4 ms·
To back up numair, and help you understand a bit better, an audit in the financial regulatory sense is a highly pedantic endeavor. Accounting can be as much art
by salawat 5y ago
To back up numair, and help you understand a bit better, an audit in the financial regulatory sense is a highly pedantic endeavor. Accounting can be as much artform as anything else, and even amongst all the accountant's in the world, exactly how you track money and classify things is open to interpretation, and it is very easy to report some cherry picked numbers and tell a completely different story than what is going on.
Independent audits do two things. They examine process, and create an environment where you have to operate in a sane manner, because your auditor can drop in at any time, pick out any particular starting point, and will expect to be able to have delivered to them where that transaction came from, and ultimately will go to based on GAAP and in house supporting process documents. This basically draws a complexity boundary around how exotic you can get within the context of one organization, because if one of the big independent auditors can't make heads or tails of you in a reasonable amount of time, it is a gigantic red flag w.r.t your operational processes.
An audit also results in a snapshot of your entire cash/value flow through an organization. This is verified and cross checked for valuation by someone who doesn't know you from Adam to ensure objectivity. This is just a guarantee that the numbers add up, and over time continue to make sense. Massive discontinuities that can't be ascribed to something in the real/business world measurable by someone else are also red flags that there may be something going on there that may be valid, but you need more info to get to the bottom of it to ensure it is sound.
t. Quality Assurance person who has spent entirely too much of his life digging into how finance works even though they allegedly hate it, but a statistical analysis of how much of my mind I devote to financial analysis, modeling and prediction tells another story.
Ironically, if you mentally audited my thoughts you'd easily come to the conclusion I love finance. I don't. I like measuring things. Measuring finances effectively is a pain in the ass, and a perennial issue, that there are more than a few groups constantly working to frustrate people like me who try to distill truth out of account ledgers.
That's why numair is absolutely right. An attestation is one slice. You know money was there. You don't know where it went afterward, which subtransactions it spawned (fees, taxes, interest accrual, etc...). That's what auditor's look at and collate. It's why it's a big deal. At the end of the day it's all arithmetic, and an audit is just having someone else run the numbers and vouching they get the same result. It's actually a little more than that, because regulation wise, the auditor's number is more reliable than yours, because they have all the incentives in place to keep their processes and interpretive liberty taking to a minimum. You want to match the auditor's numbers if you can coax your process into doing so... Of course again, there's arms race there as well. Note, auditor's don't find or investigate fraud. They just check your process is being followed, and that when someone else does it, everything works out. It just so happens that this is also a great way to shake out anything that might be used to as a basis to make fraud effectively doable.
It is a rabbit hole. One that I've explored many branches of, but after a while doing it, you can pretty quickly measure how uncomfortable a group is by their level of nervousness around letting an auditor drive.
- numair 5y agoPlease turn this comment into a blog post or whitepaper or something, because it’s excellent and deserves way more visibility than it’s going to get 4-deep in a weekend comment thread on HN. I think we will all have to take on the task of educating a lot of kids — literally, kids — about the boring-seeming world of financial auditing. They’ll be wondering where their money went, and how to avoid such situations the next time around...
- salawat 5y agoOh God... All the bibliography building that'd go into a white paper... Ugh. I might try spinning up a Blog one of these days when work slows down a bit. It's funny the insight you pick up having to dissect these types of things on a semi-regular basis. Financial Auditing, Risk Management, Quality Assurance, we're all doing the same schtick in different ways. Also, I'd add that one needs to keep in mind that what I posted only covers audits of singular organizations. If you're actually up to skulduggery, as I understand it, it's almost always distributing things between multiple corporate entities, audited by different groups so that no one auditor gets a complete view of what's going on. Once you start looking at groups of entities, especially jurisdictionally distributed, the complexity balloons, and you're now firmly in financial engineering and forensic accountancy territory. Financial engineering and forensic accountancy is a rabbit hole that I'm pretty sure encyclopedias could be written on, and would outdate themselves as quickly as you could propagate the info, as there is always an arms race going on between those wishing to move large amounts of money gained in less scrupulous ways, and those wishing to make those enterprises an impossibility.
- numair 5y agoSend me an email — numair@numair.com — I’d like to know more about what led you to do a deep-dive into these worlds, as I am sure it’s a long and interesting story!
- anon9001 5y agoVery interesting and I agree that a blog post would be great on this topic. I think intuition is what I'm lacking here. It's unclear to me just how dangerous the situation is likely to be. Before this discussion, my understanding was that only attestations are available, and that the company wouldn't become insolvent because the company would incur huge losses if that happened. That's still my understanding, though now I know the difference be attestation and audits, that a complexity boundary exists due to audit resources, and that auditors don't even look for fraud (you'd think they would without knowing how this works). Thank you for the color around it, it helps me understand how people in this business think about finance. However, I'm still of the (apparently wrong?) opinion that Gemini/Circle/Binance/Paxos at least, are staking their real businesses, with things like executive pay and employee payroll and all that, on these products. "We lost the money and our stablecoin is now worthless" should sink any of these companies. I think we're in agreement that failure to maintain peg would sink these companies, but my conclusion was that it's very unlikely. You and numair seem to conclude that it's more than just likely, it's almost inevitable. This tells me that my intuition is probably wrong, but it's still difficult for me to understand why a company like Gemini would decide to risk insolvency. And even if they did release a full audit, if it can't detect fraud, then why would you trust it? I know it's better than just attestations, but I'm not sure how great the risk of fraud is relative to the risk of other routes to insolvency.