Market-Making Interview Questions: Bid-Ask, Fair Value, Spread Reasoning

The market-making round is the central interview at firms whose business is making markets — Jane Street, Optiver, SIG, Akuna, IMC, Citadel Securities, and most options-focused prop shops. Even hedge fund and bank-strats interviews include market-making-flavored questions because the framework (fair value, edge, spread, inventory risk) is foundational to thinking about trading. This guide covers what these questions look like, the mental model interviewers want to see, and how to perform under pressure.

What Market Making Actually Is

A market maker simultaneously quotes a bid (price they’ll buy at) and an offer / ask (price they’ll sell at) for a financial instrument. The difference is the bid-ask spread, and capturing this spread on flow is the market maker’s primary source of edge.

Mathematically: if fair value is V, the market maker quotes bid = V – spread/2 and offer = V + spread/2. When customers buy at the offer, the market maker profits spread/2. When customers sell at the bid, same. Repeat across millions of trades.

The challenges:

You don’t know V exactly — it’s an estimate based on available information
Your quotes inform other market participants; trading against you signals their view of V
You accumulate inventory (positions) that create directional risk
Transaction costs (exchange fees, hedging costs) eat into the spread
Adverse selection: informed traders pick off your stale quotes

The Standard Interview Format

The interviewer plays a market with you. They reveal partial information about an unknown quantity, ask you to quote a market (bid X, offer Y), then trade against you based on whether your prices look attractive. As the round continues, more information is revealed, and you update your market.

The interviewer is evaluating:

Can you anchor an initial price reasonably?
How wide is your spread relative to your uncertainty?
Do you update appropriately when information arrives?
Do you manage inventory thoughtfully?
Do you communicate clearly under pressure?

Walk Through a Canonical Problem

Setup: “I’m thinking of the number of pages in War and Peace. Make me a market on the answer.”

Step 1: Anchor. What do you actually believe? War and Peace is famously long. Probably 1,000–1,500 pages. Maybe 1,300 as a point estimate.

Step 2: Set spread proportional to uncertainty. “I’ll bid 1,200 and offer 1,400.” Wide enough to account for my uncertainty, tight enough that the interviewer might trade.

Step 3: Interviewer trades. “I’ll sell to you at 1,200 (sells you 1 share at your bid).” Now you’re long 1 share at 1,200; if the answer is anything below 1,200 you lose money.

Step 4: Update mental position. The interviewer chose to sell, which suggests their estimate is closer to 1,200 than 1,400 — otherwise they’d buy from you at 1,400. Update down: maybe truth is closer to 1,200 than I thought.

Step 5: New information arrives. “It’s the Maude translation, hardcover edition.” This narrows the search. Maude editions tend to be ~1,300–1,400 pages.

Step 6: Update market. New market: “1,250 / 1,350.” Tighter spread because more information; centered higher because of new info about the edition.

Step 7: Manage inventory. You’re long 1 share already. If you’re risk-averse, your bid should reflect that — you need higher conviction to add to the long. New market might be “1,240 / 1,340” to discourage the interviewer from selling more.

Final answer (real): the Maude translation is roughly 1,200 to 1,400 pages depending on edition. Your final market should be reasonably close.

Pricing Frameworks

Fair value

The expected value of the unknown given all current information. Not what’s true; what you believe is true on average.

Spread sizing

Spread should be roughly 2× your standard deviation of belief about fair value. Wide enough that random one-trade error doesn’t hurt; tight enough to attract flow. As information arrives and uncertainty drops, spread should tighten.

Edge per trade

Spread / 2 is the expected profit per trade if your fair value is correct. If you collect $0.20 spread on $10,000 of notional, that’s 0.002% per round-trip. Volume turns this into real money.

Inventory adjustment

If you’re long N units of inventory, your willingness to add to long position should be lower (skew bid down) and willingness to sell should be higher (skew offer down). This naturally reduces inventory exposure as the market continues to trade.

Common Question Types

Market on a single number

“How many gas stations are in the United States?” Anchor with Fermi estimation, set spread proportional to uncertainty, update with information.

Market on the result of a probabilistic process

“I’ll roll three dice and sum them. Make a market on the sum.” Expected value = 10.5; sum ranges 3 to 18. Reasonable market: “9 / 12” with adjustment for variance.

Market on conditional information

“I’ll roll a die. If it’s even, I’ll roll again and tell you the second roll. If odd, I’ll keep that roll. What’s the expected value of the final number?” E[final] = (1/2)(3.5) + (1/2)(3.5) = 3.5. Same regardless of branching. Reasonable market: “3.0 / 4.0” reflecting uncertainty.

Market on a sequential process

“I’ll flip a coin until I get heads. Make a market on the number of flips.” E[flips] = 2 (geometric distribution with p = 1/2). Variance is 2, so std dev is √2 ≈ 1.4. Reasonable market: “1 / 3” or so.

Adversarial markets

“I’m going to choose to either buy or sell from you based on which is most profitable for me. Now make a market.” This is the realistic market-making setup; the interviewer (representing all customers) will trade against you whenever your price is wrong. Spreads need to be wider here than in friendly settings.

Things Strong Candidates Do

Anchor with Fermi estimation

For numerical questions, build a quick estimate from base rates, then commit to a market reasonably near that estimate. Don’t paralyze trying to compute the perfect price.

State your reasoning out loud

“I think US gas stations is order of magnitude 100,000. Why: 300 million people, ~3 people per car, ~100 million cars, drivers visit gas stations a few times per month, so… 50,000 to 200,000 seems right. I’ll quote 80,000 / 130,000.” This reasoning matters more than the final answer.

Update sharply on information

When the interviewer reveals new info, update visibly. “Knowing it’s the EU rather than US changes things; let me revise my market to 100,000 / 150,000.”

Acknowledge inventory exposure

“You sold me one at 1,200, so I’m long. If I’m long, I should be slightly less aggressive on the bid going forward. New market: 1,180 / 1,290.”

Manage spread with conviction

As information narrows your fair-value estimate, narrow the spread. Holding wide spreads when you have good information is leaving money on the table; the interviewer notices.

Things Weak Candidates Do

Paralyze on initial price. Spending two minutes trying to get the perfect anchor before quoting. Just pick a reasonable number and refine.
Make spread too wide and never tighten. “100 / 200” stays “100 / 200” through the whole round. Information should narrow spreads.
Ignore inventory. Keep getting traded against and don’t adjust prices to discourage further accumulation.
Compute silently. Long pauses without verbal reasoning leave the interviewer guessing whether you’re thinking or stuck.
Overcomplicate. Trying to apply textbook market-making models in 90 seconds; the interviewer wants intuition, not rigor.

Practice Strategy

Weeks 1–2: read about market-making mechanics. Heard on the Street has some classic problems; Trading and Exchanges by Larry Harris gives the broader microstructure context.

Weeks 3–4: practice Fermi estimation problems. Books on quantitative reasoning (“How Many Piano Tuners in Chicago?”) help build the skill of anchoring numerical estimates quickly.

Weeks 5+: mock market-making rounds with a peer or coach. Have them play the interviewer; trade against you on partial information. The format is unfamiliar without practice; mock rounds dramatically improve performance.

Frequently Asked Questions

How wide should my spread be?

Roughly 2× the standard deviation of your belief about fair value. If you think the answer is $100 with std dev $20, quote a market of about $80 / $120. Spread should narrow as information arrives. If you have very high uncertainty, wider spreads are appropriate; if you’re confident, tight spreads attract flow without much risk. The exact ratio matters less than directionally appropriate sizing.

What if the interviewer trades against me, then immediately reveals information that proves I was wrong?

That’s the game — market makers face adverse selection. The interviewer is testing whether you respond appropriately. Update your prices, acknowledge inventory exposure, and continue. Don’t get rattled. A trader who freezes after one bad trade loses money long-term; a trader who keeps quoting and adjusts prices rationally collects spread on average.

How does this round vary across firms?

Jane Street: emphasizes intellectual reasoning and clear thinking. Citadel Securities: tighter spreads expected, faster updates. Optiver: heavier on options-specific market making with implied vol reasoning. SIG: poker / gamesmanship overlay; more emphasis on reading the interviewer’s signals. Akuna and IMC: closer to Optiver’s style. The framework is the same; the cultural texture differs. Practicing the framework prepares you for any of them.

Is this round just probability with extra steps?

Probability matters but isn’t sufficient. Strong candidates combine probability reasoning (what’s the expected value?), risk management (how much can I lose?), inventory management (am I taking on too much directional exposure?), and game theory (what does the interviewer’s behavior tell me?). It’s a multi-skill round; pure math wins coding rounds but doesn’t win market-making rounds.

How important is this round for hiring decisions?

Decisive at market-making firms. Jane Street, Optiver, SIG, Akuna, IMC, Citadel Securities all weight market-making rounds heavily — often the highest-weight round in the loop. Strong candidates with weak market-making rounds frequently get rejected; weak candidates with strong market-making rounds sometimes pass despite weakness elsewhere. For trader-track interviews, this round is non-negotiable; for pure SWE-track at the same firms, less central but still real.