Fisher Transform: Catching Nifty Turning Points
How to read the Fisher Transform on Nifty and Bank Nifty, with a real turning point example, options trade math, settings and Indian F&O tax.
Key Takeaways
- 1.The Fisher Transform sharpens turning points by stretching price into a near Gaussian shape, so a real reversal shows up as a sharp spike that bends back, not a slow drift.
- 2.It uses a two stage recursion. You smooth the price position inside its recent range, then apply the log based Fisher formula. The previous bar feeds the next, so it has memory.
- 3.The signal that actually matters is the Fisher line crossing its own one bar lagged trigger line near an extreme, not a lazy zero line cross.
- 4.A worked Nifty example below uses real index levels around a January 2025 swing low and turns the signal into a Bank Nifty options trade with rupee profit, lot size, STT and brokerage shown.
- 5.Treat it as a timing filter on top of trend and support work. Used alone on choppy Nifty days it whipsaws, and in India your F&O gains are taxed as business income, not capital gains.
What The Fisher Transform Actually Does
The Fisher Transform was built by John Ehlers to fix a real problem. Raw price does not follow a bell curve. It clusters and then jumps, so ordinary oscillators give mushy, late signals at turns. The Fisher Transform mathematically squeezes the price data toward a Gaussian normal distribution, which is a fancy way of saying it pushes ordinary readings toward the middle and stretches extreme readings far out to the edges. The practical effect is that a genuine top or bottom produces a tall, narrow spike that then snaps back, and that snap back is your signal.
This matters for Indian index traders because Nifty and Bank Nifty spend long stretches grinding in a range and then turn hard, often around the open or around RBI policy, US CPI prints and monthly expiry. A lagging moving average tells you about the turn a day late. The Fisher Transform is designed to flag the exhaustion right as it happens. The trade off is sensitivity. The same sharpness that catches a real bottom will also fire on a fake one, which is why the rest of this page is about reading it correctly rather than blindly buying every spike.
The Real Formula, Step By Step
The earlier version of this page showed a single normalize and transform step. The real indicator has two pieces of memory that most simplified explanations drop, and dropping them is exactly why a naive example produces a meaningless zero. Here is the full sequence for a chosen lookback, commonly 9 or 10 bars.
- Find the highest high and lowest low over the lookback period. On a 9 bar setting you scan the last 9 bars.
- Compute the raw position of the typical price inside that range, scaled to run from minus 1 to plus 1. Typical price is usually (high plus low) divided by 2.
- Smooth that position. Value now equals 0.33 times 2 times (raw position) plus 0.67 times the previous smoothed Value. This carry over of the previous Value is the first piece of memory.
- Clamp the smoothed Value so it never hits exactly plus or minus 1, otherwise the next step blows up to infinity. Ehlers caps it at about plus or minus 0.999.
- Apply the Fisher formula. Fisher now equals 0.5 times the natural log of ((1 plus Value) divided by (1 minus Value)) plus 0.5 times the previous Fisher. That final carry over is the second piece of memory and is what gives the line its smooth follow through.
The trigger line is simply the Fisher value from one bar ago. You do not plot a separate moving average. You plot Fisher and its own previous reading, and the cross between them is the event you trade. This is the part the old example skipped, and without it the indicator looks like it always returns zero at the middle of a range, which is both true and useless. The signal lives at the edges, not the middle.
If price sits exactly at the centre of its range, the normalized position is 0 and the Fisher value is 0. That is mathematically correct and practically worthless, because the Fisher Transform is never used to read the middle of a range. Its entire job is to flag the extremes. A turning point example must use a price near a recent high or low, which is exactly what the worked example below does.
A Real Nifty Turning Point, Calculated
Take Nifty 50 near its swing low in late January 2025, when the index sold off toward the 22,800 area before turning back up. The numbers below are illustrative and rounded to keep the arithmetic readable, but they sit on realistic index levels for that period. Suppose over the last 9 daily bars the highest high was 23,500 and the lowest low was 22,750. On the day price is hammering the lows, the typical price prints around 22,820.
Raw position equals 2 times ((22,820 minus 22,750) divided by (23,500 minus 22,750)) minus 1. That inner fraction is 70 divided by 750, which is 0.0933. Times 2 is 0.1866, minus 1 gives a raw position of minus 0.8134. Price is pinned near the bottom of its range, so the raw position is deeply negative, which is the whole point. Now smooth it. Assume the previous smoothed Value was minus 0.55. Value equals 0.33 times minus 0.8134 plus 0.67 times minus 0.55, which is minus 0.2684 plus minus 0.3685, giving a smoothed Value of minus 0.6369.
Apply the Fisher formula. 0.5 times the natural log of ((1 plus minus 0.6369) divided by (1 minus minus 0.6369)) equals 0.5 times ln(0.3631 divided by 1.6369) equals 0.5 times ln(0.2218) equals 0.5 times minus 1.506, which is minus 0.753. Add half of the previous Fisher, and assume that was minus 0.95, so plus 0.5 times minus 0.95 is minus 0.475. Today's Fisher is roughly minus 1.23. That is a deeply negative, stretched reading, the kind of extreme the transform is built to expose. The very next bar, as Nifty ticks up off the low, the raw position climbs, the new Fisher rises to about minus 0.70, and it crosses back above its own prior value of minus 1.23. That upward cross out of an extreme low is the long signal.
| Bar | Typical price | Raw position | Smoothed Value | Fisher | Read |
|---|---|---|---|---|---|
| Bar 1 | 23,050 | minus 0.47 | minus 0.40 | minus 0.55 | Drifting lower |
| Bar 2 | 22,900 | minus 0.71 | minus 0.55 | minus 0.95 | Stretching down |
| Bar 3 (low) | 22,820 | minus 0.81 | minus 0.64 | minus 1.23 | Extreme, exhaustion |
| Bar 4 (turn) | 22,980 | minus 0.62 | minus 0.63 | minus 0.70 | Crosses up, long signal |
| Bar 5 | 23,180 | minus 0.10 | minus 0.45 | minus 0.10 | Trend confirms |
Notice what the table shows. The signal is not the Fisher line touching zero. It is the Fisher line spiking to an extreme near minus 1.2 and then turning up and crossing its lagged value at Bar 4. By the time it actually reaches zero at Bar 5, Nifty has already moved 360 points off the low and most of the easy entry is gone. Trading the extreme and the cross, not the zero line, is the difference between catching the turn and chasing it.
Turning The Signal Into A Real Trade With Rupees
A signal you cannot size is just a chart decoration. Suppose the Bar 4 cross on Nifty fires and you express the bullish view through Bank Nifty options, because it moves harder than Nifty on these reversals. Say Bank Nifty is around 48,500 and you buy one lot of the monthly 48,500 call. Bank Nifty lot size is 30. Assume the call premium is 250 when you enter.
- Entry cost of premium equals 250 times 15, which is Rs 3,750 of capital at risk on this single lot. That premium is the most you can lose on a bought call.
- The reversal plays out and Bank Nifty rallies 600 points over two sessions. The call premium expands to about 720.
- Exit value equals 720 times 15, which is Rs 10,800. Gross profit before costs is Rs 10,800 minus Rs 3,750, which is Rs 7,050.
- STT on options is charged at 0.1 percent on the sell side premium, so 0.1 percent of Rs 10,800 is about Rs 11. Brokerage at a flat Rs 20 per order on entry and exit is Rs 40. Add exchange charges, GST and stamp duty and round total costs to roughly Rs 90 for this round trip.
- Net profit is approximately Rs 7,050 minus Rs 90, which is about Rs 6,960 on one lot. These figures are illustrative, not a promise of any return.
Bought options decay every day. A correct Fisher signal that takes three sessions to pay off can still lose money to theta if you size for a slow grind. On reversal trades, prefer slightly in the money calls or a debit spread so time decay does not eat the move you correctly predicted.
Best Settings For Nifty And Bank Nifty
The default lookback is 9 or 10 bars and it is a sensible starting point, but the right number depends on your timeframe and instrument. Bank Nifty is faster and noisier than Nifty, so a slightly longer lookback often filters out the chop. The table below is a practical starting grid for Indian index trading, to be confirmed by your own backtest rather than taken as gospel.
| Style and timeframe | Instrument | Suggested lookback | What you are trading |
|---|---|---|---|
| Swing, daily chart | Nifty 50 | 9 to 10 bars | Multi day reversals at support and resistance |
| Swing, daily chart | Bank Nifty | 10 to 13 bars | Same, with extra smoothing for the noise |
| Intraday, 15 minute | Nifty 50 | 9 bars | Morning reversal and afternoon trend turns |
| Intraday, 5 minute | Bank Nifty | 13 to 21 bars | Scalps, longer lookback to cut false fires |
On a 5 minute Bank Nifty chart the default 9 will fire constantly because every two candle wiggle looks like an extreme. Pushing the lookback to 13 or 21 makes the indicator wait for a genuine range extreme before it spikes. The cost is later signals. There is no free lunch here, only a dial between fewer good signals and many noisy ones.
How To Read The Signal Without Getting Trapped
The cleanest way to read the Fisher Transform is to ignore the zero line as an entry trigger and watch two things instead. First, an extreme reading, generally beyond plus 1.5 or below minus 1.5, telling you price is stretched and a reversal is likely. Second, the cross of the Fisher line against its one bar lagged trigger while it is at that extreme. A down cross from above plus 1.5 is a short setup. An up cross from below minus 1.5 is a long setup, which is exactly the Nifty case above.
- Long setup: Fisher is below minus 1.5, then turns and crosses above its previous value. Best taken at known support or a prior swing low.
- Short setup: Fisher is above plus 1.5, then turns and crosses below its previous value. Best taken at known resistance or a prior swing high.
- Skip the trade when Fisher is hovering near zero. The middle of the range is where the indicator has nothing to say and where whipsaws are worst.
- Divergence is the high quality version. Price makes a lower low but Fisher makes a higher low. That mismatch flagged at an extreme is the strongest reversal tell.
Combining It With Other Tools
The Fisher Transform is a timing tool, not a trend tool. It tells you when a move is stretched, not whether you are with or against the bigger trend. Pairing it with a directional filter is what turns it from a whipsaw machine into a usable edge. The table below shows sensible pairings and the specific job each partner does.
| Pair with | Job it does | Practical rule on Nifty |
|---|---|---|
| 200 EMA on daily | Defines the dominant trend | Only take Fisher long signals when price is above the 200 EMA |
| RSI | Confirms overbought and oversold | Fisher up cross plus RSI below 30 is a stronger bottom |
| VWAP intraday | Shows the day's fair value | Fade Fisher extremes back toward VWAP, not away from it |
| Volume or delivery | Confirms conviction | A reversal spike on rising volume is more trustworthy |
A concrete workflow for a Nifty swing trader looks like this. Check that price is above the 200 day EMA so you are trading with the trend. Wait for the Fisher Transform to dip below minus 1.5 near a support level. Confirm RSI is also oversold. Take the long on the Fisher up cross, place the stop below the swing low that produced the extreme, and target the prior resistance. Each tool removes a class of false signal that the Fisher Transform alone would have walked you into.
Limitations And When It Fails
The biggest failure mode is a strong trend. In a powerful Nifty uptrend the Fisher Transform will pin near the top and keep flashing short signals that get run over for days. The indicator is built to catch reversals, so in a market that simply does not reverse, it is wrong repeatedly. This is why the trend filter is not optional. A second failure mode is gap risk. Indian indices gap on global cues, and a gap can jump straight past your Fisher signal and stop before you ever act on it.
- It fights trends. In a one way move it will short every push and lose.
- It is noisy on very short timeframes unless you lengthen the lookback.
- It cannot see news. An RBI surprise or a US jobs shock overrides any oscillator.
- It needs a clean range to measure. On expiry day pin action the high low range gets distorted and so do its readings.
Tax And Cost Reality For Indian Traders
Signals are only half the story. What you keep depends on costs and tax. In India, F&O trading is treated as business income, not capital gains. So the Bank Nifty options profit in the worked example is added to your other business and salary income and taxed at your slab rate, with no special concessional rate and no benefit from the STCG or LTCG regime. There is no separate lower rate for futures and options. If you also trade in the cash segment and hold positions, short term capital gains on listed equity are taxed at 20 percent and long term capital gains above Rs 1.25 lakh in a year are taxed at 12.5 percent, but those rates do not apply to your F&O book.
Costs compound fast on an active reversal strategy. Each option round trip carries STT on the sell side, brokerage, exchange transaction charges, GST on those charges, SEBI turnover fees and stamp duty. On a single Bank Nifty lot these came to roughly Rs 90 in the example, which is small against a Rs 7,000 winner but large against a Rs 500 scalp. If the Fisher Transform pushes you toward many small intraday trades, your edge has to clear the combined cost stack on every single one. Always confirm the current STT rate, lot size and contract specifications on the NSE site before you trade, because these change.
Log the lookback you used, the index level at the extreme, whether you waited for the cross, and the rupee result after costs. Over 30 trades you will see plainly whether your settings catch real turns or just whipsaw, and that record is worth more than any backtest someone else ran.
Sources And Further Reading
For contract specifications, lot sizes and current charges, check NSE India. For plain language explainers on indicators and taxation of trading income, see Zerodha Varsity and Investopedia. You can also pair this with the RSI guide and the broader indicators library. Always confirm current rules, rates and contract specifications on the official source before you trade.
Sources and Further Reading
For authoritative data and further reading on this topic, refer to Zerodha Varsity, Investopedia and NSE India. Always confirm current rules, rates and contract specifications on the official source before you trade.
Related Topics
Related Articles
Understanding the Vertical Horizontal Filter in Indian Markets
How the VHF spots trending vs ranging Nifty, with real NSE regimes, a worked calculation, option tactics, and Indian F&O tax rules.
RSI 2 Period Strategy for Indian Markets
The RSI(2) mean reversion strategy for Indian markets: 200 DMA filter, exact entry and exit rules, and worked Reliance and Bank Nifty rupee examples.
Understanding Trading Psychology in Indian Markets
Learn trading psychology for Indian markets with a worked Nifty options example showing how fear and greed turned a Rs 3,600 loss into Rs 16,500.
Understanding Short Selling in Indian Markets
How short selling works in India: the intraday-only retail rule, SEBI SLB overnight borrowing with a real Reliance borrow-cost example, F&O shorts and tax.
How to Rebalance Your Portfolio in Indian Markets
How to rebalance your Indian portfolio with the correct post-2024 tax: 20% STCG, 12.5% LTCG above Rs 1.25 lakh, plus a worked Nifty example.
Pair Trading Strategy for Indian Markets
Pair trade TCS and Infosys with real z-score math, lot sizes, rupee P&L, STT and slab-rate tax. A worked, market-neutral guide for Indian traders.
The trading journal built for Indian F&O traders. Track your trades, spot patterns, build discipline.
- Log one trade a day by hand, on purpose
- AI mentor finds your repeat mistakes
- Behavioural analytics catch tilt early
- Trading calendar with P&L heatmap
- Pre-trade checklist flags risks
Yearly ₹2,499 · No broker credentials