Concept 1 of 81 min
Making a market
Two traders, one contract
| Pays | 100 |
|---|---|
| Probability | 40% |
| One trader offers | 34 |
| Another bids | 46 |
- Value = 100 × 40% = 40
- Buy at 34, sell at 46: 46 − 34 = 12
Two quotes on the wrong sides of fair value lock in 12 whatever happens.
Quoting a coin flip
| Pays on heads | 20 |
|---|---|
| Heads | 50% |
- Fair value = 0.5 × 20 = 10
- Quote 9 / 11, tight around it
A tight market around fair value invites trades both ways.
Buying where they sell
| Their market | 20 / 22 |
|---|---|
| Your value | 25 |
- Buy at 22, their offer
- Edge = 25 − 22 = 3
You can only buy where the other trader sells.
Prices that let someone buy below and sell above fair value are taken on both sides. A very large size or a hidden direction calls for a wider spread.
A market nobody can lose on, such as 0 / 10 on a coin paying 10, is a market nobody trades; interviewers want a tight market around a reasoned estimate.
When someone quotes below your value, lift their offer; you cannot buy on a bid. If one counterpart sells below fair value and another buys above it, doing both locks in the gap whatever happens.
Large size carries more information and more market impact, so dealers widen to be paid for it.
For estimation markets, such as the number of US states whose names start with a letter, reason out loud, then quote around the estimate; for counts of random events, use the mean and the standard deviation.
- Centring on the middle of the range instead of the expected value.
- Buying on a bid, or selling on an offer.
- Quoting the normal tight price for a huge order of unknown direction.
I ask you to make me a market on how many piano tuners work in London. How do you go about it?
A complete answer covers
- Build an estimate out loud: population, pianos per household, tunings a year, tunings per tuner.
- Centre the market on that estimate.
- Set the width from your uncertainty, not as wide as possible.
- Update the market as I trade with you or give you information.
Also in our technical interviews, on camera, with a score.
Concept 2 of 81 min
Adverse selection and the winner's curse
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inConcept 3 of 81 min
Managing inventory
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inConcept 4 of 81 min
Expected values of dice and cards
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inConcept 5 of 81 min
Waiting times and optimal stopping
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inConcept 6 of 81 min
Edge, fees and variance
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inConcept 7 of 81 min
Kelly sizing
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inConcept 8 of 81 min
AdvancedWhere the break-even quotes sit when some traders know
The why, the charts, the worked example and the interview questions are in Essentials, with every quiz.
See the plansSign inRecap
- Centre your market on expected value, quote both sides tightly around it, and widen for size or uncertainty.
- A trade against you is information: someone who keeps buying is telling you your price is too low.
- Long inventory: lower both quotes; short: raise both.
- Expected value is linear, but E[X²] is not E[X]²; condition on everything you are shown.
- Kelly stakes (b × p − q) ÷ b of the bankroll, about μ ÷ σ² for small edges; most traders bet a fraction of it.
25 questions on Market-making and trading games
Timed, like the real one, drawn from a bank of 33. Every answer explained.