BankNifty Monthly Condor: The Exit That Actually Worked

Chronologically this was the first trade in the journal — deployed 03-07, before the Sensex condor and before the Nifty spread. I am writing it up last because it is the only one of the three that behaved, and I did not want to lead with the flattering one.

It got taken out on 08-07 like everything else. It got taken out for ₹1,200, while BankNifty carried on falling another thousand points without me.

That difference is not luck. It is the exit rule, and this post is mostly about why this system's exit works and the Sensex condor's did not.

The second cycle closed this morning, 20-07, for ₹7,200. That part is at the bottom, and it did not close on the signal.

The system

The signal does not come from BankNifty's price chart. It comes from a chart of the combined premium of the month's ATM straddle.

1. First trading day of the month, after the close: take BankNifty's closing price and round it to the ATM strike. July: BankNifty closed at 58,033 on 01-07, so the ATM for the month is 58,000.
2. Build the combined-premium chart: ATM CE + ATM PE of the monthly expiry, added together, plotted as a single instrument. For July that is 58,000 CE + 58,000 PE. This chart is the system. BankNifty's own candles are decoration.
3. Every day from the second onward, at 3:15 pm: check where the combined-premium candle has closed.
4. Entry: if it closes below the previous day's close, deploy an iron condor — sell the 0.4 delta CE and PE, buy the 0.2 delta CE and PE as wings.
5. Target: 50% of the max profit shown on the payoff chart at deployment.
6. Exit: on any day the combined-premium candle closes above the previous day's high.
7. Repeat: back to step 3. If the entry condition fires again, deploy again. The month can hold several cycles.

Why the combined premium

The straddle price is what the market is charging for movement. Strip out direction and you are left with a single number that says how much the market thinks is about to happen.

Falling combined premium means the market is paying less for movement than it did yesterday. That is exactly the weather a short condor wants — decay winning, volatility contracting, nobody bidding for protection. So the entry condition is not a view on BankNifty. It is a reading of what BankNifty options are charging.

The exit is the same logic in reverse. The combined premium taking out the previous day's high means the market has abruptly re-priced movement upward. Something has changed, the thing I sold is now worth more than it was, and I do not need to know why. Get out.

And note what is missing

There is no rupee stop-loss on this system. The exit is a signal, not a number. I want to flag that now, because it turns out to be the entire reason this trade cost me ₹1,200 instead of what the others cost me.

Something else is missing too, and I did not notice it until this morning. There is no rule anywhere in those seven steps about what happens as expiry approaches. Hold that thought.

Cycle 1: 03-07 → 08-07

The combined-premium candle closed below the previous day's close on 03-07. Condor went on.

• Put spread credit: 632.40 − 263.20 = 369.20
• Call spread credit: 598.00 − 214.05 = 383.95
• Total credit: 753.15 points → ₹22,594.50 (lot size 30)

Breakevens at 57,046.85 and 59,753.15. Structural worst case: 646.85 points → ₹19,405.50 if the put wing is fully breached, 546.85 points → ₹16,405.50 on the call side.

Worth noting the wings came out asymmetric — 1,400 points wide on the put side, 1,300 on the call. I did not choose that. It falls out of selecting strikes by delta on a skewed chain: puts are richer, so the 0.2 delta put sits further away than the 0.2 delta call. The structure is lopsided by construction, and my downside wing is the wider and therefore the more expensive one to have breached. Filing that away.

The exit

08-07. Trump starts rambling, the market takes it seriously, and around 2 pm the combined premium is already above the previous day's high. Exit condition, unambiguous.

Closed the whole structure for a loss of about ₹1,200.

Put that in context. ₹1,200 across 30 quantity is 40 points of widening on a structure that had collected 753 points. The condor gave back roughly 5% of the credit and 6% of its worst case. On a day that put a thousand-point hole in the index, I paid forty points to leave.

BankNifty kept going. It ended the session down near 56,800 — through my short put, past my lower breakeven of 57,046.85, a thousand points below where I had been sitting. Had I still been in that condor at the close, the put wing would have been deep in the money and this would not be a ₹1,200 post. The structural worst case was ₹19,405.50.

I want to be precise about what saved me, because it was not judgement:

The exit rule fires on volatility, not on price. The combined premium spikes the moment the market starts re-pricing risk — which is before the index has finished travelling. A price-based stop, or a rupee-based stop, only triggers after the damage is already on my P&L. This one triggered while the damage was still mostly in the future.

Compare with the Sensex condor the same week: a ₹1,000 rupee stop that I had to execute manually, that I could not execute manually, and that ended up costing ₹2,400. Same shock, same day, same trader. One system got me out early because the signal was ahead of the move. The other one asked me to be fast, and I wasn't.

The rule I broke anyway

The rule says check the combined-premium candle at 3:15 pm. I acted at around 2 pm, on a candle that had not closed yet.

It worked. It worked beautifully — waiting until 3:15 would have cost me considerably more.

Which is precisely why I am writing it down as a break rather than a clever adaptation. A violated rule that pays is the most dangerous outcome available, because it teaches me that my instinct outranks the system, and it will teach me that again next time on a day when my instinct is wrong. The honest position: either the rule becomes "exit intraday when the combined premium takes out the previous day's high" — tested, written down, applied every time — or I wait for 3:15. Right now I have neither. I have a rule I ignored and a profit-and-loss statement that rewarded me for ignoring it.

I do not get to keep both.

Cycle 2: 10-07 → 20-07

BankNifty closed the 8th sitting on 56,800, which is a level it has bounced from before. I thought it would bounce.

It did — on the 9th, and again on the 10th. Combined premium came back in as the panic drained out, and on 10-07 at 3:15 pm the entry condition fired again: candle closed below the previous day's close. Per step 7, back in.

Same rules — 0.4 delta CE and PE sold, 0.2 delta CE and PE bought as wings. Max profit on the payoff chart at deployment came out around ₹16,000, which per step 5 puts the target at roughly ₹8,000.

And here is the part I wrote down while it was still costing me nothing to admit: my support-level read had no business in this decision. "BankNifty is at support, it will bounce" is not in the system. The system does not know what a support level is. The re-entry was triggered by the combined premium closing below the previous day's close — it would have fired whether I had a view or not, and I would have taken it whether I had a view or not.

The danger is that the view and the rule agreed this time. That makes it very easy to believe next month that the view is doing some of the work. It isn't. If they ever disagree, the rule wins, and I would rather have written that down before the month where it hurts.

The close: 20-07, 09:30 am, ₹7,200

Ten sessions of decay. The combined premium never took out a previous day's high after the 10th — the exit signal never fired, and on the signal alone I would still be holding this.

I closed it anyway, this morning at 09:30 am, for a profit of ₹7,200.

Two things forced the timing, and only one of them is in the system.

The one that is in the system. Target is 50% of max profit, which here is ₹8,000 against a ₹16,000 payoff. At the open I was sitting on ₹7,200 — 45% of max, ₹800 short of target. Across 30 quantity that is 240 points captured out of a structural 533. Close enough to target that taking it is not a violation of step 5 so much as a rounding of it.

The one that is not. Expiry is tomorrow. Under the SEBI framework my broker starts ratcheting up margin as monthly expiry approaches — in practice from about a week out, tightening hard in the final sessions. The capital blocked against a structure that had already given me 90% of what I came for was no longer capital I was willing to leave parked there, and carrying it into the last day would have meant funding peak margin for the thinnest remaining slice of decay.

So: a good exit, at a sensible number, for a reason the system has never heard of.

The gap this exposes

Cycle 1 taught me that a violated rule that pays is dangerous. This is the same lesson wearing a suit.

Nothing in the seven steps says anything about expiry week. The system was written as though a position can be held indefinitely until the combined premium takes out a previous high, and that is simply not true of a monthly structure on a broker that reprices margin as expiry approaches. The constraint was always there. I just never wrote it into the rules, so this morning I improvised — and improvising happened to line up with the target, which is exactly the kind of luck that stops you fixing anything.

The fix is not complicated, which makes it worse that it is not already done. Something like: the position is closed no later than the last session before expiry-week margin escalation, regardless of signal, and the target is evaluated against that horizon rather than against an open-ended one. Written down, applied every month, not decided at 9:30 am with a margin notification on the screen.

I also want the honest version of the P&L attribution on record: ₹7,200 came from ten sessions of the combined premium doing nothing dramatic. It did not come from me reading the margin cycle well. I reacted to the margin cycle. There is a difference, and next month it will matter.

The uncomfortable part

On 08-07, Trump's ramblings took out three positions:

Three instruments. Three methodologies. One index would have been a coincidence. Three is a factor.

Every one of these was short premium, and short premium is short volatility, and volatility does not care that I had drawn my positions on different charts. When it repriced, it repriced all of them at once. I wrote in the opening post that this would happen and then had no mechanism to stop it happening, which is the least useful kind of foresight there is.

Cycle 2 does not undo that. It happens to net the journal back to roughly flat, which is a pleasant arithmetic accident and not evidence that the correlation problem has been addressed. It has not been addressed. It has been outrun by one calm fortnight.

Where the journal stands

Four trades, no positions open, and a net of two hundred rupees. Three weeks of work to end up almost exactly where I started — which is, I think, the most accurate description of what a first month of running systems properly looks like.

What I am watching

1. The 2 pm question. Still open. I need to decide, in the cold, whether the intraday exit is a rule or a rationalisation — and then write it down either way. It has now been unresolved for two weeks, which is its own answer about how seriously I am treating it.
2. The expiry-week margin rule. New, and the more urgent of the two, because it will recur every single month on a fixed schedule rather than on a Trump-shaped one. Write the horizon into the system before August's first trading day.
3. Whether the target survives a shorter horizon. If the position must be closed before expiry-week margin bites, then 50% of max profit is a target with a deadline attached. Some months the decay will not get there in time. I want to know what the system does then — take what is there, or stand aside — decided in advance rather than at 9:30 am.
4. Aggregate short-vol exposure. Not per-system P&L. That view is exactly what let ₹7,000 walk out of the door in a single afternoon while every individual position looked survivable on its own. A profitable fortnight has not fixed it.

Disclaimer

This is a personal trading journal. It is a record of my own trades, my own money and my own mistakes — nothing more.

Nothing here is a trade recommendation, a tip, a call, or advice of any kind. I am not a registered adviser and I am not qualified to tell anyone what to do with their capital. The strikes, premiums, entries and exits above are what I did, not what you should do.

If you read this blog and place a trade because of it, that trade is yours. I accept no responsibility for anyone else's losses. Do your own research, size for your own risk, and understand that options can lose you more, faster, than you expect.