Paying for Information Before You Quote

A common trading-interview scenario: you're about to quote a two-sided market on some uncertain event, and someone offers to sell you a piece of information — a peek at one relevant data point, a hint about which side of a trade a large counterparty is on — for a fixed fee, before you set your price. Should you pay?

The right way to answer isn't gut instinct, it's to actually compute the value of the information: work out your expected quoting outcome (spread captured, adverse-selection risk) without the information, then work out your expected outcome with it, assuming you use it optimally to adjust your quote or skip trading altogether when it's unfavorable, and pay the fee only if the improvement in expected outcome exceeds the fee. If the information would only rarely change your decision — say it flips your fair value estimate by less than your spread half the time — it's often not worth much, because you'd have quoted almost the same price anyway.

This question tests two things at once: whether a candidate treats information as something with a calculable, bounded dollar value rather than an unquestioned good ("more information is always better"), and whether they correctly account for the fact that a piece of information is worthless if it never changes what you'd actually do. Interviewers often follow up by changing the fee or the reliability of the tip, watching whether the candidate's answer updates consistently rather than sticking to a fixed "yes" or "no."