Tips for Assessing Level of Modelling Granularity in a Case Study?
I appreciate that the true depth you can go into on a stock can far exceed what you can reasonably accomplish when doing say a 4 hour hedge fund case study, but I have struggled with finding the appropriate detail to go into while maintaining sufficient time to find management guidance and build a variant view.
Take a company like H&M for example with a bunch of different sub-brands. Would you only granularly model the sub-brands that compose 80% of the revenues? Or just key brand?
Or do you look to understand what granularity guidance is provided for and use that to tell?
Before you start I would ask if they’d provide any guidance on time allocation or the goal of the case. Is this a quasi modeling test? If so spend more time there. I’d this more of a “run a quick screen” if so then it’s probably okay to have a very basic model (rev, gp, ebit, eps) to give you more time to think through the qualitative thesis.
Regardless, in your write up, def include a section of “near term work plan” that goes thru where you’d focus if you did have more time
That said on a broad model strategy
Def start with what they actually guide and go from there. A lot of companies provide further guidance color on calls but you aren’t going to have time to parse all that out.
To the extent you have time, I’d probably focus more on revenue drivers (likely more disclosure here) and keep the cost build simple (eg incremental gross margins and scale on sg&a.
Last thing, make a call one way or the other with differentiated numbers. If you know the funds style / 13f you can probably get a sense of where they’d lean and id prob pitch it in that direction (unless you are very convicted and already know the company well)
Adding to this - figuring out where to allocate time is part of the case study test. You need to figure out what level of granularity 1) needed 2) sufficient enough for your thesis and 3) good enough for you to defend during the walkthrough.
Also sometimes simplicity is key. Any model that requires you to rely on 10 levels of assumptions to get to a sensible number than on a tangible data point is probably worthless anyways and opens up a can of worms for the interviewer.
It depends on the information they company provides and why you're trying to get granular.
I had a quick look at H&M financials (granted on 30 seconds) and I all can see is geographic revenue. But they also provide the number of stores for the corresponding regions. Could avg sales per store per country be a good measure?
You need to ask yourself why you're trying to be granular. Just because a model has more lines doesn't make it better. You should only go granular if it adds conviction to your thesis. What's better, making an assumption about 1 variable or making assumptions about 2/3 etc variables?
If you flip like two pages further on their reports they breakdown stores / new stores by brand.
My thinking is that the different group brands have very different brand recognition, pricing power, growth, target audience etc. The H&M and COS brands account for 95% of the Group's total stores, so strike me as being what would be required to model in a 4 hour case study, lets say. There aren't that many new stores / decline in stores amongst the other sub-brands.
Given the above, would your approach in a four house pod-style case also be to focus on these two sub-brands?
Est atque odio facilis. Voluptas quam veniam qui sint qui veniam. Voluptatibus ducimus officia qui.
A veritatis illum impedit repellat. Labore tenetur vel quo sint. Animi sunt aut tempore vero eius et. Quis quia officiis voluptates laboriosam voluptates repudiandae vero. Similique voluptatem dolor molestiae quod cum consequuntur. Sint soluta molestias quia.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...