Which desk are you interviewing for

There is a specific way people fail finance interviews while knowing the material. They prepare "finance technicals" as a subject, arrive fluent in the three statements and a DCF, and then get asked to defend a purchase price, or to say what they would do with the company after buying it, or to explain why the deal they worked on happened at all. The questions were not harder. They were from a different desk.

The five desks below all draw on the same accounting and valuation. What separates them is the question they are ultimately asking you.

The overlap is real — all five can ask you to walk through a DCF — but the follow-up diverges immediately. Banking asks how you would sanity-check the terminal value. Private equity asks what return that implies at a given entry multiple. Research asks which of your assumptions the market disagrees with. Same question, three different second questions, and the second question is where the score is decided.

WatchLeland · 29:32

The three questions that open almost every banking interview, each followed by an actual spoken answer rather than a description of one. The example chapters are the valuable part — listen for how short the answers are, and how quickly each one commits to a specific reason instead of surveying.

"Tell me about yourself", "Why banking?" and "Why this firm?" — with worked examples Play “"Tell me about yourself", "Why banking?" and "Why this firm?" — with worked examples” on YouTube

The rounds, and what each one is actually testing

Fit and story. "Walk me through your resume" and "why banking". Treated as a warm-up by candidates and as a filter by interviewers. Practitioner guides converge on four question categories total — your story, fit, deals and markets, and technicals — which means the non-technical material is three quarters of the surface area and takes a fraction of the time to prepare.

Accounting. The three statements, how they link, and what a given transaction does to each. Almost always includes at least one linkage question, which is the subject of the next lesson.

Valuation. DCF, comparable companies, precedent transactions. Why the three disagree, and which you would trust for a given company.

M&A and merger models. Accretion/dilution, purchase price allocation, synergies, why a deal was done at all.

LBO. In banking, the mechanics. In private equity, an actual model — and the format varies sharply, from a mental paper LBO you do out loud with no paper, to a written paper LBO, to a one- to three-hour timed model, to a take-home model with a presentation.

The modelling test. A spreadsheet, a time limit, and a deliverable. Scored on accuracy and on whether your output supports a recommendation.

The stock pitch. Required in research and hedge-fund loops; increasingly common in banking as an optional flex. You bring it, so there is no excuse for it being weak.

The deal walkthrough. "Walk me through a deal you worked on." The single highest-leverage question for anyone with experience, and the one most often answered as a chronology when it should be answered as an argument.

The superday. Back-to-back interviews, sometimes ten or more short ones in a day, occasionally two-on-one. Reported pattern worth knowing: the hardest technical questions tend to come from the junior bankers — analysts and associates, who were asked them recently and remember them — while senior bankers ask fit and story. Any surrounding social event is part of the assessment.

Predicting the mix from the process you are in

The recruiting process tells you the round mix more reliably than the job title does.

Highly structured and fast — bulge-bracket or elite-boutique analysts recruiting on-cycle for large private equity funds. Interviews can start and finish inside 24–48 hours. Under that compression, the tests are standardised and speed matters: expect a paper LBO and a timed model, and expect your background and bank to carry weight you cannot change. Prepare for accuracy under a clock.

Unstructured and slow — off-cycle processes, smaller funds, non-traditional backgrounds. Runs for weeks or months, and the emphasis shifts to reasoning and to real deal or client experience. The same modelling test is scored differently: less on whether you finished, more on whether you can defend the assumptions.

If you are in between, the process is in between, and you will likely see both.

The one question to ask the recruiter. "What are the rounds, and what should I brush up on for each?" Recruiters answer this accurately because their incentive is for you to pass. It is a normal question with a normal answer and it is worth more than a week of untargeted reading.

The document behind the technical questions

Almost every accounting and valuation question in a finance loop is a question about something that appears in a public company's annual report. Reading one properly is the cheapest way to convert memorised answers into a system, because you see the numbers in the place they actually live.

The SEC's own investor guidance to the Form 10-K is worth reading end to end once. A few sections carry more interview weight than their length suggests:

• Item 1A, Risk Factors — listed in order of importance, and focused on the risks themselves rather than how the company mitigates them. This is where a "what would you worry about" answer comes from without you having to invent anything.
• Item 7, MD&A — management explaining its own results. The source of every "why did margins move" answer.
• Item 8, Financial Statements — audited. Worth knowing that an audit opinion can be unqualified, qualified, or a disclaimer of opinion, and that material weaknesses in internal control get disclosed.
• Item 9, Changes in and Disagreements with Accountants — short, often empty, and per the SEC's guidance, a disclosure many investors view as a red flag when it is not.
• Part III, Items 10–14 — executive compensation and related-party transactions, usually incorporated by reference from the proxy statement rather than printed in the 10-K. If you cannot find compensation in the 10-K, that is why.

Two structural facts that come up as follow-ups. Sarbanes-Oxley requires the CEO and CFO to personally certify the filing, which is why you find certifications as Exhibits 31 and 32. And any non-GAAP measure a company reports — "adjusted EBITDA" and its relatives — must be reconciled to the closest GAAP measure, which means the adjustments are itemised and you can read exactly what was added back.

The two questions scored hardest

Neither is a finance question.

"Why this desk, and why us?" The rejection here is generic answers. "I like financial markets" and "I want to learn a lot quickly" describe several thousand other candidates. What works is specific and falsifiable: a thing this firm does that you can name, connected to a thing you have actually done. Firm-specific knowledge is explicitly weighted in private-equity processes, where strategy and portfolio are fair game and "I did not look at your portfolio" is fatal.

"Walk me through a deal." The default failure is chronology — we were hired, we built a model, it closed. The version that scores is an argument: what the company was, why the transaction made sense for whom, what the valuation hinged on, what your specific contribution was, and what you would have done differently. If you interned rather than closed, the same structure applies to a pitch or a live process; what matters is a view, not a mandate.

What to prepare when you cannot tell

If the loop is opaque, prepare in this order, because it is the order of expected value:

1. The three statements as one system — the highest-frequency technical topic across every desk, and the thing every other model is built on.
2. A DCF you can defend under sensitivity — not recite. The next two lessons build both of these.
3. Merger and LBO mechanics to the level of the arithmetic — enough to compute accretion and a rough return without a spreadsheet.
4. Your story, your "why", and one deal or company you have a real view on — cheapest to prepare, disproportionately weighted.

Brain teasers and puzzle questions are worth deprioritising outside of quant and trading loops. Practitioner guides on banking interviews explicitly list them under what not to worry about, and the time is better spent on the four items above.

The rest of this module does items 1 through 3, and does them by building the models rather than describing them — because a model you have run is one you can still answer questions about when the interviewer changes an input.