What financial analyst interviews actually test

Who's on the other side of the table

The shape of the interview depends heavily on which kind of financial analyst role you're going for, and it's worth being clear with yourself about which one this is before you prepare.

In corporate FP&A, you'll usually meet the hiring manager first — often an FP&A manager or director — and then a second round with a budget holder or the CFO if the company is small enough that the CFO still interviews analysts directly. In an investment bank or private equity shop, early rounds are often with an associate or VP, and a superday or final round brings in a partner or MD who will spend less time on your CV and more time watching how you think under pressure. In equity research, you may meet the sector head and a portfolio manager on the buy side who wants to know whether your view on a stock survives contact with someone who's traded it. Each of these people is listening for something slightly different, but they share one thing: they've all built models themselves, so they can tell within a few sentences whether you have too.

Some processes also include an HR or talent screen first — a fifteen-minute call about availability, salary expectations and right to work — but that's not where the job is won or lost. The technical rounds are.

The technical test: what it actually looks like

Most financial analyst processes include some form of practical assessment, and it's rarely a written exam with model answers. It's closer to watching you work.

The most common format is an Excel-based exercise: you're given a dataset — a trial balance, a set of monthly actuals, a company's historical financials pulled from filings — and asked to build something from it within a time limit, sometimes on-site with someone watching, sometimes as a take-home with 24 to 72 hours. For corporate and FP&A roles this is often a variance analysis or a rolling forecast: explain why actual spend diverged from budget and project the next few periods. For banking, private equity and corporate development roles it's more often a three-statement model or a discounted cash flow, sometimes bolted onto a short LBO. For equity research, expect to be asked to build or critique a model for a named company in the sector you'd be covering, and to defend a price target.

What's being tested isn't whether you know the keyboard shortcuts, though fluency helps and its absence is noticeable. It's whether your model holds together: does the balance sheet balance, does cash flow from operations correctly reflect changes in working capital, have you hardcoded a number that should be a formula, and can you explain every assumption you made without checking your notes. Interviewers who do this for a living can open a candidate's workbook and find the plug in about thirty seconds.

Some employers still run brainteasers or mental-maths drills — estimate the market size for umbrellas in London, work out a percentage change without a calculator — but these have fallen out of favour at most serious shops because they test composure more than competence. If you get one, treat it as a chance to narrate your reasoning out loud; the answer matters less than whether your logic is followable.

The questions that separate a modeller from someone who's read about modelling

A lot of interview questions sound conversational but are doing diagnostic work. Here's what they're actually checking.

"Walk me through the three financial statements and how they link." This isn't a definitions test. They want to hear you trace a single transaction — say, a $10 increase in depreciation — through the income statement (lower EBIT, lower tax), the cash flow statement (added back, but tax saving is real cash), and the balance sheet (lower PP&E, lower retained earnings, higher cash than it would otherwise have been). If you can only describe each statement separately, that's the tell.

"Tell me about a forecast you built and how you handled the assumptions." They're listening for how you chose your drivers — did you tie revenue to a volume-and-price build, or headcount to a hiring plan — and whether you stress-tested them, not just whether the forecast existed.

"Describe a variance you investigated and what you found." In FP&A especially, this is the core of the job: numbers move, and someone has to find out why before the business partner asks. A strong answer names the specific driver — a vendor price increase, a timing shift in revenue recognition, a one-off reclass — and says what changed as a result of the finding, whether that's a forecast revision or a conversation with the budget owner.

"How would you value this company?" or "What's the difference between enterprise value and equity value?" These come up constantly in banking and equity research interviews and less often in corporate roles, but when they do come up in corporate finance it's usually because the role touches M&A or investor relations. The question is checking whether you can move fluidly between DCF, comparable company multiples and precedent transactions, and whether you know which one is appropriate when — a company with negative EBITDA can't sensibly be valued on an EV/EBITDA multiple, for instance.

"Tell me about a time your numbers were wrong." This is a competence question dressed as a behavioural one. They want the mechanism of the error and the mechanism of the catch, not a story about resilience.

Questions about specific tools also carry more weight than they seem to. "What's your experience with SAP, Oracle, or NetSuite?" and "Have you used Power BI or Tableau to build reporting?" aren't small talk — reporting infrastructure varies enormously between employers, and a hiring manager wants to know how much ramp-up time you'll need before you're productive in theirs.

What a shallow answer sounds like

Someone who's done this job for a while can hear a shallow answer within a sentence or two, and it usually has one of these shapes.

On valuation: "I'd look at comparable companies and see what multiple they trade at." True but empty — it doesn't say which multiple, why that multiple fits this company's stage and margin profile, or how you'd adjust for a control premium if it's a transaction comp rather than a trading comp.

On variance analysis: "I look at the numbers and see if they're up or down versus budget." That's not analysis, that's reading a spreadsheet. The real answer names a driver and a next action.

On modelling experience: "I'm very comfortable in Excel and have built financial models before." This is the single most common non-answer in the process, and interviewers have heard it hundreds of times. It survives no follow-up question. Compare it with: "I built a three-statement model for a £40m revenue business, drove revenue off a unit-economics build by channel, and linked debt schedule circularity with a switch to avoid iterative calculation errors." One of these tells the interviewer nothing; the other tells them exactly how far to push the next question.

On the CFA or ACA qualification, if it's on your CV: being unable to explain, in your own words, why you'd use FCFE rather than FCFF for a company with a complex capital structure, or being unable to walk through a simple deferred tax calculation, suggests the qualification was passed rather than absorbed. Interviewers who hold the same qualification notice this quickly.

What to do before you walk in

Build or rebuild one model from scratch in the week before the interview — a real company's, from public filings if you can find them, not a template. Be ready to open it and explain every formula. Know, without checking, how the three statements link and what happens to each one when a specific line item changes. If the role is FP&A, be ready with one real variance you found and what you did about it, not a hypothetical. If it touches valuation, know which method you'd choose for a specific company and be ready to defend the choice, not just recite the menu of options.

jobmarket.pro reads the advert in full before it prepares anything, and for a financial analyst role that means it can tell you, before the interview, which of these — FP&A variance work, three-statement modelling, valuation — the employer is actually likely to test, based on what the advert asked for.