Modeling Test - Quarterly

First off, you will have everything you need. Consensus figures, ‘house estimates’ of any, press release, etc. I’ve worked for some fairly psychotic people over the years and even they all provided this to prospective hires.

From what I can gather, it seems you don’t have any work experience in this field, which is fine, but you’re going to have to do a bit of extra leg work - e.g. not knowing what a first read note is.

I described above in multiple ways because every firm calls it something different, but it’s basically as you say, a note that talks through the main things that happened in a quarter. Every analyst is pretty delusional in thinking their note adds any value if they’re not literally first to publish. The first bullet is always something about adjusted EPS/EBITDA vs consensus. Here you will literally say, company XX delivered EPS of YY, vs cons. of ZZ. The main drivers were a beat/miss on XYZ. Second through fourth bullet (depending on size of note) talk about whatever nonsense management is peddling and other interesting stuff. Could be volumes, operating leverage, margins, Trump, anything. This is your opportunity as an analyst to add value by noticing interesting stuff or calling out oddities. 95% of the time, there isn’t anything interesting. Your last bullet in the business will always be some rehashing of your thesis like “this reinforces our buy/sell rating for all the aforementioned reasons” (note: regardless of how good/bad a quarter is, you always claim victory because you’re a sell side monkey). And you provide the dial information (time, phone number, conference code).

As for your modeling test, they may have you update one of their models for a quarter by giving you an incomplete model and a press release - the name of the game is accuracy. Your forecasts aren’t worth squat to them because you seem fairly green. Make absolutely sure your EPS and/or Adj EPS ties out to what is reported. This is an auto ding if you fuck it up (as are any other important metrics like revenue or CFO). Be fast, methodical, and don’t fuck up.

They can also make you build a model from scratch. This seems less likely based on my evaluation ofyou as someone with relatively little experience, but to prep, you should build from scratch a 3 stmt model and make sure your B/S and C/F tie out and that your I/S stuff is flowing properly into the C/F. Sky’s the limit here in terms of variability - I’ve seen kids have to build two models and then a merger overlay after we reviewed the two models; dummy models with dummy information and you having to be able to do some quick and dirty analysis on what the trends mean; normal modeling with real companies; literally anything. Since you seem junior, just make sure your shit ties out and your numbers are accurate agains the source material provided - again, your forecasts don’t really matter so long as they’re not wildly stupid.