% to 61.8% zone is where a short entry is planned.
Choosing the swing in practice
Most disagreements about Fibonacci levels are really disagreements about the swing. A few habits make the choice consistent:
- Use the most recent completed swing. A move that has made its high and started to pull back. A move still in progress has no high yet.
- Require a minimum size. A swing of at least three times the average true range on the chart's timeframe is large enough for the levels to be meaningful. On daily Nifty, with an ATR of about 190, that is a swing of 570 points or more.
- Use the extremes, not the closes. Measure from the lowest low to the highest high, the convention most traders and charting tools follow.
- One timeframe at a time. A daily swing and an hourly swing give different levels. Pick the timeframe you trade and draw from it; use the higher timeframe only to confirm the trend.
When two swings are both plausible, the larger one usually matters more: more traders see it, and its levels are further apart, which makes the zone clearer.
The rules
Long trades
- Trend: the swing is part of an uptrend, such as price above its 50-period average with higher highs and higher lows.
- Zone: the price pulls back into the 50% to 61.8% zone.
- Confirmation: a bullish reversal candle closes in the zone, such as a hammer or a bullish engulfing candle. See candlestick patterns.
- Entry: above the high of the confirmation candle.
- Stop: below the 78.6% level.
- Target: half the position at the previous high, the rest at the 127.2% extension or on a trailing stop.
Short trades
The mirror image in a downtrend, with retracements measured up from the swing low.
Why the zone, not a line
A single level, such as exactly 61.8%, is hit precisely only by chance. A zone from 50% to 61.8% is wide enough for the market to turn inside it and narrow enough to keep the stop reasonable. Waiting for a reversal candle inside the zone means the trade is taken only after the market has shown some sign of turning, rather than on the hope that the level will hold.
A worked example
Using the swing above, Nifty futures pull back to 24,700, inside the 50% to 61.8% zone, and form a hammer with a high of 24,740.
- Entry: 24,745, above the hammer's high
- Stop: 24,440, below the 78.6% level, a risk of 305 points
- First target: 25,400, a gain of 655 points
- Second target: 25,726
One lot of 65 risks ₹19,825 for a first target of ₹42,575. The numbers illustrate the rules, not a result.4
Splitting the exit
Taking half at the first target needs at least two lots. With two lots, the risk is ₹39,650. If the first lot closes at 25,400 (655 points) and the second at 25,726 (981 points), the gain before costs is ₹42,575 + ₹63,765 = ₹1,06,340. If the market turns at 25,400 and the second lot's stop is moved to the entry after the first target, the second lot closes for roughly costs, and the trade still earns the first target on one lot.
A short trade, worked through
Take the downtrend drawn earlier: Nifty futures fall from 25,400 to 24,200, below their 50-day average. The bounce that follows reaches 24,920, inside the 50% to 61.8% zone of 24,800 to 24,942, and forms a bearish engulfing candle with a low of 24,910.
- Entry: 24,905, below the engulfing candle's low
- Stop: 25,150, above the 78.6% level of 25,143, a risk of 245 points, ₹15,925 per lot
- First target: 24,200, the previous low, 705 points, ₹45,825 per lot
- Second target: the 127.2% extension of the fall, 24,200 − 326 = 23,874
A round trip to the first target costs about ₹961. Short trades in a downtrend tend to move faster than long trades in an uptrend, and the bounce into the zone is often sharp, so the confirmation candle matters even more: without it, the trade sells into a rally that may keep going.
Sizing the trade
Size on the distance from entry to stop. With a ₹10,00,000 account and a limit of 1% a trade, the budget is ₹10,000. The long example risks ₹19,825 on one lot, which is above that budget, so the trade either waits for a tighter set-up or uses a smaller instrument, such as a liquid stock, an option or a spread. The intraday example risks ₹5,200 on one lot, which fits.
The size of the swing decides the size of the stop. A Fibonacci trade on a 1,200-point daily swing has a stop of 300 points or more; one on a 240-point intraday swing has a stop of about 80. Choose the timeframe partly by what your account can carry.
Managing the trade after entry
Four rules cover most of what happens after entry. Decide them before the trade and test them:
- After the first target, move the stop to the entry. The rest of the position can then lose only costs.
- Trail the rest. Below each new swing low, or at a set multiple of ATR, so a new trend can run.
- Use a time stop. If the price has not moved away from the zone within a set number of candles, close the trade. A pullback that stalls in the zone often turns into a deeper fall.
- Respect a close below the 78.6% level. It means the move the levels were drawn from has failed. Do not redraw the levels to keep the trade.
Trading the zone with an option spread
A trader who expects the 78.6% level to hold can sell a bull put spread below it instead of buying futures. With Nifty at 24,745 after the hammer, 15 days to expiry and IV of 14%, selling the 24,450 put and buying the 24,250 put collects about ₹51.77 per unit, ₹3,365 per lot, in the model. The most it can lose is about ₹9,635 per lot if Nifty finishes below 24,250, and it keeps the full credit if Nifty finishes above 24,450.
The spread does not need Nifty to rise to the previous high; it needs Nifty to stay above the zone. It earns less than futures if the uptrend resumes strongly, and it loses less if the zone breaks. QuantPower's Bull Put Spread template in Option bots runs it as an algo with rule legs, an M2M target and stop, and an exit time.
What the trade costs
A round trip on one lot of Nifty futures costs about ₹950 to ₹990 at ₹20 an order and the rates in force from 1 April 2026, most of it STT at 0.05% of the sell value.5 Against a risk of ₹19,825, that is about 5%. On a daily chart, with swings of 1,000 points or more, costs are a small share of each trade. On a 15-minute chart, where a swing might be 240 points and the risk 50 to 60 points, the same ₹950 is about a quarter of the risk; intraday Fibonacci trades are cheaper with an at-the-money option, where STT falls on the premium.
An intraday example
Fibonacci levels work on any timeframe. Suppose Nifty on a 15-minute chart rises from 24,880 to 25,120 in the morning, a 240-point swing. The 50% retracement is 25,000 and the 61.8% is 24,971.7; the 78.6% is 24,931.4.
The afternoon pullback reaches 24,985 and forms a bullish engulfing candle. A long entry above the candle's high at 25,010 with a stop below 24,930 risks 80 points, ₹5,200 per lot. The first target at the 25,120 high is 110 points away, about 1.4 times the risk; that ratio is typical of intraday retracements, where the swings are small compared with the stop. Many intraday traders therefore treat the previous high as the only target and close the trade by the end of the session.
Bank Nifty and stocks
On Bank Nifty futures (lot 30), daily swings are usually twice as large in points as Nifty's, so a 2,400-point swing gives a 50% to 61.8% zone about 280 points wide and a stop below the 78.6% level several hundred points from entry. The rupee risk per lot ends up similar to Nifty's because the lot is smaller.
On stocks, the method works best on liquid names in clear trends, on the daily chart. Stocks gap more than the index around results, dividends and news, so check the calendar before entry and size for a gap through the stop. Avoid stocks in the F&O ban list or with thin trading, where the zone can be crossed in a single trade.
The limits of Fibonacci levels
- The swing is a choice. Two traders can pick different highs and lows and get different levels.
- There are many levels. With five levels close together, the price is near one of them most of the time. A level "working" is often hindsight.
- No proven edge on its own. Studies of Fibonacci levels have not shown that prices respect them more than other round numbers. Treat them as a way to plan entries and stops, and look for agreement with other evidence: a support and resistance strategy level, a moving average, VWAP or a round number.
- Self-fulfilling at best. If many traders watch the same levels, orders cluster there, which can make a level matter for a while. That effect is fragile and disappears when the crowd uses different swings.
When the strategy works, and when it does not
Works best when:
- The market is in a steady trend with orderly pullbacks, the kind that retrace part of the last move and resume.
- The zone lines up with another level, such as a previous high, a moving average or a round number.
- The pullback slows as it reaches the zone, with smaller candles and falling volume.
Struggles when:
- The trend is ending. The last pullback of a trend often goes straight through 61.8% and 78.6%.
- The market is in a range. Swings are small and the levels sit on top of each other.
- News drives the move. A pullback caused by a sudden event has no reason to respect ratios drawn from the move before it.
- Volatility jumps. A wider range on every candle turns a reasonable stop into one that is hit by noise.
Finding agreement with other evidence
A Fibonacci zone becomes more useful when it lines up with something else. Common combinations:
| Evidence | What it adds |
|---|---|
| A previous support or resistance level inside the zone | A second reason for buyers or sellers to act there |
| The 50-day moving average inside the zone | A level that trend followers watch |
| VWAP, intraday | The average price of the day's trading |
| A round number such as 25,000 | Option strikes and orders cluster at round numbers |
| An oversold oscillator, such as RSI below 40 in an uptrend | Momentum in the pullback is fading |
A zone with two or three of these is a better place to look for a turn than a zone with only the ratio. Write the combination into the rules before testing, so it is the same every time.
Making the method objective
To backtest or automate the strategy, the swing must be defined by a rule, for example:
- the highest high and lowest low of the last 50 candles, with the low before the high for an uptrend, or
- a zigzag rule that marks a swing after a reversal of a set percentage, such as 2%, or
- a confirmed swing: a high that has a set number of lower highs on each side, so it is known only after the price has turned.
With a fixed rule, the levels can be calculated the same way every time. The third rule is the safest for testing, because it never uses a swing before it could have been known.
Building the strategy in QuantPower
Arth.live charts inside QuantPower include Fibonacci drawing tools among their 89 drawing tools, and an Auto Fib indicator that draws the 0%, 23.6%, 38.2%, 50%, 61.8%, 78.6% and 100% levels from confirmed swings. Because it uses confirmed swings, it does not move a level after the fact, which makes it suitable for signals and backtests.
- Add Auto Fib to the chart of Nifty futures on the timeframe you trade, and set how many candles confirm a swing.
- Open Algos → Signal algos and create a signal on the same instrument and timeframe. Conditions are checked on closed candles.
- Add the conditions, all required: close above the 50-period EMA (the trend); close below the Auto Fib 50% line and above the 61.8% line (the zone); a Hammer or Bull Engulfing candle (the confirmation). Candle patterns are available as conditions, and a rule can hold up to 20 conditions.
- Choose the action. One-touch, the default, notifies you and prepares the order for one touch, which suits a method where many traders want to look at the chart before entering. Auto places the order by itself after you arm it, with every risk check.
- Protect the trade with a GTT OCO: the stop below the 78.6% level and the first target at the previous high. The first to trigger cancels the other.
A signal that places its own order costs no tokens. An algo that a signal starts is charged as a live algo, 200 tokens per 28 days.
Backtesting a Fibonacci strategy
QuantPower's rule backtest tests a condition on one instrument over up to 2,000 past candles from your connected broker, from 5-minute to daily, with a stop-loss and target in per cent, an exit rule, slippage and an estimate of charges. See backtesting for the full method.
- Fix the swing rule before you test, and do not change it to fit results.
- Compare entries at the 50% to 61.8% zone with entries at random points in the same pullbacks. If the levels add nothing, the results will be similar.
- Include trades where the price went straight through the zone.
- Test the zone itself: 38.2% to 50%, 50% to 61.8% and 61.8% to 78.6%. If only one zone works and its neighbours do not, the result is probably noise.
See look-ahead bias: a swing high is only known after the price has turned, so a backtest must not use it before then.
Read the report in this order:
- The trade list. Check that each entry really came after a confirmation candle in the zone, and that the swing was known at the time.
- The share of trades that reached the first target. It says how often the zone held. Compare it with the share for random pullback entries.
- Average win against average loss. With a first target about twice the risk, the strategy can afford to be right less than half the time.
- Results by month. Fibonacci pullbacks do best in steady trends; expect flat stretches in ranges.
- Exit reasons. Many time-stop exits mean the zone is often a pause rather than a turn.
For the daily example, the 305-point stop on 24,745 is about 1.2% and the first target about 2.6%. Then run the signal in forward test on live prices with test money before going live.
Account safety above the strategy
Every signal and algo sits under the account's own controls:
- Smart Exit sets a profit target, a stop-loss and a trailing step for the whole account. On a hit it stops the algos and squares off.
- The daily loss limit refuses new orders once today's loss reaches it.
- The kill switch stops trading in a segment until that segment's next session, and stops the algos running in it.
A daily pullback trade is held overnight, so gaps are the main risk to the stop. Size for a gap, not only for the distance to the stop.
Common mistakes
- Redrawing the levels after the fact. Pick the swing before the pullback, not after.
- Buying the level without confirmation. A falling price passes through every level on the way down.
- Ignoring the trend. Retracements are for trends. In a range, the levels mean little.
- Using too many levels. If every level is a reason to trade, there is always a reason.
- Stops just below the level. A stop a few points under 61.8% sits where many other stops sit; use the 78.6% level or the swing low.
- Forgetting costs on short timeframes. Small swings and futures STT turn many small intraday wins into losses after costs.
- Treating a famous example as proof. A memorable turn at 61.8% is easy to recall; the many pullbacks that went straight through are not.
Where Fibonacci trading is going
The levels themselves have not changed in decades. What has changed is how they are used. Automatic tools that draw levels from confirmed swings remove much of the personal choice that made the method hard to test, so traders can now check whether a zone adds anything to their pullback rules instead of relying on memorable examples. Higher STT on futures from 1 April 2026 favours longer timeframes, where swings are large compared with costs. And the retail algo framework in force since 1 April 2026 means a Fibonacci strategy that runs as an algo needs its swing rule, zone and confirmation written down precisely, which is a good discipline in itself.6 The test of the method has always been the same: whether a written rule, applied the same way every time, adds anything after costs.
Start with 500 free tokens
Connect your broker, build a strategy and test it on paper before a single live order goes out.
Frequently asked questions
Which Fibonacci level is the most important?
Traders watch 61.8% and 50% the most. A pullback beyond 78.6% often means the trend has failed, and many traders place their stop just beyond it.
Does Fibonacci work in intraday trading?
The levels can be drawn on any timeframe, but swings on short charts are small and noisy. Use them with other evidence.
What is the golden ratio in trading?
The 61.8% level, derived from the golden ratio of about 1.618.
Are Fibonacci extensions reliable targets?
They are a way to plan targets, not a forecast. Many traders take part of the profit at the previous high and trail the rest. The 127.2% extension is a modest second target; the 161.8% extension is reached only in strong trends, so a trailing stop often serves better beyond the first target.
Should I enter at 50% or wait for 61.8%?
Neither on its own. Treat 50% to 61.8% as one zone and wait for a confirmation candle anywhere inside it. Entering at a fixed level without confirmation buys pullbacks that keep falling.
Can QuantPower draw Fibonacci levels for me?
Yes. The Auto Fib indicator draws the levels from confirmed swings, and the Fibonacci drawing tools let you draw your own. The Auto Fib lines can be used in signal conditions.
Related guides
- Multi timeframe strategy: confirm a pullback on a higher timeframe
- RSI indicator strategy: another way to time pullbacks
- Option strategy builder: plan an option trade around a level
References
- A. J. Frost and Robert R. Prechter Jr., *Elliott Wave Principle* (1978)
- John J. Murphy, *Technical Analysis of the Financial Markets* (New York Institute of Finance, 1999)
- Charles D. Kirkpatrick and Julie R. Dahlquist, *Technical Analysis: The Complete Resource for Financial Market Technicians* (FT Press, 2010)
- NSE, equity derivatives contract specifications (nseindia.com)
- Union Budget 2026-27: securities transaction tax on futures and options raised from 1 April 2026 (indiabudget.gov.in)
- SEBI, circular on safer participation of retail investors in algorithmic trading, 4 February 2025, in force for every broker from 1 April 2026 (sebi.gov.in)