Timed Exercises

A timed modeling test typically asks a candidate to build a simplified LBO or DCF from a blank spreadsheet within a set time limit (commonly 30-90 minutes), given a short prompt with the key assumptions (purchase price, growth rates, leverage, and so on). It's testing raw Excel fluency and modeling muscle memory under real time pressure, can the candidate build a clean, correctly linked model fast, without needing to look up formulas or structure, more than any creative or strategic judgment.

This is exactly why the paper LBO and DCF walkthrough preparation covered elsewhere directly transfers here: the conceptual shape of the model needs to already be second nature, so the time pressure of the exercise is spent executing in Excel, not re-deriving the structure from scratch.

The most effective preparation is repetition: building the same core models (a simplified 3-statement model, LBO, DCF) repeatedly from a blank sheet, under a self-imposed timer, until the structure, formulas, and formatting habits become close to automatic. Pre-built templates and keyboard-shortcut fluency (covered in the Excel lessons) both compound directly here, every second spent navigating the mouse or looking up a formula is a second not spent on the actual analysis.

Practicing with a genuinely blank sheet (rather than only ever editing a pre-built template) matters specifically because the real test usually starts from nothing, and building the structural instinct, where revenue projections go, how the debt schedule links to the cash flow available for paydown, is exactly the skill a template can mask if that's all a candidate has ever practiced with.

A candidate under time pressure should prioritize getting a complete, correctly linked model built first, then refine, a fully built but slightly rough model generally scores better than a beautifully formatted model that's only half-finished when time runs out, since a broken link or incomplete structure means the exercise can't be properly evaluated at all. Sanity-checking outputs along the way (does this IRR look plausible, does the balance sheet actually balance) catches errors early, when they're still cheap to fix, rather than at the very end when there's no time left to trace back through the model.

Staying calm and working methodically through the buildup, resisting the urge to skip steps or rush key formulas, is itself part of what's being observed, since it's a reasonable proxy for how the candidate will behave building a real model under a live deadline on the job.

Brainteasers and mental math exercises are used less to test any specific banking knowledge and more to observe how a candidate thinks under pressure, do they work through an unfamiliar problem methodically and out loud, or freeze; can they do reasonably quick approximate arithmetic without a calculator, a real, if minor, part of client-facing and live-negotiation work where a rough number is sometimes needed on the spot.

Not every firm or interviewer uses these anymore (some firms have moved away from them), but they remain common enough that basic preparation is worthwhile, and the underlying skill they test, calm, structured thinking under time pressure, genuinely does transfer to the job.

For most brainteasers, especially estimation problems ('how many gas stations are in the US'), interviewers are far more interested in a structured, logical approach, breaking the problem into reasonable, estimable pieces and reasoning through each one out loud, than in whether the final number happens to be precisely correct. Silently working out an answer and only stating the final number misses the entire point of the exercise, since the reasoning process is what's actually being evaluated.

Stating assumptions explicitly as you go ('I'll assume roughly X million households, each needing gas roughly once a week...') both structures your own thinking and gives the interviewer visibility into exactly how you're reasoning, which is the behavior the exercise is designed to surface.

Quick, approximate mental math, percentage changes, multiplication and division of multi-digit numbers, quick estimation, is worth practicing deliberately in the weeks before interviews, since it's a mechanical skill that improves with repetition and often gets rusty without regular use, especially for anyone relying heavily on a calculator or Excel day to day. Simple daily practice (mental math apps, or just estimating totals while shopping or reading news with numbers in it) builds the speed and comfort needed to do it fluently, out loud, under mild interview pressure.

The goal isn't calculator-level precision; it's confident, reasonably accurate estimation delivered smoothly, a candidate who visibly struggles or goes silent for a long stretch on simple arithmetic creates a worse impression than one who states a clearly-labeled rough estimate quickly and moves on.

Case interviews are more common at boutiques and elite advisory firms than at bulge brackets, and typically present a business scenario, should a company make an acquisition, how should it value a target, what's driving a company's performance issue, followed by a structured discussion where the candidate builds and defends an analytical framework, sometimes with basic supporting calculations, over 20-45 minutes. It's less standardized and less purely quantitative than a consulting case, and more closely tied to the technical valuation and deal concepts covered throughout this curriculum.

Because every firm's specific case style differs, the best preparation is being extremely comfortable with the underlying technical material (valuation methods, deal rationale, basic modeling) rather than memorizing case frameworks from a different industry's interview style.

A strong case response starts by clarifying the actual question and laying out a clear structure before diving into analysis, for example, for an acquisition case: strategic rationale, valuation, financing and deal structure, and key risks, narrating the structure out loud so the interviewer can follow the logic and redirect if needed, rather than working silently and presenting only a final answer.

This is the same skill tested throughout technical interview preparation more broadly: reasoning out loud, in a clear logical sequence, using the frameworks (valuation methods, synergy analysis, financing tradeoffs) covered elsewhere in this curriculum, applied here to a less scripted, more open-ended prompt.

IB case prompts are often intentionally under-specified, without every number needed for a precise calculation, testing whether the candidate can identify what additional information they'd want, make reasonable stated assumptions to fill the gaps, and still reach a defensible conclusion, rather than stalling because the prompt isn't fully complete. Explicitly naming an assumption ('I'll assume a 20% tax rate since none was given') and moving forward is the expected behavior, not a workaround.

This mirrors real deal work closely: bankers constantly work with incomplete information under time pressure and are expected to make sound, clearly-stated judgment calls to keep moving rather than waiting for perfect data, which is exactly the instinct a well-handled case interview is meant to surface.