What is dealer positioning in options?

Why dealers matter at all

Market makers take the other side of a large share of customer options trades and hedge the resulting risk, primarily by trading the underlying. Their hedging is mechanical: it follows from their Greeks, not from a view. If you know roughly what they hold, you know roughly what they will have to trade as conditions change.

That is a rare thing in markets: a large participant whose future trades are, at least in principle, predictable from public information. It is why dealer positioning has become a standard lens for describing how a market is likely to behave, absorbing moves or amplifying them.

How it is estimated

The starting point is open interest by strike and expiry, published daily. Each contract's Greeks are calculated from a pricing model using current prices and implied volatilities. Those Greeks are multiplied by open interest and the contract multiplier to give the total exposure at each strike.

Then comes the hard part: deciding which side of each contract the dealer holds. The simplest convention assumes customers are net buyers of puts and net sellers of calls, often reflecting overwriting against stock holdings and buying of protection, which leaves dealers long calls and short puts. Another common convention assumes customers buy both, leaving dealers short both. More refined approaches classify each day's trades by whether they printed near the bid or the ask and whether they opened or closed positions, building an estimate of who added what. Some exchanges sell data that breaks volume down by participant type and by opening or closing, which narrows the guess further.

The output is a set of maps: gamma exposure by strike, delta exposure, and sometimes vanna and charm. From those come the familiar levels: the zero gamma level, the call wall, the put wall, and the Apex, the single largest gamma strike.

What it shows well

The sign of dealer gamma at the current price is the most robust output. When dealers are estimated to be long gamma, their hedging leans against moves, and markets tend to be more contained. When they are short gamma, hedging leans with moves, and ranges tend to widen. That is a statement about the character of trading, not its direction.

Concentration is the second. Where large open interest sits, large hedges are attached, and the market's behaviour near those strikes can change as price approaches them. Seeing where the heavy strikes are, and whether they are building or being unwound, is more informative than any single aggregate number.

Change through time is the third. Positioning built over days or weeks shifts around expirations and large trades. Comparing today's map with yesterday's, or with the map after expiring contracts are removed, shows how the structure is evolving.

Where it is weakest

The side assumption is the largest source of error. At strikes where a large institution sold options rather than bought them, as in collar programs and overwriting, the dealer is long rather than short, and the sign of the exposure flips. Two tools using the same open interest but different conventions can produce opposite readings at the same strike.

Open interest is also stale. It reflects the previous close and misses positions opened and closed during the day, which can be a large share of activity in same-day contracts. And dealers hedge across products, with futures, ETFs and other options, so the hedge may not land in the instrument the map describes.

Finally, dealers are not a single entity. Different firms hold different books and hedge on different schedules. Aggregate positioning describes a tendency, not the precise order flow of any individual dealer.

Using it sensibly

Treat dealer positioning as a description of how the market is likely to respond to a move, not as a forecast of the move. Read the sign at price first, then the location of the heavy strikes, then how both are changing. Be most sceptical of precise claims built on single strikes, and most attentive when several independent pieces of the map agree.

And remember that every number on the map inherits the side assumption. When a level matters to a decision, it is worth asking how the tool decided which side of that strike dealers are on.