4. The Market Maker: The Most Predictable Player in the Market

4. The Market Maker: The Most Predictable Player in the Market

You open your terminal and buy five hundred SPX puts with a single click. The trade fills in a fraction of a second. Now a question few people ever stop to ask: who sold them to you? Five hundred contracts of crash insurance — at the exact moment you wanted it. There was no line of eager sellers. On the other side of your trade — as of nearly every options trade in the world — stood a participant who had no wish whatsoever to argue with you about the future. He was simply obliged to answer.

The profession: being on the other side

A market maker (from here on, MM or dealer) is a firm whose job is to continuously quote both sides of the market — a bid and an ask — across thousands of options at once. Want to buy — he will sell; want to sell — he will buy. The exchange grants him privileges for this, and he intends to earn the spread — the penny-sized difference between his buying and selling prices, repeated millions of times.

So that this doesn't stay an abstraction: market makers are not "exchange functions" and certainly not private individuals, but the largest trading firms in the world — Citadel Securities, Jane Street, IMC Trading and their peers. Industrial machines with thousands of employees, for whom quoting SPX options is an assembly-line business.

Notice what this business model does not contain: an opinion about the market. The MM is not betting on a rise or a fall — he is selling liquidity. But over a day of such work, a position accumulates all by itself: everything the clients bought, he has sold; everything the clients dumped, he has collected. His book is a mirror image of the whole market's desires. The clients chose their trades. The MM did not choose his position.

Now recall the option seller's payoff from lesson 1: capped profit, unlimited loss. The dealer who sold you five hundred puts will lose a fortune in a crash — if he leaves things as they are. He will not. And what he does instead is the central mechanism of this entire course.

Delta hedging, step by step

The dealer's task is to earn the spread without carrying market risk. So the risk has to be extinguished. We already wrote out the instructions in lesson 3, when we took delta apart: an options position is equivalent to a position in the underlying.

Watch the numbers. Clients bought 1,000 calls from the dealer with delta 0.40. The dealer's book: minus 1,000 calls — the equivalent of a short position of 400 units of the index. The market rises a point, the book loses 400. The cure is obvious: buy futures for those same 400 units. Now a rising index earns exactly as much on the futures as the sold calls eat. The book's delta is zero; the dealer is delta-neutral and calmly earns his spread. (We are deliberately simplifying the contract multipliers — the arithmetic of conversion factors adds nothing to the point.)

That would be the end of the story if delta held still. But lesson 3 taught us: delta is alive. The index rises ten points — the delta of the sold calls creeps from 0.40 up to 0.55. The dealer's book is now equivalent to a short of 550 units, and only 400 units of futures have been bought. To get back to zero, he must buy another 150 — into a market that has already risen. The index falls back? Delta slides toward 0.40, the surplus futures weigh on the book — sell 150, into a market that has fallen.

There it is — the scramble promised in the conversation about gamma. The dealer does not want to trade futures — he is forced to: every market move upsets his neutrality, and he chases his own delta all day long. This scramble is not background noise. By volume it is one of the largest forces in the futures market, and it is entirely mechanical.

Two regimes: gamma rules the flow

The direction of the scramble depends on one single sign — the sign of the gamma of the dealers' book. And that sign has two regimes, with opposite effects on the market.

The charts show the quantity of futures the dealer is obliged to hold as a function of the index price. The hedging rule is one and the same; watch how differently it plays out.

The dealer is long gamma (left chart) — on balance, clients have sold him more options than they bought: his book is stuffed with bought calls and puts. Then the required futures position falls as the price rises: the market goes up — the dealer sells futures; the market goes down — he buys them back. Selling at the top and buying at the bottom — the dealer mechanically trades against the move. Thousands of such trades dampen the swings: the market turns viscous, moves die down, and the price seems drawn toward the big option strikes. This pull-to-the-strike effect is called a pin (pinning) — we will meet it in the data more than once.

The dealer is short gamma (right chart) — clients have bought up his options: insurance policies, lottery tickets, everything from lesson 1. The curve flips: now the higher the price, the more futures the dealer must hold. The market rises — the dealer buys after it; the market falls — he sells in its wake. Every move he is forced to amplify. A decline breeds dealer selling, which deepens the decline, which demands fresh selling. On a calm day this is a light rattle; on a bad one, a self-feeding spiral — and it is in this regime that the fastest crashes and the most vicious vertical rebounds occur.

Take in the asymmetry of blame: in both regimes the dealer does exactly the same thing — holds zero. Whether he stabilizes the market or whips it up is decided not by him but by the aggregate position of his clients.

The dealer book: from positions to a map

A real dealer holds not a thousand calls but hundreds of thousands of positions across all strikes and all expirations at once, and the longs and shorts inside the book partially cancel out. But any book, however complicated, has a bottom line at every moment: a total delta (which the dealer holds at zero), a total gamma, and its sign.

More than that — the book's contribution can be decomposed back across strikes: here are the levels where the gamma is concentrated, here it is positive, there negative, and here is the price at which the sign of the whole book flips. What you get is a map: where on the price scale lie the forces that will make the dealer buy, and where — the ones that will make him sell. This map — dealer exposure — is exactly what our GEX indicators measure. How to compute it is what the coming lessons will teach.

One honest caveat, on which the entire options-analytics industry stands: nobody publishes the dealer's book. It can only be estimated — from indirect data, with assumptions about which side of the trades the clients were on. What data exists for this, how fresh it is, and what exactly it reveals — that is what the entire remainder of the course is about.

Why this is gold

Let's pause for a second and appreciate what we have gained. Everything else in the market is opinion: an analyst can change his mind, a fund can cancel an order, the crowd can take fright. The dealer's hedging flow is the only large force that is obliged to show up: it is mechanical, its direction is computable from the book, and it is anchored to specific price levels. Forecasting other people's desires is hard; other people's obligations can be calculated.

Two caveats to close, so the map doesn't harden into a myth. The sign of gamma says nothing about direction: short gamma is not a crash forecast but an amplifier of any move, up or down; long gamma is a brake, not a guarantee of a rally. And the hedging flow is one force among many: a headline, a large fund's order flow, or a macro print can run over any gamma in the dealers' book. The book gives you a map of the terrain — where the market is viscous and where it is slippery — but not the route.

---

So the market has a player whose trades are predictable because they are mandatory — all that remains is to get hold of data that reveals his book. The industry started with what lay on the surface: the daily open interest reports. Out of them grew an entire generation of analytics — option levels, walls, magnets — and our first pack of indicators. That is where the next part begins: what the classic data could do, and where its ceiling ran.