Working Through the Berk and DeMarzo Data Cases
The data case solutions for the Berk and DeMarzo Corporate Finance textbook are essentially spreadsheet-based exercises tied to each chapter. You open the dataset, follow the problem set, and build out the financial model they're asking for. Most of the work revolves around NPV, WACC, capital budgeting, and valuation scenarios. It sounds straightforward until you actually sit down with the files and start plugging numbers in. The official datasets live on the Pearson website if your instructor gave you a course code or ISBN access. Otherwise, many students end up on various file-sharing forums or study sites. The solution sets you find floating around the internet are usually spreadsheets with the answers already filled in, but the real value comes from understanding why each cell is calculated the way it is. I'd recommend working through the problem yourself first before looking at any completed model. You'll learn significantly more that way. I ran into a specific issue once when working through the bond valuation data case. The Excel file had circular references built into the coupon payment cells because the textbook's original spreadsheet wasn't properly structured for iterative calculations. I ended up getting a circular reference warning and the solver wouldn't converge on the yield to maturity. What I did was break the loop by separating the cash flow inputs from the discount rate calculation into two sheets, then linking them with a simple reference. That took about twenty minutes and saved me from debugging the whole model from scratch. It's a common enough problem across multiple cases in the book.
Understanding How the Cases Actually Work
Each data case comes with a raw dataset and a question sheet. The dataset is usually in CSV or Excel format containing historical price data, financial statements, or market variables. You're expected to clean it, structure it, and produce the requested outputs. The cases aren't designed to be copy-paste assignments. They're meant to simulate real analyst work where the data doesn't come pre-packaged neatly. Here's something most students miss: the Berk and DeMarzo cases frequently use actual Compustat or CRSP data pulled directly from the market. That means the numbers aren't clean classroom numbers. You'll encounter missing values, adjusted prices that don't reconcile with raw prices, and corporate actions like stock splits that change the time series mid-stream. If you just run a formula without checking the data integrity first, your results will look correct but be wrong. I've seen this happen repeatedly with the cases involving merger arbitrage and option valuation where the adjustment factor for splits was buried in a footnote of the dataset description. Always verify the raw data against a secondary source before building your model. Another counter-intuitive point about these cases is how sensitive the WACC calculations are to the equity risk premium assumption. The textbook uses a standard value around 5 to 6 percent for US equities, but small changes in that input create massive differences in your final NPV. A one percentage point shift in the ERP can swing a project's NPV by tens of millions on a large-cap valuation. Students often treat the ERP as a fixed constant without thinking about whether it's appropriate for the specific firm or sector in the case. It's worth questioning that assumption even if the problem doesn't explicitly ask you to.
Practical Steps for Completing a Data Case
Start by opening both the problem instructions and the raw dataset simultaneously. Read through all the questions first before touching any formulas. You'll often find that question three depends on an output from question one, so knowing the full scope helps you structure the spreadsheet logically. Build your model in stages. Get the basic cash flow table working and verified against a simple manual calculation before adding any advanced functions. Once that checks out, layer in the NPV and IRR calculations. Then move to sensitivity analysis if the case requires it. I typically spend about an hour on a standard data case doing nothing but verifying the raw data makes sense. It's not glamorous but it prevents hours of rework later when your final answer is off by a rounding error or a misplaced decimal. The textbook cases also tend to test your ability to handle real-world complications like taxes, depreciation schedules, and working capital changes. These aren't trivial add-ons. A common mistake I see is students forgetting to include the tax shield from depreciation when calculating free cash flow for a capital budgeting case. The difference between including and excluding that line item can be substantial, especially in the later years of a project where depreciation creates a meaningful cash flow benefit. Make sure every cash flow component is accounted for explicitly.
Get the Full Details
When the Standard Approach Falls Apart
Not every case fits neatly into the template Berk and DeMarzo expect. Some of the more advanced cases involve non-linear payoff structures or real options that standard spreadsheet functions can't handle directly. For those situations, you may need to use Solver or a manual iteration approach. I've had cases where the textbook's recommended method produced a wildly incorrect result because the underlying assumptions didn't hold for the specific dataset. In one instance, the risk-neutral valuation approach broke down when the volatility input was derived from a very short historical window. The resulting option value was essentially meaningless. Switching to a binomial tree with a larger number of steps gave a more reasonable answer and matched the intuition better. It took longer to set up but the result was defensible. The biggest limitation of relying on pre-made solution sets is that they don't teach you how to think through messy problems. If you're using these cases for a class, your professor will likely adjust the parameters or ask follow-up questions that require you to understand the mechanics, not just reproduce a spreadsheet. The solutions you find online can be a useful reference point, but they're not a substitute for actually working through the logic yourself. Budget roughly two to three hours for a standard data case if you're being thorough, though experienced students with a well-organized approach can often complete one in under an hour. If you get stuck on a particular case, the best approach is to revisit the relevant chapter sections and trace each formula back to its theoretical origin. The Berk and DeMarzo text is generally rigorous about explaining the derivation, and understanding why a formula exists makes it easier to adapt when the standard application doesn't work. I still refer to the textbook even after completing dozens of cases because new edge cases keep coming up that remind you how much nuance is built into these models.
