Option selling in the last 30 minutes of expiry: what the short ATM straddle actually did, 21 expiries measured

Every gamma blast clip has a counterparty: the writer who sold that option at 15:00 to collect the last of its time value. This page is the last 30 minutes of expiry from the seller's side, measured the same way as the buyer's. For 21 NIFTY and SENSEX expiries we took the at-the-money straddle at 15:00, the call plus the put at the strike nearest the index, and followed what a writer short that pair would have owed at every ten-second sample to 15:30. The numbers are dated; the table is complete, including the days it went badly.

The writer's bet in the last 30 minutes

At 15:00 on expiry day an at-the-money option is almost all time value, and all of it must be gone by 15:30. That is the seller's case: sit still for half an hour and collect it. What the case leaves out is the settlement price. The exchange computes the index close as the weighted average of the last 30 minutes, so the writer is short an option on a number that is still being formed, and short it at the moment its gamma is highest. A twenty-point drift in the walk moves the at-the-money leg by more than the whole premium collected. And a writer who is on the wrong side with size faces the choice that produces the blast itself: close now at whatever the market asks, or be assigned at settlement.

What the short straddle actually did, 21 expiries

Read as a coin with a heavy tail. Half the time the writer kept most of what was collected, and on eight days kept more than half of it. The other half of the time the pair cost more at the end than it fetched at 15:00, and on five days it was worth double at some point inside the window. The median end value was 1.1 times the 15:00 value, so the middle outcome was a small loss, not a small gain. The SENSEX split is the striking one: seven of ten Thursday expiries finished with the writer paying more than collected. SENSEX strikes are 100 points apart and its index level is three times NIFTY's, so "at the money" on SENSEX is a wider bet on the same half hour; the sample is ten days and the reading is an observation, not a rule.

Every expiry, straddle at 15:00, its worst moment, and the end

Each expiry links to its recap with the chart. A high at "15:00" means the pair never traded above its opening value in the window.

Pin risk: the part the premium does not price

Look at 8 September 2026. The 23,650 straddle was ₹75 at 15:00 and ₹15 at the end; the writer kept 80 percent. Inside that same window the put alone went from ₹33 to ₹91 at 15:21, so a writer short only the put, or short the pair but hedging the put leg, sat through a moment where the leg was worth 2.7 times the entry. The straddle table shows the end, not the ride. On 27 August the SENSEX pair went from ₹178 to ₹453 at 15:27 and closed at ₹266; the writer paid 1.5 times what was collected, having been down 2.5 times a few minutes earlier. The mechanism both days was the same: the strike with the largest open interest on one side sat within a few points of the settle walk, on the wrong side for its writers, and the writers closed. Gamma Terrain shows that configuration while it forms: the pin, the wall, and the resting depth on each side of the pin, which on 8 September thinned to a quarter of its baseline three and a half minutes before the put spike. The 21-expiry study has the buyer's side of these same days.

What this page does not say

It does not say whether to write or to buy. Half the days rewarded the writer and half charged them, and the charged days were larger than the rewarded ones. It does not include what a writer would have done between 15:00 and 15:30, and most writers do something. And the ten-second chain samples have holes on 16 of these 21 days where the exchange rate-limited the poll, so a pair's true high may be higher than shown, never lower. The expiry recaps add every new expiry to this record the same evening.