Clarify the assignment before beginning
A public-markets case can be open-ended or tightly specified. Confirm the security, recommendation type, time horizon, information cutoff, required model, page limit, presentation time, and whether external data or AI tools are permitted. The best analysis can be judged incomplete if it ignores the requested output.
Find the security debate
The case shouldn't become a company encyclopedia. Identify the variables that determine the security’s value: demand durability, customer retention, pricing, margin structure, capital intensity, credit risk, regulation, capital allocation, or an event. State what the price or consensus appears to assume and where the analysis may differ.
Build the source file
Use filings, transcripts, competitors, regulators, industry data, legal documents, and relevant public sources. Record the date, definition, and limitation of each important input. Separate facts, estimates, and judgment in the model and memo.
Model only what changes the answer
A model should connect the thesis to earnings, cash flow, balance sheet, or recovery. It doesn't need every historical line if those lines don't inform the debate. Include scenarios that show the range of outcomes and the assumptions responsible for it.
The written case
A concise case usually contains:
- recommendation, security, price date, and horizon;
- market expectation and variant view;
- two to four thesis points;
- estimate differences and valuation;
- catalyst path;
- downside and falsifiers;
- position or risk considerations;
- source and information cutoff.
The appendix can hold detailed model outputs, industry work, and source notes.
The live defense
Expect questions about the strongest contrary evidence, why the market is wrong, what happens at another price, what can delay the catalyst, and which fact would reverse the recommendation. The candidate should be willing to update in real time rather than defend the initial document mechanically.
Common case failures
Cases fail when they summarize the company without identifying the expectation gap, use a target price without showing the method, treat management guidance as fact, ignore the balance sheet or dilution, provide risks without monitoring indicators, or use a model whose detail can't be explained.
A clear “no position” conclusion can be strong if the analysis shows why the expected return, evidence quality, or downside doesn't justify capital.