Financial Analysis interview preparation

Questions

100

Traced to a firm

42

Firms

28

Updated

September 2026

Asked at

All firmsMoody's7Bain Capital3SSState Street3AMAres Management2BLBlackRock2DED.E. Shaw2MSMorgan Stanley2Oaktree Capital Management2S&P Global2Bridgewater Associates1Citadel1FTFranklin Templeton1Golub Capital1HWHarris Williams1Houlihan Lokey1J.P. Morgan1Jane Street1MWMarshall Wace1Millennium Management1Morningstar1PIMCO1Sycamore Partners1TSTruist Securities1Two Sigma1Vanguard1WMWellington Management1Wells Fargo Securities1Wolverine Trading1

Topic

All topicsThree statements9Accounting policy and standards5Working capital and cash7Ratio analysis8Forecasting and budgeting9Variance and management reporting7Unit economics and costing8Capital budgeting7Cost of capital and valuation7Markets and rates5Modelling, Excel and data8Business partnering6Brainteasers and estimation4Fit and career10

Level

AnyCoreIntermediateHard

Type

AnyTechnicalCaseBrainteaserMarket viewFit

1. 087If a date is a Monday, what day of the week will the same date be a year from now?Brainteasers and estimationIntermediatetechnicalOaktree Capital ManagementGeneralist · Los Angeles · 2022

   Say this

   Tuesday, unless a 29 February falls in between, in which case Wednesday. A normal year is 365 days, which is 52 weeks plus one day, so the day of the week advances by one. A leap year adds two.

   Then walk it

   1. The mechanism: 365 divided by 7 is 52 remainder 1. The remainder is the whole answer, so a year later is one weekday forward.
   2. 366 divided by 7 leaves remainder 2, so across a leap day you advance two, landing on Wednesday.
   3. The condition to state, because it is the actual test: does 29 February fall inside the twelve months between the two dates? So a date in March 2023 to March 2024 crosses the leap day and advances two; a date in March 2024 to March 2025 does not and advances one.
   4. The same modular arithmetic answers the harder version. Ten years forward is ten plus the number of leap days, modulo seven. Ten years with two leap days is 12 mod 7 equals 5, so five weekdays forward.
   5. Say the assumption out loud, that you are on the Gregorian calendar and ignoring the century rule where 1900 was not a leap year but 2000 was. That earns credit because it shows you checked the edge case rather than got lucky.
   6. And the reason these get asked in a credit or corporate finance interview: they are testing whether you reach for the structure, remainder after dividing by seven, rather than counting.

   Where candidates lose it

   Answering Tuesday with no condition attached. The interviewer is waiting to see whether you raise the leap year yourself. Saying 'Tuesday, or Wednesday if a 29 February falls in between' answers it completely in one sentence.

   Expect next

   • What about ten years from now?
   • What day of the week were you born on?
   • How many leap days between 2000 and 2100?

   Reported by candidates at Oaktree Capital Management (Generalist, Los Angeles, 2022). Source: Wall Street Oasis.

2. 088Quick mental arithmetic: a business does 4.2 crore of revenue a month at a 23 percent EBITDA margin. What is annual EBITDA, and what is the business worth at 9 times?Brainteasers and estimationIntermediatetechnicalMSMorgan StanleyUtilities · Baltimore · 2025

   Say this

   Annual revenue is about 50 crore, EBITDA is about 11.6 crore, and at 9 times that is roughly 104 crore of enterprise value. Round as you go and say the rounding out loud.

   Then walk it

   1. Revenue: 4.2 times 12 is 50.4 crore. Call it 50 and note you rounded down slightly.
   2. EBITDA: 23 percent of 50 is easiest as 25 percent minus 2 percent. A quarter of 50 is 12.5, 2 percent of 50 is 1, so 11.5. The precise figure on 50.4 is 11.6.
   3. Value: 9 times 11.6. Do 10 times minus 1 times, so 116 minus 11.6, about 104 crore. Never multiply by 9 directly in your head.
   4. The techniques worth having automatic: break percentages into round components, multiply by 10 and subtract, use the rule of 72 for doubling times, and remember that a 10 percent margin means enterprise value equals 0.9 times revenue at a 9 times multiple.
   5. Then add the detail that gets you the point: say whether this is enterprise or equity value. At 104 crore of enterprise value with 30 crore of net debt, the equity is 74. Interviewers ask the mental maths question and then check whether you know what the number is.
   6. And narrate. An interviewer cannot mark a silent pause, but they can mark a clean decomposition even if you fumble the last digit.

   Where candidates lose it

   Going silent and trying to be exact. Speed with narration beats precision in silence. Also, quoting the 9 times number as equity value. State that it is enterprise value and that you would deduct net debt.

   Expect next

   • Now tell me the equity value if net debt is 30 crore.
   • If EBITDA grows 15 percent a year, how long until it doubles?
   • What multiple would make this worth 150 crore?

   Reported by candidates at Morgan Stanley (Utilities, Baltimore, 2025). Source: Wall Street Oasis.

3. 089How would you fix the city of Baltimore?Brainteasers and estimationHardsuperdayMSMorgan StanleyUtilities · Baltimore · 2025

   Say this

   Treat a city like an entity with a P&L, a balance sheet and a customer base. Define what 'fix' means, diagnose whether the problem is revenue, cost or population, then pick two interventions with the best return per rupee and say how you would measure them.

   Then walk it

   1. Define the objective first, out loud, because the question is deliberately unbounded. I would take it as: reverse population decline and restore a sustainable budget, because most city problems are downstream of those two.
   2. Then the diagnosis as a financial structure. Revenue is property tax, which depends on population and property values. Cost is pensions, debt service, policing, schools and infrastructure, most of which is fixed. That is a high-operating-leverage entity facing a declining top line, which is the actual problem.
   3. So the doom loop is the mechanism to name: population falls, tax base shrinks, services get cut, more people leave. Any fix has to break the loop rather than balance the budget for one year, because balancing it with service cuts accelerates the decline.
   4. Then prioritise interventions by leverage, not by sympathy. Two candidates: make the tax base competitive with the surrounding county so families stop leaving, and concentrate capital on a small number of blocks rather than spreading it thinly, because urban recovery is non-linear and needs density to hold.
   5. Then fund it honestly: the balance sheet has assets, land, parking, utilities, and the pension liability is the real constraint, so any plan that does not address pension cost is arithmetic theatre.
   6. Then measurement and a time frame: net migration, vacancy rate, and tax collection per capita, reviewed annually, with a stated expectation that this is a ten-year programme. And I would say plainly what I am not confident about, because pretending to solve a city in four minutes is the wrong answer to this question.

   Where candidates lose it

   Listing policy opinions. It is a structuring test. Define the objective, name the revenue-cost-population mechanism, pick two prioritised interventions and say how you would measure them. Political commentary scores zero.

   Expect next

   • Which intervention first, and what would it cost?
   • How would you fund it?
   • How would you know in two years whether it was working?

   Reported by candidates at Morgan Stanley (Utilities, Baltimore, 2025). Source: Wall Street Oasis.

4. 090If you had infinite time and money, what invention would you make?Brainteasers and estimationCoresuperdayDED.E. ShawConsulting · New York · 2026

   Say this

   Pick one specific thing, explain the problem it solves, and then reason about it like an analyst: who benefits, what it would cost, what would be hard, and what could go wrong. The invention is a prompt; the thinking is what is being marked.

   Then walk it

   1. Specificity beats ambition. 'Cure all disease' is a non-answer. Something like a cheap, accurate soil-nutrient sensor that tells a smallholder farmer exactly what to apply, in the local language, from a phone, is answerable and testable.
   2. Then the problem size, with a rough number. Roughly a hundred million-plus smallholdings in India, most fertilising by habit rather than measurement, which wastes input cost and damages yields. That frames the value without pretending to precision.
   3. Then the mechanism and the hard part, because interviewers listen for whether you can identify the real constraint. The sensor is not the hard part; distribution, calibration across soil types and getting the recommendation trusted are.
   4. Then the economics: what it would have to cost per unit to be adopted, who pays, and whether the value is captured through hardware, subscription or through the input supplier.
   5. Then the second-order effects, which is where you can be genuinely interesting: it shifts bargaining power toward farmers, changes fertiliser demand mix, and has an environmental effect through runoff. Good answers hold both the upside and the unintended consequence.
   6. Keep it to two minutes and show some actual enthusiasm. This question is also a personality screen: they want to know whether you think about anything outside a spreadsheet.

   Where candidates lose it

   Giving a grand abstraction, or a joke answer. The point is structured curiosity: one specific invention, the problem, the hard constraint, the economics, and an unintended consequence. And do not choose something you cannot discuss for three minutes.

   Expect next

   • What would be the hardest part to actually build?
   • How would you fund it if money were not infinite?
   • What would be the unintended consequence?

   Reported by candidates at D.E. Shaw (Consulting, New York, 2026). Source: Wall Street Oasis.

Firm tags come from public, anonymous candidate reports on Wall Street Oasis: strong signal, not sworn testimony. Firms are named as the places a question was reported, not as partners of Fin Maverick. Answers are written for this page to show how to think out loud; they are not scripts to recite.