Investment Banking interview preparation

Questions

100

Traced to a firm

100

Firms

46

Updated

September 2026

Asked at

All firmsTSTruist Securities10Rothschild & Co8Centerview Partners7CSCredit Suisse7HWHarris Williams6Houlihan Lokey6Lazard6Mizuho6Barclays5Citi5Deutsche Bank5Evercore5Moelis & Company5MSMorgan Stanley5Piper Sandler5RCRBC Capital Markets5Goldman Sachs4Nomura4TD Securities4Bank of America3GSGuggenheim Securities3J.P. Morgan3Jefferies3Moody's3Perella Weinberg Partners3WPWarburg Pincus3WBWilliam Blair3HSBC2Lincoln International2Scotiabank2TPTPG2UBS2Wells Fargo Securities2Advent International1Apollo Global Management1Bain Capital1Balyasny Asset Management1BLBlackRock1BPBNP Paribas1General Atlantic1Invesco1Morningstar1PIMCO1STSociété Générale1SSState Street1WMWellington Management1

Topic

All topicsAccounting14Valuation21M&A10Markets and deals10Capital markets3LBO8Leveraged finance3Restructuring2Credit3Debt capital markets2Capital structure2Case and estimation11Brainteasers6Fit5

Level

AnyCoreIntermediateHard

Type

AnyTechnicalCaseBrainteaserFitMarket view

1. 081How would you value your favourite animal?Case and estimationIntermediatesuperdayRothschild & CoGeneralist · New York · 2026

   Say this

   Pick an animal with an obvious cash flow so the question becomes tractable. A racehorse: value it on prize money, breeding fees and resale, less training and stabling costs, discounted over its career.

   Then walk it

   1. Choose the animal strategically. A racehorse, a dairy cow or a breeding bull all have identifiable revenue. A panda does not, and you will spend the whole answer fighting your own example.
   2. For a racehorse: expected prize money, weighted by the probability of winning at each grade, plus stud fees after retirement, which for a successful stallion dwarf the racing income.
   3. Costs: training fees, stabling, vet, insurance, jockey and entry fees. These are substantial and largely fixed, so most horses are value-destructive.
   4. Finite life with a terminal value: the residual breeding or resale value at the end of the racing career.
   5. Then the honest framing, which is the point of the question: the expected value is the probability-weighted average of a few enormous outcomes and many zeros. It is an option, not an annuity, so the way to value it is scenario-weighted, and the market price of a yearling at auction is your best cross-check.

   Where candidates lose it

   Freezing on the absurdity, or picking an animal with no cash flow and then trying to force a DCF onto it. Reframe the question as 'value any finite-life risky asset', choose an example that cooperates, and name your framework before you touch any number.

   Expect next

   • What is your personal beta?
   • How would you value a business with the same payoff shape?
   • How would you cross-check your number?

   Reported by candidates at Rothschild & Co (Generalist, New York, 2026). Source: Wall Street Oasis.

2. 082Why should I buy your college, and how much would you sell it for?Case and estimationHardsuperdayWMWellington ManagementInvestment Research · Boston · 2024WMWellington ManagementEquity Research · Boston · 2024

   Say this

   Pitch it as a subscription business with pricing power and a real estate portfolio attached. Revenue is tuition times enrolment plus research grants and endowment income; the assets are the campus and the brand.

   Then walk it

   1. The investment case: extremely sticky revenue, since a student enrolled is contracted for three or four years, pricing power that has historically exceeded inflation, and a brand that is effectively impossible to replicate.
   2. Revenue build: enrolment times net tuition after scholarships, plus housing and dining, plus research funding, plus endowment draw. Be explicit that gross tuition overstates it badly because of discounting.
   3. Cost base: mostly faculty and staff, largely fixed, which means high operating leverage in both directions. A 10 percent enrolment drop is devastating; a 10 percent rise is almost pure margin.
   4. Valuation on two bases and take the higher. As a going concern, a DCF or an EBITDA multiple on the operating surplus. As an asset play, the campus real estate plus the endowment, which for many institutions exceeds the operating value.
   5. Then the risks that make the price: demographic decline in the applicant pool, regulatory dependence on public funding and visa policy for international students, and the fact that you cannot actually cut faculty quickly. And I would flag that the brand is inseparable from the non-profit status, so a buyer might destroy the asset by acquiring it.

   Where candidates lose it

   Treating it as a whimsical question. It is a full valuation case wearing a joke. The two highest-value moves are separating gross from net tuition, and recognising that the real estate and endowment may be worth more than the operations.

   Expect next

   • How would you IPO it?
   • How would a college increase revenue?
   • What would you do in the first year as owner?

   Reported by candidates at Wellington Management (Investment Research, Boston, 2024); Wellington Management (Equity Research, Boston, 2024). Source: Wall Street Oasis.

3. 083How would a college increase its revenue?Case and estimationIntermediatetechnicalHWHarris WilliamsInvestment Banking · Richmond · 2018

   Say this

   Price, volume, mix, and new revenue lines. Raise net tuition by discounting less, grow enrolment, shift mix toward full-fee and postgraduate students, and monetise the assets that sit idle.

   Then walk it

   1. Price: the lever is usually the discount rate, not the headline tuition. Most institutions discount heavily; recovering a few points of net tuition is worth more than a sticker price rise and is less visible.
   2. Volume: more students, but constrained by capacity and by admissions standards, since taking weaker students damages the brand that supports the price.
   3. Mix is the highest-return lever. International and out-of-state students pay multiples of the domestic rate. Postgraduate and professional programmes carry better margins. Executive education has almost no marginal cost against existing faculty.
   4. New lines: online programmes that break the capacity constraint entirely, summer and short courses that use the campus in the off-season, conference and event hire, and licensing the brand.
   5. And the asset side: parking, retail on campus, research commercialisation and licensing, plus the fundraising engine, since alumni giving is a genuine revenue line that responds to investment.
   6. The reason mix beats price and volume: operating leverage. Faculty cost is already committed, so an incremental full-fee student in an existing class is almost entirely margin.

   Where candidates lose it

   Listing ideas without ranking them by margin impact. The interviewer wants commercial prioritisation. Naming operating leverage as the reason mix wins turns a brainstorm into an analysis.

   Expect next

   • Which would you do first?
   • What is the risk of the online strategy?
   • How would you value the business after those changes?

   Reported by candidates at Harris Williams (Investment Banking, Richmond, 2018). Source: Wall Street Oasis.

4. 084How would you value an insurance brokerage that operates in one country that has just had a coup and writes only one line of coverage?Case and estimationHardsuperdayPerella Weinberg PartnersFinancial Institutions Group · New York · 2026

   Say this

   Start from the normal brokerage framework, which is a commission stream on premium, then attack it with the two facts they gave you: extreme country risk and total product concentration. The answer is a wide range with a real chance of zero.

   Then walk it

   1. The base framework: a broker earns commission on premium and holds no underwriting risk, so it is a capital-light, high-margin, recurring revenue business that normally trades at a premium multiple on EBITDA.
   2. Now the coup. The currency may be unconvertible, so you may not be able to repatriate cash at all. That alone can make a profitable business worth little to a foreign buyer.
   3. Country risk enters the discount rate through a sovereign spread, and in a post-coup situation that could be well over 1,000 basis points. It also enters the cash flows, because premium volumes fall when economic activity stops.
   4. Single line of coverage means no diversification. If that line is motor and vehicle imports halt, or it is trade credit and trade stops, revenue can go to near zero. So I would model scenarios rather than a base case: functioning state, prolonged instability, and asset seizure.
   5. So: probability-weight the scenarios, discount at a rate that reflects the sovereign, and cross-check against what a local buyer would pay, because a domestic acquirer does not face the repatriation problem and will value it far higher than a foreign one.
   6. The honest conclusion is that the identity of the buyer determines the value here more than the cash flows do.

   Where candidates lose it

   Running a standard brokerage multiple and ignoring the two facts in the question. The coup and the single line are the question. And missing the repatriation point, which is the specific insight that makes the foreign buyer's value different from the local buyer's.

   Expect next

   • Who would actually buy it?
   • How would you size the country risk premium?
   • What if the currency is pegged but not convertible?

   Reported by candidates at Perella Weinberg Partners (Financial Institutions Group, New York, 2026). Source: Wall Street Oasis.

5. 085Given a B2B SaaS company with this EBITDA and this P/E, what would you do to improve its operations and financials?Case and estimationHardsuperdayHoulihan LokeyInvestment Banking · New York · 2026

   Say this

   Work the SaaS levers in order of value: pricing, then retention, then sales efficiency, then cost. In software, a point of net revenue retention is worth more than a point of cost saving, because it compounds.

   Then walk it

   1. Pricing first. Most B2B software is underpriced relative to the value it delivers. Move to value-based or usage-based pricing, introduce tiers, and raise prices on renewal for the existing base. This is near-pure margin.
   2. Retention second. Net revenue retention above 110 percent means the installed base grows without new sales. Reduce churn in the weakest cohort and upsell modules into the strongest. This changes the growth rate and therefore the multiple.
   3. Sales efficiency third. Look at customer acquisition cost payback and the magic number. If payback is over 24 months, the problem is targeting or pricing, not effort. Reallocate spend to the segments with the fastest payback.
   4. Cost fourth, and deliberately last. Consolidate the cloud bill, rationalise the product portfolio, offshore support engineering. Real money, but it does not change the growth story.
   5. Then the bolt-on question: in a fragmented software vertical, acquiring adjacent modules at a lower multiple and cross-selling them into your base is usually the single largest value-creation lever available.
   6. One flag on the question itself: P/E is an odd metric for a software company, since GAAP earnings are suppressed by growth spend and stock compensation. I would work off EV/ARR and EV/EBITDA instead, and I would say so.

   Where candidates lose it

   Jumping to cost cutting. In software, growth and retention drive the multiple, and the multiple drives the value far more than a margin point does. Also worth noticing that P/E is the wrong lens here; naming that is a real signal.

   Expect next

   • What is the formula for net revenue retention, gross retention and churn?
   • Which of those levers moves the multiple?
   • How would you verify the pipeline to forecast revenue?

   Reported by candidates at Houlihan Lokey (Investment Banking, New York, 2026). Source: Wall Street Oasis.

6. 086What are the formulas for net revenue retention, gross retention and churn?Case and estimationIntermediatetechnicalPiper SandlerInvestment Banking · Burlingame · 2026

   Say this

   All three measure the same cohort a year later. Gross retention counts only what you kept, capped at 100 percent. Net retention adds expansion, so it can exceed 100. Churn is the revenue you lost as a percentage of what you started with.

   Then walk it

   1. Gross revenue retention: starting recurring revenue from a cohort, less churn and downgrades, divided by starting revenue. Expansion is excluded, so it can never exceed 100 percent.
   2. Net revenue retention: starting revenue, less churn and downgrades, plus upsell and expansion, divided by starting revenue. Above 100 percent means the base grows by itself.
   3. Gross churn: revenue lost divided by starting revenue. It is one minus gross retention. Logo churn counts customers rather than revenue, and the two can diverge sharply if you lose many small accounts or one large one.
   4. The critical rule: neither retention metric includes revenue from new customers. Mixing new business into retention is the most common error and it flatters the number badly.
   5. Benchmarks worth knowing: best-in-class enterprise SaaS runs gross retention above 90 percent and net above 120. SMB software runs materially lower on both because small customers fail.
   6. Why it matters for valuation: net retention above 110 percent means the business compounds without selling, which is exactly what justifies a high revenue multiple.

   Where candidates lose it

   Including new customer revenue in the retention calculation. It is a cohort metric. And not knowing which one can exceed 100 percent, which immediately reveals whether you have actually used these numbers.

   Expect next

   • What is the rule of forty?
   • Which matters more for valuation, growth or retention?
   • Why can logo churn and revenue churn diverge?

   Reported by candidates at Piper Sandler (Investment Banking, Burlingame, 2026). Source: Wall Street Oasis.

7. 087How would you verify the validity of a client's sales pipeline in order to forecast revenue?Case and estimationHardsuperdayHWHarris WilliamsInvestment Banking · Richmond · 2025

   Say this

   Test it historically before you believe it prospectively. Take last year's pipeline, see what actually converted by stage, and apply those real conversion rates rather than management's assumed ones.

   Then walk it

   1. Back-test first. Pull the pipeline as it stood 12 months ago and compare it to what closed. If management said 60 percent of late-stage would convert and 30 percent did, you now have the real number and the size of their optimism.
   2. Test the stage definitions. A verbal indication is not a late-stage opportunity. Ask what evidence is required to move a deal between stages, and whether that discipline is enforced in the CRM.
   3. Check the vintage of each opportunity. Deals sitting in the pipeline for three times the average sales cycle are usually dead and not yet marked dead. They inflate the total.
   4. Check concentration. If three opportunities are half the pipeline, the forecast is not a probability distribution, it is three binary bets. Diligence those three individually and talk to those customers if the process allows.
   5. Cross-check against capacity. Does the forecast require more closed deals per salesperson than the team has ever achieved? And check whether headcount to deliver it is actually in the plan.
   6. Then rebuild the forecast bottom-up with your own conversion rates, and present it as a range against management's case. The gap between the two is one of the most valuable things you can hand a buyer.

   Where candidates lose it

   Accepting the pipeline and only sanity-checking the arithmetic. The technique is historical back-testing of conversion by stage. If you do not say that, you have not answered it.

   Expect next

   • What if they have no historical pipeline data?
   • How would that change your valuation?
   • What would you do if the top three opportunities were all with one customer?

   Reported by candidates at Harris Williams (Investment Banking, Richmond, 2025). Source: Wall Street Oasis.

8. 088Comparing two identical buildings, how would you value them differently?Case and estimationIntermediatetechnicalApollo Global ManagementReal Estate · Williamsport · 2022MSMorgan StanleyInvestment Banking · London · 2025

   Say this

   Identical bricks do not mean identical value. The difference is in the leases, the tenants and the debt. Value is net operating income divided by cap rate, and both terms can differ completely for the same building.

   Then walk it

   1. Net operating income first: what rent is actually contracted, at what escalations, with what vacancy and what recoveries of operating expenses. One building leased at above-market rent is worth more than its twin at below-market, today.
   2. Then lease duration and tenant credit. Ten years remaining to an investment grade tenant supports a much lower cap rate than two years remaining to a weak covenant. Duration and credit are the risk in real estate.
   3. Then the cap rate itself, which is where location micro-differences show up: the side of the street, the transport access, the parking, the floor plate efficiency.
   4. Then the debt in place. Assumable below-market fixed-rate debt is a real asset and can be worth several percent of the value. Expensive debt with prepayment penalties is a liability.
   5. Then everything outside the four walls: property tax assessment, ground lease versus freehold, capital expenditure deferred by one owner and not the other, and zoning or development rights above the building.
   6. So the short answer: I would value the cash flows and the risk of those cash flows, not the building. Two identical structures can easily differ 30 percent in value.

   Where candidates lose it

   Assuming the question is a trick with no answer, or listing only location. Leases and tenant credit are the substance. Naming assumable debt is the detail that marks out someone who has looked at real deals.

   Expect next

   • Walk me through getting to exit value from gross potential rent using a cap rate.
   • What is the cash-on-cash return at a given LTV and cap rate?
   • How does a cap rate relate to a multiple?

   Reported by candidates at Apollo Global Management (Real Estate, Williamsport, 2022); Morgan Stanley (Investment Banking, London, 2025). Source: Wall Street Oasis.

9. 089Given a 20x P/E, 10x EV/EBITDA, $20 of interest at a 5 percent rate, $200 of market cap and $20 of depreciation, calculate the tax rate.Case and estimationHardsuperdayEvercoreMergers and Acquisitions · San Francisco · 2026

   Say this

   Work backwards from the multiples to build the income statement. Net income is $10, debt is $400, so enterprise value is $600 and EBITDA is $60. EBIT is $40, pre-tax income is $20, so tax is $10 on $20 — a 50 percent rate. That is above any statutory rate, so I would flag that the inputs are inconsistent rather than just hand you the number.

   Then walk it

   1. Net income: market cap $200 at a 20 times P/E means net income of $10.
   2. Debt: $20 of interest at a 5 percent rate implies $400 of debt.
   3. Enterprise value: $200 equity plus $400 debt is $600, assuming no cash. At 10 times EV/EBITDA, EBITDA is $60.
   4. Then EBIT is EBITDA less depreciation, so $60 less $20 is $40. Pre-tax income is EBIT less interest, so $40 less $20 is $20. Net income is $10, so tax is $10 on $20, which is a 50 percent rate.
   5. I would then flag it: 50 percent is above any statutory rate, which usually means the question contains rounded or inconsistent inputs, or there is a non-operating item I am not being told about.
   6. The way to handle this live is to lay out the chain clearly, state the answer the arithmetic gives, and then say what would make it plausible: cash on the balance sheet would lower enterprise value and therefore EBITDA, and minority interest or a one-off charge below the line would change the bridge.

   Where candidates lose it

   Either freezing on the arithmetic or reporting an absurd tax rate with a straight face. The test is whether you can chain multiples backwards into an income statement, and whether you have the judgement to flag an implausible output rather than just handing it over.

   Expect next

   • What if there were cash on the balance sheet?
   • Which assumption are you least comfortable with?
   • Redo it with $100 of cash.

   Reported by candidates at Evercore (Mergers and Acquisitions, San Francisco, 2026). Source: Wall Street Oasis.

10. 090You have a rope 30cm long and another 45cm long. Each centimetre takes one minute to burn. How do you measure exactly 25 minutes?BrainteasersHardtechnicalMoelis & CompanyInvestment Banking · New York · 2025

    Say this

    The only tool you have is that lighting a rope at both ends halves its time. With these two lengths the times you can construct are 15, 22.5, 26.25, 30 and 37.5 minutes — 25 is not one of them. I would show you that working and tell you the puzzle as stated has no exact solution.

    Then walk it

    1. The single insight: a rope lit at both ends burns out in half its length in minutes, regardless of where it burns unevenly.
    2. Light the 30cm at both ends and it is gone at 15 minutes. Light the 45cm at both ends and it is gone at 22.5.
    3. Best combination for something near 25: light the 45cm at both ends and the 30cm at one end at time zero. At 22.5 minutes the 45 is gone and the 30 has 7.5cm left. Light its second end and it burns out 3.75 minutes later, at 26.25.
    4. The other combination goes the wrong way: light the 30cm at both ends and the 45cm at one end. At 15 minutes the 45 has 30cm left; lighting its second end gives 15 more minutes, so you land on 30.
    5. So the reachable set is 15, 22.5, 26.25, 30, 37.5 and the un-halved 45. There is no route to exactly 25.
    6. In the room, say the halving principle immediately, construct the timings out loud, then say plainly that 25 is not reachable and 26.25 is the closest. Interviewers garble the lengths on this puzzle constantly, and catching that scores better than forcing a wrong answer.

    Where candidates lose it

    Forcing an answer because you assume the question must have one. Narrate the halving principle, build the reachable times, and say if the target is not among them. Confidently asserting a solution that your own arithmetic contradicts is the actual failure here.

    Expect next

    • What is the general set of times you can construct?
    • How much water can you measure with a 3-litre and a 4-litre bottle?
    • What angle do the clock hands make at 3:15?

    Reported by candidates at Moelis & Company (Investment Banking, New York, 2025). 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.