Market making, explained properly

Everything you need to play these games well, in plain English. No finance background needed. It takes about fifteen minutes, and you can skip around.

The game in one minute

Someone asks a question with a number for an answer, like "how many bones are in the human body?". You don't need to know the answer. You need to make a market: a price you'd buy at and a price you'd sell at.

Then people trade with you. At the end the real answer's revealed and every trade settles against it. Buy at 180 when the answer's 206 and you made 26 a lot. Sell at 180 and you lost 26.

That's it. The skill is picking a sensible price, then learning from what people do with it.

The six golden rules

If you only read one bit, read this.

1. Centre on your best guess. Not on what you think anyone wants to hear.
2. Width = how unsure you are. Wide enough to be honest, tight enough to be useful. Never silly-wide.
3. Ask "does that trade tell me something?" If the other person might know more, a trade is a clue. Move.
4. When you're hit again and again, make three moves: shift your centre properly, widen a touch, show less size on that side.
5. Know your position and break-even at all times. Say them out loud.
6. Process beats result. Losing a bit early is normal. What matters is how fast you adapt.

Bids, offers and the spread

A market is two numbers. The bid is where you'll buy, and it's the lower one. The ask (or offer) is where you'll sell, the higher one. The gap between them is the spread, and halfway between is the mid, roughly your best guess.

Traders say it out loud as "200 at 210": bid 200, offer 210. Add a size and it's "200 at 210, 5 up", meaning you'll trade 5 lots either way.

When lots of people quote, all the bids and offers stack up into an order book:

If you want to buy straight away, you pay the best offer (210). If you want to sell straight away, you get the best bid (200).

Buying at someone's offer is called lifting it. Selling at someone's bid is hitting it.

Your first price

Your opening market matters most. Every trade before you've adjusted costs you money, so a sensible start saves a lot.

1. Get a rough number fast. Anchor on something you know, then adjust (more on this below). Fifteen seconds of thinking beats five seconds of panic.
2. Put that number in the middle. Your bid and offer go either side of it.
3. Pick a width you can defend. See the next section.
4. Say it like you mean it. "I'm 280 at 320" sounds a lot better than "erm… maybe around 300?".

How wide to go

Your width is your protection against being wrong. Wide is safe, but nobody trades with you and you learn nothing. Tight shows confidence and gets trades, but costs more when you're off.

A good starting point is 10–20% of your price when you're genuinely unsure, getting tighter as you learn. In this simulator the interviewer refuses anything much wider than 35% on Warm-up, 20% on Standard and 12% on Hard, and a bit more on genuinely niche questions.

Is that trade telling you something?

This is the question good traders ask after every fill. It depends on who's trading with you:

• If they might know more than you (an interviewer who's looked up the answer, or a friend who clearly knows football), a trade is real information. If they buy, it's probably higher. If they keep buying, it's definitely higher. Move your centre.
• If they know no more than you (someone who's guessing too), a trade doesn't change what the answer is likely to be. But it does change your position, so you might lean your price to manage that risk. More on this in skew or widen.

In an interview, assume they know more. Saying this out loud ("you bought, and you probably know more than me, so I'm moving up") is exactly the sort of thinking they're listening for.

Reading the flow

The flow is the trades coming at you, and what they're telling you. Here's the classic situation: you open at 120 at 200 on the number of bones in the body, and the interviewer keeps buying.

Each buy says "too cheap". When they finally pass, your market contains their number, and here it contains the answer too.

When they keep buying, don't just nudge your offer up. Make three moves:

1. Raise your centre. This is the big one, and the one people skip, because it means admitting your first guess was wrong.
2. Widen a touch. Someone who keeps trading one way probably knows something, so charge a bit more for it.
3. Show less size on the side they're hitting. If they keep buying, offer fewer lots so you're not building a big short.

The same works in reverse when they keep selling: drop your centre, widen a touch, show less on your bid.

How far to move

There's no exact formula, but these rules of thumb work well:

• One trade: shift by about half your width. It's a clue, but it could be noise.
• Same side again: shift by a full width or more. Two in a row is far stronger than one.
• A small trade (1 lot): weaker than a big one. Often it means your market is already close and they're just leaning one way. Nudge, don't jump.
• "Can you tighten that?" Good news: they like your level. Keep the same middle and bring both sides in.
• They pass: you're about right. Tighten slightly to invite a trade and learn more.

Don't swing wildly the other way either. Jumping from 200 to 500 off one trade is as bad as not moving. On Hard, the interviewer will occasionally bluff, trading the wrong way to see if you panic. One odd trade shouldn't undo everything else you've learned.

Skew or widen?

These are two different tools, and mixing them up is very common. Widening changes the gap and keeps the centre. Skewing moves both prices together and keeps the gap.

The dotted line is your original centre (200). Widen for uncertainty. Skew for your position, or because you think the answer's moved.

Why does skewing up help when you're short? Your higher bid makes people more likely to sell to you (which flattens you out), and your higher offer makes them less likely to buy from you (which would make you even shorter).

Position and size

Your position is how many lots you hold. Buy 5 and you're long 5 (+5): you want the answer to be high. Sell 5 and you're short 5 (−5): you want it low. Every time someone buys from you, you get shorter. Every time they sell to you, you get longer.

The bigger your position, the more a wrong answer hurts. Two habits keep you out of trouble:

• Size with your confidence. Small while you're working things out, bigger once you've got a real view. Taking every trade that looks good, without noticing your position's got huge, is a classic way to lose the lot on one bad answer.
• Lean against a growing position. Getting long and staying long (or adding to it) is the single most common way to blow up. If you're getting long, skew down a bit.

When news comes in

Part-way through a round you might get an announcement. It comes in two kinds, and you should react to both straight away, before anyone else does:

• A clue: a related fact, never the answer. Asked how many toilets Wembley has, you might hear "Wembley holds 90,000 people". Work out what it means for your number. If it changes your view, move your whole market; if it backs you up, hold your level.
• A change of contract, like "we now settle at 70% of the answer" or "the answer plus 50". Work out what your old middle is worth under the new rule (70% of 300 is 210) and move your whole market there. Then keep reading the flow.
• A random range, like "somewhere between 50% and 80% of the answer". Use the middle (65%) for your price, and go a bit wider because there's more risk now.

Don't forget your position. If you're long 10 and the contract drops to 70%, those 10 lots just got cheaper. That's why big positions are dangerous when news can land at any moment.

PnL and break-even

When the round ends, the answer's revealed and your position settles at it:

Cash goes down when you buy (you paid) and up when you sell (you got paid). Using the bones example above, where the interviewer bought 5 at 200 and 5 at 204:

A really useful number is your break-even: the answer where your PnL would be exactly zero. Here it's 2,020 ÷ 10 = 202.

Short positions make money when the answer's lower than break-even. Long positions are the mirror image.

Notice you lost a bit even though you adjusted well. That's normal. The interviewer knew more and got the first trades in. What they're marking is how quickly you adapted.

Getting picked off

Being picked off means trading with someone who knew better. Here's the uncomfortable truth: when someone trades with you, it's often because your price was wrong. The trades you get are skewed towards the ones you'd rather not have done.

You can't avoid this completely, and interviewers will sometimes tell you right after a trade that you got done badly, just to see how you react. The right reaction is boring: update your price, note your position, keep quoting. A trader who freezes after one bad trade loses more than one who calmly adjusts.

Playing with friends

In the group game everyone trades the same question at once on a live order book. A round has a natural rhythm:

You don't have to trade from the first second. Often the best edge comes once the market's formed.

If you don't know much, go in a bit later

Here's a trick good players use. If a question's outside your comfort zone, you don't have to guess blind. Let others quote first. The best bid and offer quickly become the crowd's combined guess, which is often better than any one person's. Then you only trade where you genuinely disagree, or you make a tight market around that level and earn the spread.

Making vs taking

Making means quoting your own bid and offer and waiting for people to trade with you. On average you earn the spread. Taking means hitting someone else's price (the BUY and SELL buttons). That costs you the spread, so only do it when you're fairly sure they're wrong.

A few more that win rounds

• Don't chase. If the price has run away, let it go. Paying up because you're scared of missing out is how money disappears.
• Watch who's trading. If your mate who supports Arsenal keeps buying on a Premier League question, take note.
• Mind the bell. A big position on a guess, with ten seconds left, is a coin flip. Trim it if you're not sure.
• Use the "What if" box to see your PnL at any answer, and the order book to see where everyone else is.

In the interview

Trading firms use these games because they show how you think when you don't know the answer. From what candidates and prep guides report, they're mostly marking:

• A reasoned first price. Say your logic out loud, briefly.
• Updating on information. The big one. Move when you're traded with.
• Consistency. Your bid's below your offer, your width makes sense, and you don't contradict yourself.
• Keeping track. Your position and rough PnL, without being asked twice.
• Composure. Losing early is fine. Freezing isn't.

They mark your reasoning as much as your numbers. Someone who loses a little with clear thinking and a controlled position usually beats someone who wins by getting lucky.

Getting a rough number

For "how many / how big" questions, break it into pieces you can reason about. This is called Fermi estimation.

• Anchor on something you know. "A storey's about 3 m, so a 100-storey tower is about 300 m."
• Split it up. "52 weekends is 104 days off, plus 8 bank holidays, so a UK year has about 253 working days."
• Bound it. "Definitely more than 100, definitely less than 500." That gives you a starting width too.
• Round aggressively. You want the right ballpark, not the right digit.

For maths questions, look for a shortcut: 17 × 23 = (20 − 3)(20 + 3) = 400 − 9 = 391. If you're sure, quote tight. Confidence you've earned is a good look.

Why firms care

Market makers are firms that quote buy and sell prices all day in things like shares, options and commodities, using their own money. They make a small amount on lots of trades by earning the spread, and they help keep markets running smoothly for everyone else. Optiver, IMC, Susquehanna (SIG), Jane Street, Maven Securities, Flow Traders and Onyx Capital Group (in oil derivatives) all do some version of this.

The daily job is exactly this game: price something uncertain, notice when you're being picked off, manage your position, and don't panic. Several of these firms grew out of trading floors. Optiver, for example, started as a single trader on Amsterdam's options exchange. That's why market-making games turn up so often in their interviews: it's the job in miniature.

Most of the quoting is done by computers now, but people still decide the prices, the risk and the strategy. The instincts you build here are the same ones.

Common mistakes

Glossary