IMC Trading Interview Guide

IMC filters with game-based cognitive tests before any human sees your application, then spends a Superday watching whether you update a price when the information changes.

Sample IMC questions

Higher or Lower on the Next Card

A dealer shuffles a standard 52 52 -card deck and turns one card face up. Ranks run from ace low (1 1 ) to king (13 13 ); suits are irrelevant. Looking at that card, you call whether the *next* card off the deck will be higher or lower in rank. A tie -- the next card matching the rank of the face-up card -- counts as a loss.

Playing optimally, what is your probability of winning, and what is the fair price to enter a game that pays $1 \$1 on a win and nothing on a loss?

Gondola Headcount Puzzle

A mountain gondola makes three stops on its descent, and the same thing happens at every stop: three-quarters of the riders currently aboard step off, and then 77 new riders climb on. Nobody recorded how many riders started at the summit. What is the smallest possible number of riders aboard after the third stop?

The Spread That Informed Flow Demands

You are the sole market maker in a one-off contract. Everyone agrees the contract will settle at either $90 \$90 or $110 \$110 , each equally likely, so its unconditional fair value is $100 \$100 .

Orders arrive one at a time, in unit size. Each order comes from an informed counterparty with probability α=0.2 \alpha = 0.2 -- that trader already knows the settlement value, and will buy from you whenever your offer sits below it and sell to you whenever your bid sits above it. Otherwise (probability 0.8 0.8 ) the order comes from a noise trader who buys or sells with probability 12 \tfrac{1}{2} each, regardless of price.

(a) Why is quoting $100 \$100 bid at $100 \$100 offered a losing strategy, even though $100 \$100 is the honest fair value? (b) Find the bid and offer at which you break even on the next order. (c) What happens to your quote if informed flow rises to α=0.4 \alpha = 0.4 ?