What changes between interview rounds, and what is a bad sign if it changes? This note explains the difference, using a first-round question about the SPY spread and a later conversation about the same mechanism as the public example.
Depth changes, the standard does not
A later round should go deeper on the same habits, not swap in a new personality test. The first conversation samples whether a candidate can reason and be corrected. A later conversation samples whether that reasoning survives a longer problem and a second listener. What should not change is the boundary: internships and experienced interviews alike are about the work, not about a guessed performance, and live capital decisions are not the prize at the end of a loop. New York, London, and Hong Kong may host different rounds of one process. They do not host different definitions of a good answer.
A bad change is a first round that is a puzzle sport and a last round that is a sales pitch, or the reverse. The candidate cannot prepare for a process that does not know what it is sampling. The firm owes the candidate a stable sample. Interviews that drift into headcount, pay bands as theater, or a request to speculate about results have drifted off this note. The candidate may say so.
Between rounds, the questions should get deeper on the same habits. They should not change from a test of reasoning into a request for a pitch or a performance story.
The same spread, asked harder
A first round might ask what the spread in SPY is for. A competent answer says it is the distance between the best bid and the best offer, that crossing it is a cost of immediacy, and that a last price hides the cost. A later round can stay with SPY and ask what happens when an order is larger than the size at the offer, or how a backtest that uses midpoints has assumed the cost away. The mechanism did not change. The number of consequences the candidate is asked to hold did. That is a fair change between rounds.
An unfair change is a later round that abandons the mechanism and asks the candidate to pitch a trade in SPY, to guess the firm’s positions, or to treat an internship as a shortcut to sending orders. Interns are trusted to ship research tools. They are not interviewed as if the summer were a license to act for the firm in the market. The word license is the wrong word anyway. Nothing in a round confers authority over an order path. That is true in the New York room, the London room, and the Hong Kong room.
Logistics may change and should not be confused with standards. A later round might be at a different hour because London and New York are both trying to be awake, or because Hong Kong is joining a debrief. The hour is not the test. Candidates should not interpret a video call as a lesser round or a city as a veto. If a round is decisive, someone should be willing to say what it is deciding. Silence plus a harder puzzle is not a decision procedure.
How to read the arc
Between rounds, listen for whether the questions got more precise. Precision is a good sign. A sudden request for a narrative about why the firm is impressive is a bad sign, including when the candidate is tempted to enjoy it. The public example remains available as a compass. If the conversation can no longer be tied back to something like the SPY spread, an opening auction, or a missing sequence number, it may have stopped being an interview for this work.
What changes between rounds should be depth. What stays is the habit and the boundary. Internships and experienced loops in New York, London, and Hong Kong can share that shape. This note does not promise how many rounds exist.
This piece does not fix the number of rounds, a timeline, or an offer. It describes what a change between interviews is for, in New York, London, and Hong Kong, including internship loops. The SPY spread, asked once and then asked harder, is the example.
