Method

Nothing here is an opinion. Every figure is computed from recorded market data, and this page says exactly how, so you can judge whether a signal means what you think it means.

Where the data comes from

Option prices are captured tick by tick from the exchange feed during market hours and stored as they arrive. Stock data is one-minute candles for every NSE F&O stock — about 211 names — pulled each minute while the market is open, with a rolling year of history (365 sessions) kept for context.

Live Index Options

• Candles are built from ticks at one-minute resolution; the newest candle updates as ticks arrive.
• Support and resistance are swing highs and lows on the candles currently on screen, clustered together when they sit within a tolerance scaled to the option's own volatility. Only the two strongest levels each side are shown; the count (×3) is how many times the level was touched.
• OI sentiment compares the change in call and put open interest over the last 15 minutes. Puts adding faster than calls means put writers are dominant, which is bullish; the reverse is bearish.
• Move scores reprice each strike with Black-Scholes at its current implied volatility, assuming the index moves a chosen number of points and time decays by the chosen window. Gamma and theta are both included. The "if wrong" column prices the opposite move. Implied volatility is assumed unchanged — in practice it usually falls on rallies and rises on sell-offs, so gains on calls are often flattered.

Market Edge

Gamma per strike. Implied volatility is solved from the last traded price of the out-of-the-money side of each strike by bisection on Black-Scholes, then gamma is evaluated at the current spot. The in-the-money quote is deliberately not used: it is mostly intrinsic value and inverts poorly. A quote sitting below intrinsic is discarded rather than clamped, and the strike inherits the at-the-money volatility.

Gamma exposure = gamma × open interest × lot size × spot² ÷ 100, reported in crore of delta per 1% move. The positioning convention is that the sell side writes both calls and puts and the buy side pays for them, which is how this market trades. A book short both sides is short gamma at every strike, so the two sides add rather than cancel, there is no level at which net gamma changes sign, and no "gamma flip" is published. Concentration is the share of total gamma within 0.5% of spot — raw gamma explodes into expiry, so that ratio is the part comparable from one day to the next.

Writer rotation scores the balance of two flows over a rolling 20 minutes, across strikes within 2% of spot, using open-interest changes only:
bullish flow = calls covered above spot + puts written below spot;
bearish flow = puts covered below spot + calls written above spot;
score = (bullish − bearish) ÷ (bullish + bearish) × 100, bounded to ±100. A rotation is published when |score| ≥ 35 and the flow is at least 2% of near-money open interest, so a quiet chain drifting by a few hundred contracts cannot score 100. Outcomes are measured over the following hour against a 0.25% move in the rotation's direction, with re-triggers inside that hour suppressed so one rotation counts once.

The flat level is where the writers' book carries no net delta, found by recomputing that delta at each of 121 candidate index levels spanning ±6% of spot and interpolating the crossing. Because puts dominate far below and calls far above, the crossing always exists.

Expected move comes from the at-the-money straddle: since a straddle is worth about 0.7979 × spot × volatility × √time, the one standard deviation move to expiry is 1.2533 × the straddle price. No positioning assumption is involved.

The pin is the gamma-weighted centre of open interest across strikes within 1.5% of spot. Pin strength per strike is its total gamma as a percentage of the largest on the chain.

First-hour regime classifies 09:15–10:15 by where the index closed within that hour's range (top 30%, bottom 30%, or neither) crossed with whether the hour was wider or narrower than the median first hour on record. Six buckets, kept coarse so each one holds enough sessions to quote.

Writer capitulation fires when a strike whose peak open interest is at least 55% of the largest within 1.5% of spot loses 12% or more of that peak, while the index sits within 0.35% of the strike and the option's own premium is up 20% or more over 30 minutes. Puts below spot read as support being abandoned, calls above spot as resistance being covered.

Base rates. Every rate shown on that page is recomputed nightly by replaying each recorded session minute by minute through the same code that produces the live numbers, then measuring what followed. Rates are reported with both the number of instances and the number of distinct sessions they came from, because intraday triggers cluster and counting them as independent would flatter them. Where a bucket holds fewer than 25 instances or fewer than 20 sessions, the page prints "not enough history yet" instead of a percentage. The guide on gamma exposure sets out what this can and cannot tell you.

Calibrating the rotation trigger

The score at which a rotation is called is chosen by search, not by judgement. Every combination of score threshold (20 to 70) and minimum flow (0.5% to 5% of near-money open interest) is replayed over every recorded session at five-minute samples. At each signal the at-the-money option on the side the rotation points to is bought at its price that minute and followed for an hour.

The threshold selected is the one where that option most often reached +50% before it fell 25% — ordering matters, so the full minute-by-minute path is walked rather than comparing best and worst case, since a gain that only arrives after a deep drawdown is not a trade anyone holds through. A rotation that stays above the threshold is one event: further samples inside the follow window are suppressed, because counting each five-minute reading of the same rotation separately would inflate the sample twelvefold and make one signal look like a dozen confirmations.

A threshold is only adopted if it produced at least 30 signals across at least 15 sessions; otherwise the untuned default stands, on the grounds that a rule fitted to nine signals is worse than no fitting at all. The live publisher reads the result, so re-running the search changes what the site calls a rotation without any code change. The page shows the chosen threshold, the rate it was chosen on, and how many signals a day it produces.

Session replay

The OI page can replay a whole session. A separate file per index holds the chain sampled every five minutes from the open — open interest per strike per side, plus the index at each frame — built from the live feed's own tick history during market hours and from the recorded session file afterwards. It covers the 40 strikes nearest the money, which is where open interest actually moves; the wings would triple the file for nothing.

Replayed bars mean exactly what live bars mean: with a window selected, the change is measured over that window ending at the frame on screen, not from the open. The verdict and totals are recomputed from the frame being shown, so they always describe the same moment as the bars underneath them.

Bullish Stocks

Each F&O stock scores one point per condition met, out of seven: above today's VWAP; 5-minute EMA9 above EMA21; above the opening-range high (09:15–09:30); higher highs and higher lows over the last 30 minutes; volume at least 1.5× normal for that time of day; beating NIFTY after adjusting for the stock's beta; and within 0.5% of the day's high.

Freshness. Each row also reports whether the score rose in the last 15 minutes. The same number means different things arriving and three hours old, so the two are measured separately and the base rates are published for each.

Risk and invalidation. The level the reading depends on is the higher of today's VWAP and the lowest low of the last 30 minutes — lose either and the score is no longer true. The table shows that level and the distance to it as a percentage, which is the room the setup needs rather than a recommended stop.

Sector breadth is the share of the stock's own sector also scoring 5 or more, so a leader moving with its peers can be told apart from one moving alone.

Base rates. Every session in the intraday store is replayed at half-hourly stamps through the day, scored with the same module that produces the live page, and followed for 30 and 60 minutes and to the close. Results are bucketed by score, by time of day, and by whether the score had just risen; each bucket reports the share that were higher afterwards, the median move, and the median excursion for and against. Excursions are clamped at zero so the for-minus-against figure means what it says. Rates carry both the number of readings and the number of distinct sessions, and a bucket below 200 readings or 20 sessions is shown as "not enough history yet" rather than as a percentage, because a scanner produces hundreds of correlated rows a day and counting those as independent observations would flatter every number on the page.

F&O Support Bounces

Support zones are clusters of daily swing lows, each defended at least twice, first formed at least three months ago, and weighted by how much of the year's volume traded at that price. A stock is testing a zone when it falls into it from at least 3% above without closing more than 2% below. It is bouncing once it is back above the zone and off the day's low. The score combines zone strength, bounce size, the share of minutes that held the zone, volume against normal, and how recent the test is.

Holding Up Best

A low-beta stock always falls less than the index, which is not strength. So each stock is compared with its own history: from a year of 15-minute bars we measure its down-capture (the share of NIFTY's falling moves it normally takes) and up-capture. In the chosen window we compare what actually happened with what that history predicts. The z-score divides the gap by the stock's normal deviation from its own pattern, so a naturally jumpy stock cannot top the list by chance. A reading of 2 or more is a clear deviation. Stocks that barely track NIFTY at all are filtered out by correlation.

Which index is covered each day

The tick recorder follows the expiry calendar rather than running both indices at once, so every session is captured at full depth for the contracts that actually matter that day: SENSEX on Wednesday and Thursday, around its Thursday expiry, and NIFTY on the other weekdays, around its Tuesday expiry. The pages show whichever index is being recorded, so an index switch that offers only one name is expected rather than a fault.

How often things update

Option charts update tick by tick over a live connection. The stock scanners recompute every minute from 09:15 to 15:40 IST. Support zones and strength baselines are rebuilt nightly after the close. Outside market hours every page shows the last completed session and says so.

What this is not

These are measurements, not recommendations. Nothing here accounts for your position size, risk tolerance or costs, and no calculation predicts the future. Option pricing assumes a model that markets do not always follow, particularly around events and expiry.