Williams Fractals in Indian Markets: A Worked Nifty Example

What a Williams Fractal Actually Is

A Williams Fractal is a simple swing-point marker built from five consecutive candles. A down fractal (the bullish one) appears when the middle candle has the lowest low of the five, with two candles on each side making higher lows. An up fractal (the bearish one) appears when the middle candle has the highest high, flanked by two candles each side making lower highs. The indicator simply prints a small arrow above or below that middle candle once the pattern completes.

The key word is completes. Because two candles must form after the middle candle, a fractal cannot be confirmed until the fifth candle closes. This means every fractal you see on the chart is, by definition, at least two bars old. On a 15-minute Nifty chart, a fresh down fractal confirms roughly 30 minutes after the actual low was made. Traders who forget this lag chase signals that have already moved.

Bill Williams designed fractals as part of a larger system in his book Trading Chaos, not as a one-line buy or sell rule. In his original framework, a fractal only counts as a real signal once price trades beyond it, and ideally only in the direction of the Alligator (his three-moving-average trend tool). On the NSE and BSE this distinction matters because Indian indices spend long stretches range-bound, where raw fractals fire constantly and most lead nowhere.

Do Williams Fractals Repaint? The Two-Bar Delay

A fractal needs two bars to close after the middle bar before it exists. Platforms then draw the arrow on the middle bar, which is already two bars in the past. LuxAlgo's fractal script describes it this way: with a period of 2, the indicator must wait for 2 bars to close after the high or low to confirm the pattern. Community scripts on TradingView say the same, that the marker always appears in the past.

So what looks like repainting is really a delay. Once a standard five-bar fractal has been drawn, the closed bars that formed it do not change. Two things can still fool you:

• The live bar. A pattern can look complete while the last bar is still open. If price then makes a new high or low, the pattern may disappear. Wait for the bar to close.
• Looking at history. On an old chart every arrow is already there, so it looks as if the indicator called each turn in real time. It could not have: each arrow became visible two bars after the turn.

To avoid acting on an unconfirmed fractal, use the level and not the arrow. Wait for a close beyond the fractal's high or low, then decide. The example below shows how.

The Five-Bar Rule, Step by Step

To read a fractal correctly, label the middle candle as bar 3 and count two bars on each side. For a confirmed down fractal you need: low of bar 3 lower than low of bar 1, bar 2, bar 4 and bar 5. For a confirmed up fractal you need: high of bar 3 higher than high of bar 1, bar 2, bar 4 and bar 5. If even one neighbouring candle breaks the rule, there is no fractal.

Some charting platforms let you widen the pattern to seven or nine bars. A wider fractal is rarer and marks a more significant swing, but it confirms even later. For Indian intraday work on Nifty and Bank Nifty, the standard five-bar fractal on a 15-minute chart is the common default. For positional swing trades on stocks like Reliance or TCS, the same five-bar fractal on the daily chart filters out most intraday noise.

How Bill Williams Used Fractals

In his book Trading Chaos, Bill Williams treated a fractal as a breakout level and not as a prediction. The idea, as it is usually described, is that the most recent up fractal marks a level where buyers failed once. If price later trades above it, that is a sign that buying pressure has taken over, and a trader may place a buy order just above the fractal's high. A down fractal works the same way in reverse for a sell order below its low.

Williams paired this with his Alligator, three moving averages that show whether the market is trending or asleep. A breakout fractal inside a sleeping Alligator is treated as weaker than one that forms away from it. The Alligator and fractals were two parts of a larger method, and he did not offer fractals as a stand-alone system. Remember that any breakout order fills at the next available price, which can be worse than the fractal level.

A Worked Example on Nifty

Here is a fully worked down fractal long using made-up index levels. All figures are illustrative, not a prediction or a guarantee.

Imagine a 15-minute Nifty chart during a pullback inside an uptrend. Five candles print, and the middle candle bottoms at 24,920. The two candles on each side hold higher lows, so a down fractal confirms when the fifth candle closes at 24,985. The fractal low is 24,920. The trend filter, a 20-period EMA sitting at 24,930 and sloping up, supports going long. Your rule: enter long on a break above the fractal-forming swing high at 25,010, stop just below the fractal low at 24,900, target the prior up fractal at 25,160.

• Instrument: Nifty weekly options, buying 1 lot of the 25,000 CE (call). Lot size is 65.
• Entry trigger: Nifty trades through 25,010 (above the fractal swing high). At that moment the 25,000 CE is priced around Rs 95.
• Stop logic on the index: a close below the fractal low 24,900 invalidates the setup. On the option, that roughly maps to the premium falling to about Rs 60.
• Exit: Nifty pushes to the prior up fractal near 25,160 and you book the 25,000 CE at about Rs 175.

Now the rupee maths on 1 lot of 65. Buy premium Rs 95, sell premium Rs 175, so the gross gain is (175 minus 95) times 65, which is Rs 5,200. From this you subtract real costs. A typical discount broker charges a flat Rs 20 per order, so Rs 40 round trip. STT on options sold is 0.15 percent of the premium (since 1 April 2026): 0.15 percent of (175 times 65) equals 0.15 percent of Rs 11,375, which is about Rs 17. Exchange transaction charges, SEBI fee, stamp duty and 18 percent GST on (brokerage plus exchange charges) add roughly Rs 15 more. So total costs land near Rs 72, leaving a net profit close to Rs 5,128 on this single illustrative trade.

The same example with the trade going wrong is just as important. If Nifty reverses and the 25,000 CE is sold at the Rs 60 stop, the gross loss is (95 minus 60) times 65, which is Rs 2,275, plus about Rs 58 of costs, so roughly Rs 2,333 lost. Notice the structure: risking about Rs 2,333 to make about Rs 5,128 is a reward-to-risk near 2.2 to 1. That ratio, not the indicator alone, is what makes a fractal strategy survivable over many trades.

A Bank Nifty Example With Correct Lot Size

Bank Nifty moves faster, so fractals fire more often and stops must be wider. Suppose on a 15-minute chart Bank Nifty prints an up fractal (bearish) with the middle candle topping at 56,400, while the 20-EMA at 56,300 is sloping down. This is a sell setup. You go short by buying 1 lot of the 56,400 PE (put). The current Bank Nifty lot size is 30.

• Entry: Bank Nifty breaks below the fractal swing low at 56,150; the 56,400 PE is priced around Rs 210.
• Stop: a move back above the up-fractal high 56,400 invalidates the idea; the put roughly falls to Rs 150.
• Target: the prior down fractal near 55,600, where you sell the 56,400 PE at about Rs 470.

Gross gain is (470 minus 210) times 30, which is Rs 7,800. STT on the sell side is 0.15 percent of (470 times 30), about Rs 21, and total costs including flat brokerage and GST come to roughly Rs 77, so the net is near Rs 7,723. If stopped instead, the loss is (210 minus 150) times 30 plus costs, about Rs 1,859. Again the maths is the discipline: a small defined loss against a larger planned gain. These are illustrative numbers, and real fills differ with liquidity and slippage.

Using Fractals for Stop-Loss Placement, the Most Reliable Edge

The single most dependable use of Williams Fractals in Indian trading is not entries at all. It is trailing stops. In an uptrend, each new confirmed down fractal marks a fresh higher low. Move your stop up to just below each new down fractal as price climbs. In a downtrend, trail your stop down to just above each new up fractal. This turns a subjective gut decision into a mechanical, chart-defined rule.

For a swing trader holding Reliance in the cash segment, this is powerful. Say you bought 100 shares of Reliance at Rs 1,420 after a breakout. As the stock trends to Rs 1,520, three new down fractals print at 1,440, 1,470 and 1,495. You trail your stop to just under each one. If price finally cracks the 1,495 fractal, you exit with most of the move locked in rather than giving it all back. Remember that if you sell within 12 months, the gain is short-term and taxed at 20 percent; held longer, it is long-term and taxed at 12.5 percent on gains above Rs 1.25 lakh in the year.

Fractals vs Swing Highs, Swing Lows and Pivot Points

A Williams fractal is a rule-based swing high or swing low. A swing high is any peak between lower prices, and a fractal is the strict version: the high must beat the two bars on each side. In practice they often mark the same turns, but the fractal rule removes judgment and ignores smaller wiggles.

The main difference from pivot points is that pivot levels are known in advance from yesterday's prices, while fractals only appear after price has turned.

Combining Fractals With the Alligator and Moving Averages

Raw fractals on a sideways Nifty chart are noise. The fix Bill Williams himself recommended is to filter every fractal through a trend tool. His own choice was the Alligator, three smoothed moving averages (jaw, teeth and lips) offset forward. A down fractal that forms while price is above the Alligator lips is a higher-quality long; an up fractal forming while price is below the lips is a higher-quality short. Fractals that form inside the tangled Alligator are ignored.

If you do not use the Alligator, a single 20 or 50 period EMA works well on Indian indices. The rule becomes simple: only act on down fractals when price is above the EMA and the EMA is rising, and only act on up fractals when price is below the EMA and the EMA is falling. Adding the RSI as a second filter, for example requiring RSI above 50 for longs, cuts down the false signals that plague Bank Nifty during choppy, news-driven sessions.

Best Timeframes and Settings for Indian Markets

Timeframe choice should match your style and the instrument. Index option buyers who scalp intraday tend to use 5-minute or 15-minute fractals, accepting more signals and more noise. Intraday Bank Nifty traders often prefer 15-minute because 5-minute fractals on such a fast index produce too many false breaks. Positional swing traders on stocks like HDFC Bank, Infosys or TCS use daily fractals, which capture meaningful multi-day swings and are far cleaner.

• 5-minute chart: aggressive intraday scalps on Nifty and Bank Nifty options, high noise, strict trend filter required.
• 15-minute chart: the common intraday default for Indian index F&O, a sensible balance of frequency and reliability.
• 1-hour chart: useful for short-swing positional option trades spanning the weekly expiry.
• Daily chart: best for cash-segment swing trades on liquid NSE stocks and for trailing stops over weeks.

Keep expiry mechanics in mind when you trade fractals through options. NSE index options like Nifty now have a single weekly expiry plus monthly expiry, and theta (time decay) accelerates sharply in the last two sessions before expiry. A fractal signal that looks great on Monday can be eaten alive by decay if you hold a same-week option into the Tuesday expiry without the move arriving. For positional fractal trades, buying the monthly or next-week option reduces this decay drag.

Changing the Period: 3, 5 or 7 Bars

The standard fractal compares the middle bar with two bars on each side, which gives the five-bar pattern. Some scripts add a three-bar mode that compares one neighbour on each side, and a wider mode with three on each side for seven bars. The trade-off is simple.

There is no best setting that works for everyone. Test your timeframe on past charts and note which setting would have kept you out of noise without making you late. Write the setting down in your trade notes so you can compare results across weeks.

Limitations and Common Mistakes

Fractals are lagging by construction and they ignore volume entirely. In a strong trend, price can make dozens of small fractals against the move that all fail, because the indicator has no idea whether momentum supports a reversal. This is why a standalone fractal strategy on volatile Indian indices produces a stream of small losses. The lag also means that in fast gap moves, common around RBI policy or budget days, the level you wanted is often already far away by the time the fractal confirms.

• Trading every fractal without a trend filter, which generates constant false signals in range-bound Nifty.
• Entering at the fractal low or high itself instead of waiting for price to break the level, ignoring the two-bar lag.
• Using tight stops on Bank Nifty, which whipsaws traders out before the planned move develops.
• Ignoring transaction costs and STT, which quietly turn small winning scalps into break-even or losing trades.
• Holding same-week options on a slow fractal setup and losing to time decay even when direction is right.

Fractals on Intraday Nifty Charts: Expect Noise

On a 1-minute or 5-minute Nifty chart, price makes many small swings, so fractals appear often and most are minor. On a 15-minute or hourly chart they are fewer and tied to bigger moves. If you trade intraday, mark only the fractals that are also near an obvious level such as the previous day's high or low, and skip the rest.

How to set a fractal alert on TradingView: add the Williams Fractal indicator (or a community script that includes alert conditions) to your chart, open the Create Alert dialog, choose the indicator as the condition, and pick the fractal-up or fractal-down option if the script offers one. Because the arrow is drawn after two bars close, the alert also fires after confirmation, not at the turn. Alerts come from TradingView. OneTradeJournal has no price alerts and no chart data.

Taxes, Costs and SEBI Rules to Remember

How your fractal profits are taxed depends on the segment. Intraday equity and all F&O activity are treated as business income and taxed at your applicable slab rate, with the ability to set off expenses and carry forward losses if you file correctly. Equity delivery is capital gains: short-term (held under 12 months) is taxed at 20 percent, and long-term gains above Rs 1.25 lakh in a financial year are taxed at 12.5 percent. Because F&O is business income, frequent fractal traders should keep clean records and consider a tax audit threshold check with a CA.

On costs, STT on options is 0.15 percent of premium on the sell side, and on futures it is 0.05 percent on the sell side (rates since 1 April 2026). Add exchange transaction charges, SEBI turnover fee, GST at 18 percent on brokerage plus exchange charges, and stamp duty on the buy side. SEBI also requires that you trade only through registered intermediaries and never act on insider information or attempt to manipulate prices; a technical tool like fractals is purely analytical and does not exempt you from these rules. Always confirm current contract specifications and tax rules on the official NSE and SEBI sources before trading, since rates and lot sizes are revised periodically.

Sources and Further Reading

For authoritative data and further reading on this topic, refer to Zerodha Varsity, Investopedia, SEBI (Securities and Exchange Board of India) and NSE Indices (Nifty Indices). Always confirm current rules, rates and contract specifications on the official source before you trade.