The weekly premium strategy was tested and it failed
Measured at the price a seller is actually paid, no horizon from 7 to 60 days clears t = 2. The apparent edge at 0 DTE was a σ error. The gate fired on 73% of 0 DTE observations and 0% at 3+ DTE, so it could only ever open where the model was broken: at 0 DTE it claimed 78.2% safety and delivered 57.9%. Drawdown reached 33.3% of cumulative profit.
What that σ error actually is, measured rather than assumed. This paragraph said the market prices same-day volatility "about a third higher" than India VIX. That was an estimate and it was too large. Across 39 readings on 14 sessions, at-the-money implied volatility runs 7% above India VIX once dead weekend days are taken out of the clock — 12% at the shortest horizons, not 33%. The raw comparison points the other way, and that is the part worth knowing: σ is spot × vol × √(dte/365) with calendar days, while India VIX is a thirty-day number whose per-day variance already averages in a normal weekend. A Friday-to-Tuesday weekly is half dead time, so on Fridays this site's σ is roughly 26% too large, and on Mondays about 12% too small. Every POP, EV, band and gate reading here is quietly weekday-dependent. Selecting σ from this expiry's own chain removes both errors, which is what ?sigma_source=chain_iv on the API does.
Everything below is kept because the measurements are worth reading, not because the signal is worth taking. The gate, the sizing and the Kelly figures on this tab are not a validated trading system. The POP and EV columns on the Chain tab inherit the same VIX-derived σ. Today's paper trade is on the Intraday tab.
Implied volatility by expiry — the curve one VIX flattens
Solved from each expiry's own chain, against the single thirty-day India VIX this site applies to all of them. The horizontal axis is days, not one notch per expiry — on equal notches a curve sampled unevenly slopes differently everywhere between its points.
What each expiry is actually priced at
Position sizing — margin, risk rule and Kelly
Two limits, and the smaller one wins: how many lots the margin can carry, and how many the risk rule allows before a stop-loss would cost more than you said you were willing to lose. Kelly is a third opinion on whether the edge justifies size at all — a negative full Kelly means the maths says do not take it, whatever the margin allows.
Scanner — ranked by EV per rupee of margin
Not ranked on rupees of EV. A bigger index produces bigger credits simply by being bigger, and it consumes proportionally more margin to do it — so the only fair comparison is what each earns per rupee tied up. Only underlyings the collector actually holds appear here.
Payoff at expiry — –
All seven structures — click a row to chart it
Ranked on net EV after the corrected 0.0495% exchange rate and measured slippage. The gate is on credit/σ ≥ 0.9346 — the workbook's POP ≥ 0.65 expressed on one axis. A structure can show positive EV and still not be recommended, because the gate is about whether the premium is worth selling at all, not which shape to sell.
Legs
Click a premium in the chain to sell it, click again to buy it, a third time to drop the leg. Everything recomputes against the same engine the Telegram signal uses — no second implementation to drift.
Payoff — nothing selected
Compare — pin two structures
Build a structure, pin it as A, build another, pin it as B. Each panel is scaled to its own structure by default, which is the clearest way to read one payoff — and a misleading way to compare two, because a smaller structure is rescaled to fill the same box. The axis range is printed under each chart so you can see when they differ, and same scale forces both onto one range when you want the comparison to be honest by construction rather than by inspection.
Chain — click to add legs
Volatility smile — implied, both sides
Black–Scholes at zero drift, solved off the parity forward. Calls and puts at the same strike should sit on top of each other; where they separate, one side's quote is stale.
ATM straddle through the session
The price of the at-the-money straddle, minute by minute, from your own bars. This is what you are selling — when it falls through the day with the index still, that is theta arriving.
Chain
The same column you are reading in the table, drawn across strikes — calls solid, puts dashed, as on the smile. A table tells you the value at a strike; a curve tells you where the maximum is, which is the question you are usually actually asking.
Priced at the mid, never at close: a far strike that
did not trade in the last minute carries a stale print. Quotes that are one-sided
or absurdly wide are dropped and counted above rather than guessed at. Bars behind
the rows are open interest, scaled to the largest on screen.
VWAP is volume-weighted for the whole session, from the
pv the collector accumulates per bar — not an average of the minute
closes, which would weight a minute that traded one lot the same as one that
traded a thousand. A mid far below VWAP means the option has been sold down
since most of its volume traded.
B/U reads price and open interest together, measured from this session's first bar. LB long buildup, price up and OI up — new buyers. SB short buildup, price down and OI up — new sellers. LU long unwinding, both down — holders closing. SC short covering, price up and OI down — writers buying back, which on a strike you have sold is the one worth noticing. Moves under 0.5% are shown as flat rather than dressed up as a signal.
Open interest by strike
Calls above the line, puts below. The tallest call bar is where the most writers have bet the index will not go — which is why it so often behaves like resistance. The tallest put bar is the mirror. Neither is a forecast; both are where the pain is concentrated.
Max pain — total loss to option writers
At each possible settlement price, what every open call and put would cost the people who wrote them. The minimum is max pain. It is a statistic about crowd positioning, not a prediction — but expiries do tend to gravitate toward it, which is why the distance is worth watching.
Per-strike buildup — ranked by OI change
Price and open interest moving together, per contract, from the session's first bar to the latest. Rising OI means new positions; falling OI means positions closed. The pairing names who was doing it. Read the agreement panel below before trusting the pattern — ranking by OI change sorts near-money strikes to the top, and that is where the index move drowns everything else.
Out-of-the-money view — forward move removed
Per-minute structure price
Price, VWAP and moving averages
VWAP is each leg's own session volume-weighted average, summed — the average cost of assembling the structure leg by leg, not a price the combination itself ever traded at. Price above it means the structure costs more now than the session's average buyer paid.
RSI
On the structure's own price, not the index. A short straddle bleeding to theta drifts down all session, so a low reading here is often decay rather than weakness — which is exactly why the anchor matters.
Cumulative net — paper, per lot
Every settled session, in order, after the measured spread and real charges. The shape matters more than the level: a line that climbs gently and then drops a long way is the short-volatility payoff, and it is the reason the median and the mean below disagree.
Where the implied premium goes
Journalled sessions
Calibration — modelled POP against realised
Each dot is one structure's whole record. On the diagonal means the model predicts its own hit rate. Above means pessimistic, below means overconfident — and overconfident is the one that costs money.
Realised P&L per unit, recommended trades
Per unit, so a lot-size change cannot flatter or flatten it.