Slashdot Mirror
Oracle Whistleblower Suit Raises Questions Over Cloud Accounting (nbcnews.com)
Svetlana Blackburn, a former senior finance manager for Oracle claims that the company has fired her for not "inflating" revenues in its cloud services division. She alleges that her bosses had instructed her to add "millions of dollars of accruals" for expected business "with no concrete or foreseeable billing to support the numbers." Oracle eventually inflated the numbers without her assistance, anyway, she adds. From NBC News report: The lawsuit, filed on Wednesday in U.S. District Court in San Francisco by former Oracle senior finance manager Svetlana Blackburn, also revives longstanding questions about proper accounting when software and computer services are bought on a subscription basis rather than as a single package, analysts said. Those questions are becoming more urgent as companies including Oracle, IBM, Microsoft and SAP race to transform their businesses for an era in which customers no longer own and operate their own information technology systems and instead lease computing services and software from cloud vendors using vast data centers.A spokesperson for Oracle says that Blackburn's claims are wrong, adding, "We are confident that all our cloud accounting is proper and correct."
"We are confident that all our cloud accounting is proper and correct." == "We are confident we paid off the right politicians to get nothing beyond a slap on the wrist for our doctored books."
Oracle Overcharges!
News at 11.
Is it Oracle, for allegedly inflating their sales numbers? Or is it the accounting firms that audit the books and sign off on it? I have no horse in the race, just an honest question.
"We are confident that all our cloud accounting is proper and correct." == "We are confident we paid off the right politicians to get nothing beyond a slap on the wrist for our doctored books."
I'm a certified accountant (among other things). What it probably really means is that the rules are sufficiently poorly defined that Oracle figures they have enough weasel room to claim what they are doing is permissible. And they may be correct in that assertion. It also probably means that what they are doing won't result in any issue worse than a fine and probably will have minimal impact on their stock price or bottom line long term. I don't have any idea if Oracle is doing something illegal here or not but I'm not surprised they would make a statement like that. Could result in some fines after years of litigation but probably won't amount to anything material for them even if it should.
One of the problems (especially with intangible goods) is consistently determining when to book the goods as a sale. This often isn't as straightforward as you might think and there are often several potential acceptable answers. The important thing is that A) what method they are using is understood and made clear and B) that they follow it consistently. For example my company books a sale when we receive a purchase order from a customer. A cloud software company might choose to book the sale based on expected retention rates and then adjust the numbers later for bad debts based on actual cash received. There are accounting guidelines for what is generally considered acceptable and what isn't through the FASB, the IRS and some other government agencies like the SEC. Obviously making up sales out of whole cloth is clearly illegal but if there is anything backing up the numbers at all then it can get a lot fuzzier real fast.
Is it Oracle, for allegedly inflating their sales numbers? Or is it the accounting firms that audit the books and sign off on it?
The answer is Yes. As in both of them are at fault. If the auditing firm misses something big like that then they are potentially liable if Oracle really is inflating their numbers. Also the company management bears substantial responsibility as does the board of directors.
One thing to bear in mind however is that when an auditor signs off on the books they are NOT claiming that the books are absolutely correct to the penny. That would be impossible for a company of any significant size. What they are saying is that they believe the books are materially representative of the financial picture of the company. It would be impossible for them to claim that there were no errors or flaws in the books. Basically the best they can do is say "yeah, this is pretty close to reality and you can depend on these numbers for decision making". If they miss a big fraud then obviously the auditor screwed up and that is a big problem. But it also means that management screwed up (or lied) too.
Dear Oracle,
Please add millions of dollars of accruals from expected business in to my Oracle ERP cloud account. Thanks.
Sincerely,
Drone CFO
The accounting firm can only comment on whether the records they receive are consistent and have a correct process based on their inputs.
That's not true at all. I'm an accountant and audits absolutely can and generally should investigate whether the records they review are factual and evidence based. In fact an auditor is supposed to look for evidence of fraud or mismanagement when doing a financial audit. If an accounting firm is not doing this when auditing the books for a large company then they are not doing their job properly.
If the wrong stuff is far enough upstream of the books, there's a GIGO problem.
That's when the auditor is supposed to decline to sign off on the books. Unfortunately audit firms have something of a conflict of interest. If they don't sign off on the books they might lose that client ($$$) and so they sometimes aren't are independent as they really should be.
My wife is a CPA and has changed jobs several times due to this kind of crap. the executives want the accounting department to lie on the books and fudge numbers.
She refuses to and has quit several places from time to time over this. Reporting it to the SEC every time. It's normal for executives to demand that laws be broken to make the books look better than they really are.
They must use RIAA/MPAA Accounting practices.
ie:
We old 10 subscriptions, we did not sell 300M subscriptions, therefore 300M people are pirating, so we will declare a loss, but will sue 300M people, assuming all of them will settle, so we will book 600M worth for subscription revenue.
I'm going to use that on my taxes.
I made 100K, but people could have stolen that money from me, so I made $0. Gimme a refund, some more free money, gimme gimme gimme.
Considering more and more companies are using non-GAAP accounting measures [marketwatch.com] to deceive investors, it appears they've already determined what method to use while not making it clear what they're doing.
Well if you are a global company you aren't using GAAP outside the US anyway. You probably are using IFRS. If you are an investor seeing words like Pro-Forma you should . Pro-Forma is roughly translated from latin as "Fairy Tale" or "Bullshit we made up".
That being said, any accountant worth their per-diem can use GAAP accounting that is perfectly legal and still obfuscate what is really going on. Plenty of companies do it. Go pull the financial records of any large bank and if you claim you can make sense of them you are either in line for a Nobel prize or a liar and I'm going to lean towards the later. I'm a certified accountant and I can't make heads or tails of them.
Accounting rules are crazy. Whenever I have something explained to me by an accounting, I'm often baffled at how it works.
The most recent example was a customer who sold prepaid punch cards to their members for an activity. The member bought a punch card for $100, good for 10 activities which normally would have sold for $12.
Strangely (to me anyway), even though the organization got all $100 at once from their member when a card was bought, they only accrued income when a punch was used.
My customer didn't have time to teach me accounting 101 and I sort of get the short version that was explained to me, it still seems kind of bizarre that you actually *gain* $100 but don't actually count it except $10 at a time over time.
In their case, it seems extra weird because the punch cards never expired and so there's the risk they would be never redeemed. I didn't get the explanation as to how that part is accounted for. I mean, if they took in $10,000 for punch cards but members only redeemed $5,000, it seems weird that you would carry $5,000 in liabilities essentially forever yet still have the $5,000 in cash already in some other accounting-speak category.
I imagine that Oracle has internally justified their accounting methods simply because their business model is changing out from under them. In short, there are no guidelines so make up your own rules. The problem with Cloud is you're selling the customer the _potential_ to use a service in the future usually. Knowing when to book that as an actual sale seems to be hard with this model.
I'm not an accountant, but I've taken introductory accounting courses twice, once in the 90s and once pretty recently. It's strange because back in the 90s, all the examples were straightforward. This was of course back when companies did LUDDITE things like taking raw materials as input and producing a finished product. Sales, cost of goods sold, etc. are easy to understand in that environment. Now US companies don't manufacture anything tangible - it's Apps(!) and eyeballs and cloud computing environments. That's tricky. I'm not saying Oracle is blameless, but new world accounting is definitely a different beast.
Accounting with the head in the clouds.
you actually think corporations like this pay taxes.
Probably not, but...
If you remember the dot com bust, there were a few precipitating events that got everyone looking closer at sky-high valuations. Enron is probably the most important of those. While Enron wasn't a dot com, they used some very creative accounting practices to book revenue and inflate their value (innovative, disruptive, new economy, and all that). Their crash helped highlight the funny math going on at internet companies and helped get investors and regulators asking the hard questions.
The current bubble has a few possible candidates for an Enron-style scandal. I've thought that Theranos might be it - they fit the broader narrative of youthful founders disrupting stodgy industries and minting new billionaires. Then there's Uber, but they have enough investor money in the bank to hide any structural issues with their business model for years. Palantir is another that may help burst the bubble as they learn that the outrageous consulting rates and blind faith in their methods they received from the military are difficult to replicate in the business world when you have some level of accountability to shareholders and customers. They have less cash than Uber to mask their situation much longer.
But, maybe it will be just like before: good old math that no longer adds up. At some point, all the SaaS companies using the clouds will run out of credits and goodwill from the providers and be asked to pay real rates for the services (just in case you don't know: almost no SaaS company actually pays for their hosting on the major cloud providers for the first year or so, and even then they pay deeply discounted rates. And if you're at a SaaS paying full rates to Amazon or similar, you need to call your rep - your competition is paying less than you. The cloud providers do this in hopes that strong businesses will emerge that they can eventually reap profits from - Salesforce is the classic example). I'm guessing* that all the major cloud providers are using some form of funny math to hide this.
So, Oracle's probably not going to suffer the fate of Enron. But, just like in the last bubble, "clever" accounting may be what finally pops it.
-Chris
*ok, I know from direct experience how at least two of them are doing it, and yes, they're doing it
Most big corps do crap like this all the time. Where are the posters defending such actions as looking to the shareholders best interests?
Cloud and subscription is just a return to the good old days of mainframes.
The corporation has the computer, the user has a dumb terminal.
Control has been taken from the user.
Remember the great revolution the IBM PC brought about ?
Well we're helping them to reverse the power balance.
Go well
Everyone who could investigate or prosecute Oracle is probably using Oracle products, and wouldn't dare go after them for fear of a license audit.
Should have included Oracle's accounting firm so we know who is going to shaft them, or go on the list of accounting firms that the criminal organizations like to use.
Is there Cloud bubble on the horizon as more of these issues come to light?
Accounting rules are crazy. Whenever I have something explained to me by an accounting, I'm often baffled at how it works.
No they aren't crazy. You just have to understand some basic principles.
Strangely (to me anyway), even though the organization got all $100 at once from their member when a card was bought, they only accrued income when a punch was used.
That's not strange at all. To understand it you have to understand double entry accounting. Every transaction has two entries - a debit and a credit. Think of them kind of like and In and an Out - they have to balance. There are basically five categories of ledger accounts. Assets, Liabilities, Equity, Income, Expenses. Every transaction has a debit of an account in one of the five categories and a credit in an account of one of the five categories. The debit and credit can be in the same category or different ones.
That transaction you mentioned works like this:
When the card is bought Assets (cash) increased (debit) by $100. Liabilities (future punches) increased (credit) by $100.
Later when a punch is used Liabilities decrease (debit) by $10 and Income increased (credit) by $10. (this is when the sale occurs)
Basically the company has traded an liability (the obligation of services for the punch) for an asset (cash). As that obligation is fulfilled the company realizes the income from that sale. It's done this way for several reasons but most important is the matching principle - basically that expenses and income should be booked when they occurred, regardless of when the actual cash transfer took place. The company didn't actually sell it's product until the customer used the punch (the punch is the product) so it would be inappropriate to say they sold something until the punch transaction results in Income.
it still seems kind of bizarre that you actually *gain* $100 but don't actually count it except $10 at a time over time.
It seems weird because you are thinking of the $100 as income when in fact it is an asset. To make it easier instead of getting $100 for the card let's say the customer gave the money to buy an orange on their behalf. Orange slices are just another form of asset like cash. So when the customer gets their card punched the company would give the customer an orange slice. The company had to pay something to buy that orange slice so until they actually trade it to a customer a sale hasn't occurred. It seems confusing because we usually associate receiving cash with income but they aren't necessarily the same thing.
A customer can't buy a liability which is what the customer is doing when he pays $100 for the card. He can only buy a good or a service which is what they can do with the card. So until that good or service is rendered to the customer then it isn't income to the company. The fact that cash changed hands before the good or service was rendered is immaterial to when the company actually realized the income. It's a bit like buying something on layaway.
In their case, it seems extra weird because the punch cards never expired and so there's the risk they would be never redeemed.
That is EXACTLY what happens and is why you typically see expiration notices on checks and cards. If they never expired and were never used then the company would have to carry them on the books forever. If they don't expire and the customer loses the card then the company has to carry that obligation in theory forever. In practice there are ways to work around it but they are considerably more problematic.
But being "the cloud" is generally new, what would be an "acceptable" way to make an estimate for renewal rate?
Ahh, you've hit on the crux of the problem. With any new business model there exists the problem of trying to figure out what accounting practices are considered reasonable for that business. Usually this takes a bit of time and eventually there is usually guidance from the IFRS, IRS, SEC and other bodies for acceptable booking practices. Remember that GAAP stands for Generally Accepted Accounting Principles and much of GAAP (and IFRS even more so) is principle based rather than black and white rule based. Because every company is a bit different the accounting treatments can be different as well to best reflect the structure of the business. It wouldn't surprise me at all if Oracle was being rather aggressive with their accounting treatments. In all likelihood the subscriptions will probably be treated much like the accounting for magazine subscriptions.
Personally I'm not overly familiar with the nuances of accounting for subscription based businesses but here is a little primer.
Or, could Oracle just say "it's a new field so we simply guessed" in their official accounting? Or, would they be expected to survey somewhat similar services, such as mainframe or super-computer subscriptions, to find a rough approximation?
Basically yes. They can probably hide behind the fig leaf that they weren't sure because there wasn't official guidance on the matter. In reality I'm sure Oracle's accountants are pretty bright so it's more of plausible deniability thing. Personally I don't see how cloud subscriptions are any different from other types of subscriptions and that is well understood in accounting practice. What will probably happen is that the IRS will probably weigh in on the matter at some point (if they haven't already) and that will effectively write the law about how it should be handled going forward. If there isn't fraud in play then it probably is a matter of overly aggressive accounting practices.