Basic Prerequisites: none - this page assumes only that you know a bid is the price at which someone will buy and an ask is the price at which someone will sell.


1. Intuition & Practical Objective

Imagine a currency-exchange booth at an airport. It posts buy €1 at $1.08 and sell €1 at $1.10. If a tourist sells €100, the booth pays $108. If the next tourist buys €100, the booth receives $110. The booth ends the day flat in euros and $2 richer. That $2 is the spread - the booth’s gross revenue. This is exactly what a market maker does, one hundred million times a day and one cent at a time.

The intuition a beginner must absorb first is that the spread is a price for a service, not a gift. The booth is selling immediacy: it guarantees that any tourist can trade right now, at a known price, without waiting for another tourist to appear. That service has a cost, and the whole economics of market making is the accounting of that cost:

Gross spread − adverse selection − inventory cost − fees + rebate = true profit.

Almost every intuition failure in this topic comes from looking only at the first term. The quoted spread is the headline, and every other term is a deduction from it. This page gives you the five-term picture; the rest of the folder makes each term precise.

The one-sentence essence. “A market maker is a shop that sells immediacy; its revenue is the spread, its cost of goods sold is adverse selection and inventory, and the exchange’s rebates are a supplier discount - profit is what’s left.”


2. Mathematical Ground Truth & Derivations

Work per share. Let the maker post a two-sided quote around the mid-price : bid , ask , so the quoted spread is and is the half-spread the maker captures on each fill (assuming the mid does not move between the two legs).

Term 1 - Gross spread capture. Each fill earns the maker per share.

Term 2 - Adverse selection . Some of the counterparties know something. When an informed buyer lifts your ask, the mid is about to rise; you sold above the pre-trade mid but the mid itself jumps. If the mid moves against your new position by on average, you lose per share. (This is the whole content of Glosten–Milgrom: the winner’s curse.)

Term 3 - Inventory cost . Fills are one-sided in bursts. The inventory you accumulate must be financed (repo/capital), hedged, and carried through volatility - and it is a non-diversifiable risk you must be paid for (Stoll 1978; Grossman & Miller 1988). Write it as a per-share penalty .

Term 4 - Fees and rebates. If you post a resting order and it is hit, you are a maker and the exchange pays you a rebate (equivalently a make fee ). If you cross the spread, you are a taker and pay .

The master equation (per share, maker side):

Break-even half-spread. Set and solve for the quote that just covers costs:

Two immediate readings, both central to this folder:

  1. The rebate lowers the break-even spread. A maker receiving can quote tighter and still break even - this is the mechanism by which exchanges believe rebates improve liquidity (see 03 · Maker-Taker Fees & Rebates).
  2. Cash terms are the only tradable lever. is set by competition, by the counterparties you cannot avoid, by your capital and risk appetite. The rebate is the one line an exchange can hand you directly - which is exactly why it became a central piece of market-structure policy.

3. Computational Implementation - the five terms, in dollars

Standard library only. This is the mental model made arithmetic: a desk that captures a penny of half-spread and pays for it in adverse selection and inventory, with and without the exchange rebate.

Read the last two lines carefully. The rebate is 40% of net profit when it is present, and removing it cuts the desk’s profit by 40% (300k). This is the empirical reason the maker-taker debate is so bitter: a payment that looks like a rounding error on the spread is a first-order term in the margin. That is the subject of the rest of this folder.


4. Failure Modes & First-Principles Breakdowns

  1. Reading the quoted spread as revenue. The quoted pencil-and-paper spread is gross. The maker keeps . In the example above, the headline spread is and the true take is - the spread overstates profit by 4×. This single error is the most common one in casual commentary.
  2. Treating the rebate as free money. It is not: is financed by the taker’s , and under competition it is largely passed through into a tighter quote (Colliard & Foucault 2012). Ignoring the pass-through double-counts the subsidy.
  3. Ignoring adverse selection because it is invisible on the confirm screen. Every fill prints a price; no blotter line says “you were picked off.” is an expectation, measurable only statistically (→ VPIN, 02 · Market-Maker P&L).
  4. Assuming both legs fill. The two-tourist example is the ideal: bid and ask both filled at the same mid. Real spreads are earned when two-sided flow arrives; when only one side arrives you are left holding inventory, which is Term 3.

5. References

  • Stoll, Hans R. (1978), The supply of dealer services in securities markets, JF 33(4)
  • Grossman & Miller (1988), Liquidity and market structure, JF 43(3)
  • Hasbrouck (2007)
  • Demsetz (1968), The cost of transacting, QJE 82

6. Connected Graph Bridges