Gamma — how fast delta itself changes
The answer
Gamma answers one question. When the underlying moves by 1 unit, how much does delta change? Delta tells you how fast your option is moving. Gamma tells you how quickly that speed itself is changing. High gamma means your position is becoming more directional, fast, without you doing anything.
Why this costs you money
Gamma is the reason a short option position goes from comfortable to uncontrollable in a single session, and it is the reason most people who write options do not see it coming.
Here is what happens, step by step, and it happens every time.
You write an out-of-the-money call. Its delta is −0.15, so the position behaves like being short a small amount of the underlying. That feels manageable, and it is.
The market rises 2%. The option is now closer to the strike, so its delta has risen to −0.35. Your position is now more than twice as directional as when you opened it, and it is directional in the losing direction.
The market rises another 2%. Delta is now −0.65. You are now short an amount of underlying you never chose to be short of, and every further rise costs more per point than the last one did.
Nothing in that sequence involved a decision by you. The position got larger and worse automatically. That automatic growth is gamma, and it is why a short option loss is not proportional to the market move — it accelerates.
The buyer's version of the same mechanism is the reverse and it is favourable: a long option becomes more responsive as it moves in your favour, so gains accelerate. But the buyer pays for that every single day, in time decay. Gamma and theta are the same trade seen from 2 sides. Whoever has the acceleration pays the rent. Whoever collects the rent has the acceleration working against them.
The practical loss for a buyer is this: you can pay for gamma for weeks and never receive a move large enough to use it. The practical loss for a writer is this: you collect rent for months and then, on 1 day, the acceleration takes back more than the rent ever paid.
How it works
Gamma is the rate of change of delta. If you have met the idea of speed and acceleration in physics, delta is the speed and gamma is the acceleration. If you have not, here is the version that needs nothing.
Delta is not a fixed number. Gamma is the number that says how unfixed it is.
Three facts about gamma cover almost every practical situation.
Fact 1 — gamma is highest at the money
An option deep in the money already has a delta near 1. It cannot rise much further, so a move in the underlying changes its delta very little. Gamma is small.
An option deep out of the money has a delta near 0. A small move does not change that either. Gamma is small.
An option at the money is on the boundary between the 2 states. A small move genuinely changes whether it is likely to finish in the money or not, so delta moves a lot. Gamma is highest exactly where the outcome is most uncertain.
Fact 2 — gamma rises sharply as expiry approaches
An option with 3 months to run has plenty of time for the underlying to move back and forth. A 1% move today does not settle much, so delta changes slowly. Gamma is modest.
An option with 3 hours to run is close to being decided. A 1% move now settles the question, so delta swings from 0.3 to 0.8 quickly. Gamma is enormous.
At expiry itself, at-the-money gamma becomes extreme: delta flips between near 0 and near 1 as the underlying crosses the strike. This is the mathematical reason expiry-day trading behaves unlike any other day of the week.
Fact 3 — buyers are long gamma, writers are short gamma
- Buy an option — you have positive gamma. Your delta improves as the market moves, in either direction. Your gains accelerate and your losses decelerate.
- Write an option — you have negative gamma. Your delta worsens as the market moves, in either direction. Your losses accelerate and your gains decelerate.
Read that again with the word "either" in mind. Short gamma is not a directional risk. A short straddle is short gamma both ways: it loses if the market rises sharply and loses if it falls sharply. The only outcome it wants is stillness.
What being short gamma feels like
This is worth describing, because the feeling is a reliable warning and most people ignore it.
A short gamma position requires you to trade in the direction that is hurting you, and to do it repeatedly. If you are short calls and the market rises, the correct hedge is to buy the underlying — at a higher price than a minute ago. If the market then falls back, you sell what you bought, at a loss. Every oscillation costs money.
That is not bad execution. It is the structure. A short gamma position is paid to be still and charged for every movement, and it cannot decline to pay.
Long gamma is the mirror image. You buy low and sell high mechanically as the market oscillates, and each round trip earns a little. You pay for that privilege through theta.
What it tells you, and what it does not
Gamma tells you how stable your delta is. That is the single most useful thing to know before an event with a known date, because it tells you whether your position will still be the position you chose after the announcement.
It tells you where the risk in an option book actually sits. A book that looks delta neutral and is heavily short gamma is not neutral in any meaningful sense. It is neutral for the next few points and dangerous after that.
What gamma does not tell you is direction. Gamma is symmetric: a long option position has positive gamma whether it is a call or a put. Somebody who says "gamma is bullish" has confused it with delta.
It does not tell you magnitude of loss on its own, either. Gamma describes the rate of change of the rate of change. To know what a 6% move costs you, do not try to reason from gamma. Just reprice the position at a 6% move. Every broker's option calculator will do it.
The decision rule
Never hold a short option position through an event, or into the final day of expiry, without repricing it at a 5% and a 10% adverse move first.
The reason is gamma. Your current delta is not the delta you will have when it matters, and the difference is not small.
Three practical rules follow.
- Short gamma has a time limit. If you write options, the last day is the
most dangerous day, not the safest. The premium is smallest and the acceleration is greatest.
- Long gamma has a cost limit. If you buy options for the acceleration, you
are paying theta every day. Decide in advance how many days of that you will pay before the position has to justify itself.
- Never judge a position by its current delta alone. Judge it by its delta
at 3 prices: here, 5% up, and 5% down.
Try this now
Five minutes on your own option chain, and it makes gamma visible without any mathematics.
- Open the option chain for the current expiry on any index you follow, with the Greeks displayed.
- Find the strike nearest the current price. Write down its call delta, and the call deltas of the 2 strikes immediately above and immediately below it.
- Compute the difference in delta between each pair of neighbouring strikes. That difference, per unit of strike distance, is gamma made visible. You have measured it without a formula.
- Now switch to a much later expiry — 2 or 3 months out if available.
- Repeat step 2 and step 3 at the same 3 strikes.
What you should see. The delta differences between neighbouring strikes are much larger in the near expiry than in the far expiry. That is Fact 2 from this article, appearing on your own screen.
You should also see that the differences are largest around the money and smaller at strikes far from the current price, in both expiries. That is Fact 1.
Now the part that changes a decision. Pick the at-the-money call in the nearest expiry. Note its delta. Then find the strike about 4% above the current level and note that call's delta. If the market moved up 4% overnight, your at-the-money option would take on approximately the delta profile of a contract that is now 4% in the money — near 1. If you had written that option, you would be short close to a full lot of the index, overnight, having gone to sleep short about half of one.
If you write options, do this exercise on a Thursday or a Friday of an expiry week. The numbers are much larger and the point is much sharper.
Three real cases
1. Volkswagen AG, 26 to 28 October 2008 — hedgers forced to buy Porsche disclosed that through shares and cash-settled options it controlled a very large proportion of Volkswagen's ordinary shares, at a time when a substantial short interest existed in the stock. The free float available to short sellers was far smaller than they had assumed. Volkswagen's share price rose extremely sharply over 2 sessions, briefly making it, on paper, the most valuable listed company in the world. Everybody who had sold calls or was short the shares had to buy into a rising market, and their buying pushed it higher still. This is short gamma at the level of a whole market: a group of participants obliged to buy more, the more the price rises.
2. GameStop Corporation, January 2021 — the hedging loop Very large volumes of short-dated call options were bought by individual traders. Market makers who sold those calls hedged by buying shares. As the share price rose, the deltas of those calls rose, so the hedgers had to buy more shares, which pushed the price up further, which raised the deltas again. The share price reached an intraday high of about $483 on 28 January 2021. The US Securities and Exchange Commission published a staff report on the episode in October 2021. The loop is gamma. It also unwound: the same mechanism ran in reverse as the price fell, and most option buyers who arrived late lost everything.
3. Indian index options on expiry day, 2024 and 2025 — the regulator naming the mechanism Indian index options concentrate enormous volume into the final hours before expiry, when gamma is at its highest and premiums are almost entirely time value. SEBI's measures announced in October 2024 addressed this directly, including additional margin on short options positions on the day of expiry and a restriction to one weekly index expiry per exchange. Separately, SEBI passed an interim order in 2025 concerning alleged manipulation of index levels around expiry by a large trading firm, with a substantial disgorgement direction. Whatever the eventual outcome of that case, the structural fact stands: expiry day is when gamma is at its maximum, which is when a small move in the underlying has the largest possible effect on option positions. That is exactly when a market is most sensitive to pressure, and it is exactly when an individual writing options is least protected.
The question that resolves it
A novice asks: what is my position worth now?
An expert asks: what will my position have become, by the time I am able to do anything about it?
Gamma is the difference between those 2 questions. On a calm day with a long-dated option, the difference is small. On expiry afternoon, or through an overnight gap, the difference is the entire risk.
What would make this wrong
If delta were stable, gamma would be an irrelevance and short option positions would be as safe as they feel. Delta is not stable, and the instability is greatest precisely when the position is under stress. That is arithmetic derived from the shape of an option payoff, not an empirical claim.
Three honest limits.
Gamma is only useful over small moves. Like delta, it is a local approximation. For a large move, do not extrapolate from gamma. Reprice the position at the new level using a calculator.
Being short gamma is not automatically bad. Selling options and collecting theta is a legitimate and widely used institutional strategy. The failures in this article involve short gamma without a defined boundary. Buying a further option to cap the loss removes the unbounded part while keeping most of the income, and that is the correct retail version.
Gamma squeezes are widely over-claimed. After any sharp rally, somebody will say it was a gamma squeeze. Often it was simply buying. Establishing that dealer hedging drove a move requires positioning data that individuals do not have. Treat the explanation with suspicion unless there is evidence.
In India
Weekly expiries concentrate gamma into a small number of hours. For most of the week, index option gamma is moderate. On expiry day it is extreme, and the final 2 hours are the most extreme part of that. Volume follows the gamma.
Index options are European-style and cash-settled, so the gamma risk resolves into a cash difference at a single settlement value. There is no delivery complication, but there is also no way to escape the final settlement other than by closing the position before it.
Additional margin applies to short options positions on expiry day under SEBI's October 2024 measures. Read that measure as the regulator pricing gamma risk explicitly, having concluded that participants were not pricing it themselves.
Overnight gaps are the Indian gamma problem. The market is closed for around 17 hours out of every 24. A short gamma position cannot be adjusted during that window, so the acceleration happens in a single jump at the next open rather than progressively. In a market that trades continuously, short gamma is expensive. In a market that gaps, short gamma is occasionally fatal.
Single-stock options are physically settled, which adds a second layer at expiry: a position that finishes in the money creates a delivery obligation requiring the full contract value.
In the United States
Same-day-expiry index options are a very large share of index option volume. These contracts spend their entire life in the maximum-gamma regime, which is why they behave unlike any other product.
Equity options are American-style, so a short position can be assigned at any time. Combined with high gamma near expiry, a writer can be assigned on a position whose delta moved overnight, and wake up holding or owing 100 shares per contract.
Trading is close to continuous through index futures, so a professional running short gamma can adjust the hedge at any hour. That does not remove the risk, as October 1987 and March 2020 both demonstrated, but it changes its character from a single jump to a fast sequence of adjustments.
Pin risk is a recognised operational problem. An option that finishes almost exactly at the strike leaves the writer uncertain whether they will be assigned, so they do not know their Monday morning position.
Where they differ, and what that tells you
The window in which gamma acts differs, and that changes what a hedge is worth. An American short gamma position can be adjusted continuously, so its risk is the risk of moving fast. An Indian short gamma position cannot be adjusted overnight, so its risk is the risk of moving while you sleep. What that tells you is that continuous hedging is a strategy available to Americans and a fantasy for Indian retail traders, and any Indian following American material about "managing gamma" is following advice built for a market that never closes.
The regulator's response differs in kind. The United States has largely allowed the same-day-expiry market to develop and has studied it. India has priced it, through additional expiry-day margin, and constrained it, through fewer weekly expiries and higher minimum contract values. Both regulators looked at the same mathematics. One imposed a cost on the risky end of it. That difference tells you which regulator concluded its retail population was being harmed.
Expiry concentration is far higher in India. A single weekly index expiry per exchange concentrates all of that gamma into 1 session rather than spreading it across several. Concentration is not neutral. It makes the expiry session itself an event, which is why so much Indian retail volume is now in contracts that live for a few hours.
Carry this
- Gamma answers: how much does delta change when the underlying moves 1 unit?
- Gamma is highest at the money and rises sharply as expiry approaches.
- Buyers are long gamma. Writers are short gamma. There is no third option.
- Short gamma means losses accelerate and gains decelerate, in both directions.
- Never hold a short option through an event or into expiry without repricing it at a 5% and a 10% adverse move.