Citadel, SIG & DRW “Make Me a Market” Interview Game

"Make me a market on the length of the Amazon" is a live exercise in trader interviews at firms like Optiver, SIG, IMC and Flow Traders: you quote a bid and an ask on a number you don't know exactly, the interviewer trades against your quote, then asks you to mark your P&L and quote again. What's being graded is not trivia — it's whether your mid reflects a calibrated estimate and whether your spread is tight enough to be a real market without getting you picked off.

This page is a free version of that round. You quote two-sided markets on real fact and guesstimate prompts; a counterparty with a noisy read of the true value lifts your offer, hits your bid, or passes — sizing up to 5 lots the more mispriced you look — and every trade settles at the truth.

Scoring tracks what interviewers actually probe: how centred your mid was in standard deviations, spread discipline, the edge you captured over the session, and short break-even and P&L quizzes — the follow-up questions you will get in the room.

What is the "make me a market" interview question?

An interviewer asks you to quote a bid and an ask on an uncertain quantity — a fact (height of Everest) or a Fermi estimate (piano tuners in a city). They trade against your quote, then ask you to compute your P&L, re-quote, or justify your spread. It tests calibrated estimation and quoting discipline, not trivia recall.

Which firms ask market-making questions in interviews?

It is standard in trader interviews at market-making firms — Optiver, SIG, IMC, Flow Traders, DRW, Akuna — and appears in trading rounds at hedge funds and prop shops. It usually comes up in early rounds and is often repeated several times with the interviewer trading against each quote.

How wide should my bid-ask spread be?

A reasonable benchmark is about twice your honest one-standard-deviation uncertainty. Quote tighter and an informed counterparty picks you off; quote wider and you rarely trade and signal you have no estimate. This game caps quotes at three times the fair spread, so you cannot show a market that can never lose.

How does this market making game score you?

Four components: centring (how close your mid sat to the true value, measured in standard deviations), spread discipline, edge (normalized P&L captured across the session), and a technical quiz on break-even and P&L. A bot with a noisy read of the truth trades against your quotes and all positions settle at the true value.