4 ms·
I used to professionally audit spreadsheets that were used to back up million dollar decisions. I say “back up” because it was fairly clear the decisions were m
by moomin 6y ago
I used to professionally audit spreadsheets that were used to back up million dollar decisions. I say “back up” because it was fairly clear the decisions were made independently. But they needed to tick a box saying they’d had it audited so they paid us for a two day audit. In that limited time, which included writing the report, you’ll appreciate we rarely had time to audit anything more than the top-level summary sheet.
There was always at least one mistake that put the numbers out by 8 digits.
- imperfectcats 6y agoIs your job to help the justify the justification, or to really fix things?
- thaumasiotes 6y agoI assume you fix the math errors and then whatever the result is turns out to be the justification for the decision. The result doesn't matter, so you don't need to aim at a particular one.
- moomin 6y agoNeither. You report what you found. You make recommendations. You state which areas you haven't had the time to look at. Under no circumstances do you provide any guarantees that might put you at the wrong end of a multi-million dollar lawsuit. This, incidentally, is every audit ever. I've been audited many times, every audit finds stuff and make recommendations. Then management gets to pick which recommendations they adopt and which they'll derogate. I'm not saying it's a completely useless process, but it doesn't provide the cast-iron guarantees people think it does. Then, of course, there's financial audits, which are meant to provide a cast-iron guarantee the firm is still a viable entity. But nearly every big firm that went bankrupt had passed such an audit in the previous year.
- xivzgrev 6y agoI had no idea this was a thing, but makes total sense!
- conductr 6y agoQOE drove me nuts in a past life. M&A of small companies with crappy financial data. Biggest caveat is basically garbage in and garbage out. But often they already had the deal done and would just feel better if I told them the garbage didn’t stink.
- moomin 6y agoOne of my best behaviours for actually writing one of these things is to have a single page that only has inputs/assumptions on it and put an annotation/note on every last data item/series saying where you got it from, because someone on the warpath will want to know where you got that BS number from two months later, and it's very satisfying to show it was them.
- conductr 6y agoAgree. That’s a must. You have to cover your ass when it comes to this type of work. It’s high stakes enough that I save off copies of emails (.msg) and create a Sources folder for these, spreadsheets, and I’ll even do a screen capture if I have to login to some system and pull the data myself. I am usually given the parameters on what to pull, and it’s very easy for someone to accuse me of pulling the data wrong a month or more after the fact. I did this on every analysis. I can relate to what your saying. Two months later something like this always happens; What was you assumption of growth? Infinity. What?! That’s ridiculous why would you assume that? forward back an email you sent to me when I pointed out the growth rate seemed high but you insisted it was a slam dunk (aka the only way the deal would go through and we’d get that bonus) At times I feel like I put so many footnotes in the resulting presentation later the information should be interpreted as meaningless.