Gamma Exposure, and What It Can Actually Tell You

What Gamma Exposure Measures

The per-strike calculation is one line:

Gamma Exposure at a strike = gamma × (call open interest − put open interest) × spot² × 0.01

Gamma is the rate at which an option's delta changes as the underlying moves. Open interest is how many contracts stand at that strike. The spot squared term and the 0.01 together convert the answer out of "delta per point" and into rupees per one per cent move — which is the form that lets you compare a reading on NIFTY against one on RELIANCE without the index level dominating the comparison.

Calls carry a positive sign and puts a negative one. That is the convention, and the next section is about why it is a convention rather than a fact.

One detail that matters more than it looks. A call and a put at the same strike, same expiry, must share one gamma — gamma is a property of the strike, not of the side. This site recomputes that single gamma from a volatility solved off the traded premium rather than reading a per-leg value from the data feed, because feed values for the two sides at one strike do not always agree, and gamma exposure sums across the whole board where any such disagreement compounds.

The Assumption Underneath Every Figure

Exchanges publish open interest. They never publish who is on which side of it.

So the sign convention above — calls positive, puts negative — is not derived from data. It encodes a specific belief: that the people who wrote the options are long the calls and short the puts. That belief comes from American index markets, where customers characteristically buy protective puts and write covered calls, leaving the other side holding the mirror image.

This is not a reason to ignore the numbers. It is a reason to read them as one description of positioning among several, rather than as a measurement. Two practical consequences follow.

The absolute reading survives the doubt. Adding call and put gamma together instead of differencing them produces a figure that does not depend on who holds what. It cannot tell you whether hedging damps or amplifies, but it does tell you which strikes carry enough gamma for hedging to matter at all. On this site that is the "both sides" column and the heaviest-gamma strike.

Weekly expiries compress the whole picture. A very large share of Indian index option turnover sits in contracts expiring within days. Gamma rises sharply as expiry approaches, so a reading taken on the near series can be dominated by contracts that will not exist by the end of the week. This site reads one expiry at a time and refuses to publish a figure once a series has settled, rather than summing tenors that behave nothing alike.

The Gamma Flip, and Why the Method Matters

The gamma flip — also called the zero-gamma level — is the price at which the measured total would change sign. Above it, hedging a positively-signed book means selling into strength and buying into weakness, which works against each move. Below it, the arithmetic reverses and hedging works with the move instead.

There are two ways to find that level, and they are not equivalent.

The cheap way is to run a cumulative total across the strike ladder from the bottom up and interpolate wherever the running sum crosses zero. It requires no extra computation because the per-strike figures already exist.

The honest way is to ask the actual question: if the underlying were at some other level, what would net gamma exposure be? That means recomputing every strike's gamma at each candidate price and looking for where the total crosses zero.

This site uses the second method. The cumulative curve is still drawn, because seeing a line cross zero is the clearest way to understand what "flip" means — but the level itself is solved, not read off the curve.

A level near spot deserves less weight than a level far from it, for the ordinary reason that any estimate is most fragile where it is closest to being tested. When spot and the flip sit within a strike of each other this site merges them into one marker rather than drawing two lines a few pixels apart and implying a precision it does not have.

The Gamma Walls, and How They Differ from Open-Interest Walls

The call gamma wall is the strike above spot carrying the most call-side gamma exposure; the put gamma wall is the strike below spot carrying the most put-side gamma exposure. Both are restricted to their own side of spot deliberately — an unrestricted search puts both walls on the same at-the-money strike, because that strike carries the most gamma on both sides at once, which is arithmetically true and completely useless as a level.

Both walls are shown alongside the heaviest open-interest strike on the same side, so the two readings can be compared directly rather than discovered to conflict on another page.

A wall is not a barrier. It is a description of where the option board is concentrated. Price moves through walls routinely, and the concentration itself changes through the session as contracts are opened and closed.

Where the Reading Stops Being Worth Much

Four situations flatten the usefulness of everything above, and this site marks each of them rather than publishing a number that looks the same as a good one.

Expiry day. Gamma rises without limit as time runs out. Measured on a real expiry session, a net reading swung from strongly positive to strongly negative inside ninety minutes on the same board, while the following series stayed stable throughout. Readings on a settling series are suppressed, not estimated.

A book that cancels itself. When call and put exposure are close to equal, the sign of the total carries almost no information — a small shift moves it either way. This site publishes the share of strike-level exposure that survives the sum, and declines to name a regime when that share is small, rather than reporting a coin-flip as a verdict.

Thin or illiquid boards. Single-stock ladders gain strikes during the session as new contracts are listed. A total that grows partly because the board grew is not the same as a total that grew because positions were taken, so any such growth is stated explicitly.

Strikes with no usable price. A volatility can only be solved where something traded. Strikes that carry open interest but no price contribute nothing, and the share of the book that could not be measured is reported alongside the figure.

The general principle is the one the rest of this site follows: a figure that could not be measured is named as such, never rendered as a zero or a dash. A blank cell reads as "nothing is happening" when the truth is "this could not be computed", and those are very different statements.