Investment Banking Case Studies: Spring Week vs Analyst vs Associate

The four exercises, at a glance

Work out which of these you are sitting before you prepare for any of them. Almost all wasted preparation at this stage comes from candidates reading private equity case study material — which dominates search results because it is the most dramatic version — when they are actually sitting a group discussion.

• Spring week group case: 45-60 minutes prep, 10-minute group presentation, no model, behavioural scoring. First-year students.
• Summer internship / full-time analyst case: 1-3 hours, usually written or a short valuation exercise, sometimes with a group element. Penultimate and final-year students.
• Associate / lateral modelling case: 3-6 hours or a take-home, a full LBO or merger model plus a recommendation. Experienced hires and MBA candidates.
• Private equity case: 2-6 hours in-office or 24-72 hours take-home, an LBO model and investment recommendation presented to the deal team. Banking analysts recruiting to the buyside.

The spring week group case study

Four to six candidates are given a short business brief — a company considering an acquisition, a client weighing a financing decision, occasionally a live news injection partway through. You get roughly 45-60 minutes to prepare together and about ten minutes to present a recommendation to a panel, followed by questions.

Nothing is modelled and no finance background is assumed. The assessors are scoring group behaviour, and they are scoring it on a rubric: does this person build on others, do they manage time, do they bring in the quiet candidate, can they be disagreed with without becoming defensive. Technical correctness barely features, because at first-year level there is no technical work to be correct about.

The failure mode here is counterintuitive and extremely common. Candidates arrive believing they must stand out, so they talk over people and drive every point. Assessors read that as poor collaboration, and it scores below the candidate who spoke less and structured better. Taking a visibly useful role — timekeeper, synthesiser, the person who writes the group’s structure on the flipchart — outperforms trying to dominate.

The analyst case: summer internship and full-time

By penultimate year the expectations change. The exercise is usually individual and written: a short company overview and a set of financials, one to three hours, and a recommendation — sometimes a rough valuation, sometimes a strategic judgement supported by numbers you have pulled from the pack.

What is scored shifts from behaviour to judgement. Assessors want to see that you can pick the three numbers that matter out of a fifteen-page pack, state a clear view, and defend it under questioning without either collapsing or digging in. A defensible wrong answer beats a hedged non-answer at this stage, every time.

• Lead with the recommendation, then support it. Do not build to a conclusion — the reader may stop after your first paragraph.
• State your assumptions explicitly. An assumption you have flagged is judgement; the same assumption unflagged is an error.
• Get the arithmetic right on the two or three numbers you actually cite. Nothing costs credibility faster than a headline figure that does not tie.
• Leave time to sanity-check. A valuation you cannot explain in one sentence is one you should not present.

The associate and lateral modelling case

At associate level the exercise becomes a genuine modelling test: three to six hours in office, or a take-home over a day or two, building a full LBO or merger model from a set of assumptions and presenting the output. Here technical correctness is the assessment, not a side condition of it.

The differentiator is rarely model mechanics — most candidates who reach this stage can build the schedule. It is whether the model answers a question. A returns bridge that shows where the IRR actually comes from, a sensitivity table that isolates the assumption the deal turns on, and a one-page recommendation that a partner could read in the lift: those separate offers from near-misses.

The private equity case study

This is the exercise most search results are describing, and it belongs to buyside recruiting rather than banking. A real or disguised company, a data pack, two to six hours in-office or 24-72 hours as a take-home. You build an LBO, form a view, and present an investment recommendation to the deal team.

The scoring is investment judgement first and modelling second. A perfectly built model with no view fails. What the deal team wants is a decision, the two or three things the decision depends on, and an honest account of what would have to be true for you to be wrong.

How to tell which one you are sitting

The invitation email usually tells you, if indirectly. Duration is the single most reliable signal, and group format is the second.

• Under an hour of preparation, and the word "group" appears → spring week group case. Prepare structure and group behaviour, not valuation.
• One to three hours, individual, financial statements attached → analyst written case. Prepare a clear recommendation format and quick, correct arithmetic.
• Three hours or more, or a take-home with a deadline → modelling case. Prepare to build from a blank sheet, including the debt schedule.
• The words "investment recommendation" alongside a company data pack → buyside case. Prepare to have and defend a view, not just a model.
• If the email genuinely does not say, ask the recruiter. Nobody has ever been penalised for asking what format an exercise takes, and turning up prepared for the wrong one is a real risk.