One trade. Two minutes. Minus ₹19,800.
That is the whole story of Tuesday, 15 September 2026, and we are going to tell it properly — because the days we lose money are the days that actually tell you whether a system has rules or just hope.
The machine maps support and resistance levels on SENSEX before and during the session. It does not trade a level being near. It waits for a confirmed break through it, and only then buys an option in the direction of that break.
At 13:03 IST a downside break confirmed. The machine bought the 74700 PE at ₹442.45, 900 quantity. That is the entry — no prediction, no view on where the market "should" go, just a rule that fired.
The break did not hold. Price came straight back through the level it had just broken. The premium slid, and at 13:05 the fixed stop-loss triggered at ₹420.45. Exactly 22 points below entry. Exit done, machine flat.
| Time | Side | Strike | Entry | Exit | Qty | Result | P&L |
|---|---|---|---|---|---|---|---|
| 13:03 → 13:05 | Buy PE | 74700 | ₹442.45 | ₹420.45 | 900 | Stop-loss hit | −₹19,800 |
Think of a batsman who edges the ball to the keeper and walks off without waiting for the umpire's finger. He does not stand there arguing. He does not hope the umpire missed it. He knows the rule, the rule applied, he goes.
That is what 13:05 was. The level failed, the stop was already sitting 22 points away, and the machine walked. It did not average down, it did not "give it five more minutes", it did not convert a trade into an investment.
A bad day is when a system breaks its own rules. Today the system followed them perfectly and still lost money, which is a completely normal thing for a stop-loss to do. If you buy options on confirmed breaks, a certain share of those breaks will fail. That is the cost of being in the trade at all.
The number we watch is not "did today make money". It is whether a losing trade cost roughly what a losing trade is supposed to cost. Twenty-two points, two minutes, no bleeding into the afternoon. It did.
Eight sessions have printed this month. Five green, three red.
| Date | Day P&L |
|---|---|
| 1 September | −₹19,350 |
| 3 September | +₹38,723 |
| 4 September | +₹3,600 |
| 8 September | +₹11,192 |
| 9 September | +₹25,287 |
| 10 September | −₹26,200 |
| 11 September | +₹5,723 |
| 15 September | −₹19,800 |
That leaves the month at +₹19,175. Note what carries the month: 3 September alone is bigger than the whole month's total. Take that one day out and September is negative. That is the honest shape of a break-trading book — a handful of sessions do the work, and the rest is the cost of showing up for them.
We did not take a second trade to "make it back". After the stop fired at 13:05, the rest of the session had over two hours left in it, and the machine sat out. Revenge is the most expensive trade in retail options, because it is the one trade you take without a reason.
Thursday is SENSEX expiry. We will report that session exactly the way we reported this one, whichever direction it goes.