On Wednesday 10 Jun 2026 the 23,250 Nifty straddle on the weekly expiry cost Rs 346.05 a unit: a call at Rs 163.60 and a put at Rs 182.45, with 6 days to go. Over the next four sessions the index moved −53, 461, 231 and 135 points. One of those four days did all the work. Gamma scalping is the habit of buying that kind of day and paying for the others. I write it down as arithmetic because the arithmetic is simple and the social media version of it is not.
3By the end you can
- Work out the break even daily move of a long gamma position from implied volatility.
- Read a four day table of moves against priced moves and name the day that decided the result.
- Explain why futures lots of 65 units and trading costs limit a small scalper.
The idea, plainly
A gamma scalper owns options, so the position is long gamma (module I09). The scalper sets delta to zero with the future, and resets whenever the index has moved enough to matter. After a rise the delta is positive, so the scalper sells future. After a fall the delta is negative, so the scalper buys back. Sell high, buy low, again and again, in small pieces. Each piece is a little gain locked in.
Day's P&L = Gamma*move*move/2
less Gamma*implied*implied/2
The maths: the break even move
For one day, with F the future and s the implied volatility, the move priced is F times s times the square root of one day as a share of a year. That is about F times s divided by 19.1, which is the square root of the days in a year. Gamma times half the squared difference of the two moves is the day's P&L, so the break even is a move equal to the priced one. For a weekend, Friday to Monday, the chain counts three calendar days, and the priced move is the one day move times the square root of three, about 1.73. Some desks weight weekends less than three days. The chain's own clock, calendar days as a share of a year, is what the figures here use.
Gamma sits at the strike, so a scalper's edge fades once the index runs away from it
Read 2 more paragraphs, 116 words
The bill for it is theta. The straddle above loses 28.72 rupees a unit each day to time, Rs 1,867 for a lot, whether the index moves or not (module F22). So the daily result is the scalping gain less that bill, and module A01 showed they have one formula. This is the same identity from the buyer's side.
Here move is the index change that day in points, and implied is the move the option's volatility priced for that day. The second term is exactly the theta bill. The scalper earns on a day when the move is larger than the implied one and loses when it is smaller. There is no third case.
The lot problem
This is the Indian detail. A future trades in whole lots of 65 units. One lot of this straddle adds 0.119 units of delta for every point the index moves. The index has to move about 544 points before the delta has drifted by one whole future lot. A trader holding one straddle lot can therefore hardly scalp it with the future: even rounding to the nearest lot, the first hedge waits for a move of about 272 points, and then overshoots by half a lot. Hedging in lots needs a bigger book of options, or a different hedge instrument, and both have costs of their own.
Worked example: four sessions, 10 to 16 Jun 2026
The straddle was bought on the weekly chain (nearest the parity forward of 23,229.26, since the weekly has no future of its own). Implied volatility was 14.6%, so the priced move was about 177 points a day. Friday to Monday is three calendar days, so the priced move for 15 Jun is 306. The table uses the closes of the Nifty index: 23,214.95 on 10 Jun, then 23,161.60, 23,622.90, 23,853.90 and 23,989.15 on 16 Jun. The last column is the gain in rupees: half the gamma times the squared move less the squared priced move, for one lot, with gamma frozen at its 10 Jun value.
| Session | Move | Priced | Gain a lot |
|---|---|---|---|
| 11 Jun | −53 | 177 | −1,700 |
| 12 Jun | 461 | 177 | 10,844 |
| 15 Jun | 231 | 306 | −2,422 |
| 16 Jun | 135 | 177 | −777 |
| Total | 5,946 |
The term this section turns on: Priced move
Read one more paragraph, 99 words
Three days lost and one won. On 12 Jun the move was about 6.8 times the priced move in squared terms, and that one day carried the whole result. Over the four sessions the squared moves added up to 2,87,297 against 1,87,782 priced, a ratio of 1.53. The scalper needed the ratio to be above one, and it was. The total is an upper estimate. Gamma was frozen at its starting value, but by 16 Jun the index stood 3.1% above the strike and most of the gamma was gone (the chart above shows how fast it fades).
Not the same as owning the straddle
The straddle itself went from Rs 346.05 to Rs 739.15 at settlement, a gain of Rs 393.10 a unit and Rs 25,552 a lot. Most of that is direction: the index ended well above the strike. A scalper sells future on the way up, so gives that direction away and keeps the variance gap instead. The two numbers answer different questions, and the scalper's is smaller here.
The straddle needs a move to pay, and the curve above the V is the time still to be paid for
The other exposure: vega
A bought straddle is long vega as well as long gamma. The 30 day at the money implied volatility fell from 14.9% on 10 Jun to 12.6% on 16 Jun, a drop of 2.3 points, in a week when the index ran. That pairing is common: a big move often takes some of the fear out of the price. The straddle's vega was 23.7 rupees a unit for each volatility point. A weekly option does not move one for one with the 30 day number, so read this as an order of magnitude and not a measurement. At the 10 Jun vega, a fall of that size would take about Rs 3,501 off the mark of a lot while it is open, a large share of the scalping gain in the table. A straddle held to expiry is settled on the index close, so this hits a scalper who closes early or marks to market along the way. Gamma and vega are bought together, and the second can take back what the first earns.
Expiry day, when gamma is largest
Gamma is highest on the last day, and so is theta. A scalper who holds an at the money straddle into the final session pays the biggest daily bill and owns the most curve. The index on that day can only be priced by what it does, which is why expiry day gets so much attention (module I25). Since 20 Nov 2024, a sold index option that expires that day carries extra margin, so the seller on the other side of the scalper pays to hold the risk the scalper has bought.
What it costs to scalp
If the index moves at the priced pace, the gross result is zero. Every reset then costs money. STT on one future lot sold at the 10 Jun level is about Rs 755 at 0.05%, before the exchange charge, stamp duty, GST, brokerage and the gap between bid and ask (module F06). Costs are paid on a quiet day too. So a scalper needs the index to beat the priced pace by enough to cover them, not just to beat it.
The long run: who pays whom
A buyer of gamma pays the implied volatility and earns the realised one. Across the 300 sessions from 1 Jul 2025 to 15 Sep 2026, the 30 day at the money implied volatility averaged 12.4% and the 20 day realised volatility of the index 11.5%. Implied was above realised on average. That gap is the variance risk premium (the paper in the sources): sellers of volatility are paid it, and they are paid for carrying the days when it goes wrong. It is an average over a short sample and says nothing about any one week.
Implied volatility sat above realised on average, which is the price a gamma buyer pays
Who is on the other side
Whoever sold the straddle is short the same gamma. In module A01 the seller of the option pays when squared moves beat the priced ones, and earns the theta that the scalper pays. The two are mirror images, and neither is the safe side. For a scalper the questions are how the cost of hedging compares with the gap between realised and implied, and how big the book must be to hedge in lots. For a seller the questions are the gap days and the margin (module A14). The simulator below lets you set the volatility paid and the volatility delivered and see the scatter of outcomes.
The common mistake
The claim: Gamma scalping earns from volatility whichever way the market moves, so it cannot lose on direction.
It is true that direction drops out. It is false that nothing else decides the result. In the table three of four days lost, and the total was positive only because one move was large against the priced one. The scalper pays Rs 1,867 a day to be there, resets in whole lots of 65 units, pays costs on each reset and, over the sample above, bought implied volatility that sat above realised on average. The idea is sound. The edge is not automatic. The short answer is on What is gamma scalping?.
Key takeaways
- A scalper earns when the index moves more than the option priced.
- The break even daily move is the move that implied volatility priced.
- In four sessions of the example, three lost and one won.
- One straddle lot needed 544 points before one future lot was off.
- Implied averaged 12.4% against 11.5% realised across the sample.
- Costs make a zero gross result a net loss.
Two terms to hold on to from 14.16
Before you trade. Derivatives trading carries a high risk of loss. SEBI's study "Profitability of Individual Traders in the Equity Derivatives Segment (FY25 to FY26)" (August 2026) found that 87.7% of individual traders in equity F&O made net losses in FY26, with aggregate net losses of Rs 91,685 crore. Read the SEBI study