Many people enjoy mystery stories or crime thrillers; in the same vein of savoring the whodunnit and howdunnit, I like a good accounting scandal. My fascination with cooking the books started when I was young with the “great salad oil swindle”, which wound up causing losses in excess of $1 billion in today’s money and even threatened a Wall Street collapse. This disaster was averted by the assassination of President Kennedy, which kept markets closed on Monday, November 25, 1963, and gave the parties involved an extra day to resolve the matter. Nowadays I look forward to receiving FIRST, a compendium prepared by IBM’s Risk Analytics group of the previous month’s financial shenanigans. So, the recent Hewlett-Packard-Autonomy imbroglio fascinates me.
HP’s current management says “serious accounting improprieties” on the part of Autonomy have cost it $5 billion as part of an overall charge of $8.8 billion taken in the company’s fourth quarter of fiscal 2012 and articulated by HP CEO, Meg Whitman. HP asserts that Autonomy accelerated revenue recognition and misconstrued lower-margin hardware sales as higher-margin software business, and the related hardware cost was classified as marketing expense. This kind of revenue misrepresentation can begin small but ultimately mushroom; it puts companies on what auditors like to call a slippery slope of having to deal with ever-mounting shortfalls. It can start when management wants to believe that a permanent change in the business (say, some unfavorable trend in demand or pricing) is simply a one-time anomaly.
Misclassifying hardware revenue as software licenses, labeling cost of sales as an operating expense, would not have had an impact on reported earnings (since the aggregate debits and credits are unaffected) but could have an impact on the valuation of the company. This would be the case if the practice was used to disguise shortfalls in license revenue that would have led to a drop in the company’s share price or was intended to show faster growth in the more attractive software license revenue category. Lynn Turner, once the chief accountant at the U.S. Securities and Exchange Commission (SEC), was quoted as being puzzled as to how $200 million of disputed revenues is connected to the $5 billion number. Well, it’s through the magic of a revenue multiplier used to justify an acquisition price.
One of the lessons I’ve learned about situations like this one is not to jump to too many conclusions if you suspect you don’t have all of the facts. It's not entirely clear how a public company that had its financial statements audited annually committed fraud in an acquisition process. At the same time, I’m not naïve enough to think it's impossible or even improbable. Indeed, it’s not as if Autonomy’s accounting policies went unquestioned when it was a public company. The Center for Financial Research and Analysis (CFRA) published multiple notes from 2001 to 2010 pointing to what it believed were dodgy accounting and disclosure practices. In my career I witnessed accounting by British listed firms that would have been hooted down in the U.S. but that were accepted by British markets. It’s also true that CFRA is not the final word. It has had its share of false positives in the past, where its red flags have proved not be serious issues.
At the same time, I don’t think you have to be too much of skeptic to regard this as a political maneuver on the part of Meg Whitman, who gains a great deal by taking some of the sting out of a poorly conceived acquisition by blaming it on fraud. Although she was not running the company when the decision was made, she was a member of the board. It may have been an astute move from the standpoint of internal politics, but whether it was a smart decision is questionable since it tarnishes the Autonomy brand and does not enhance HP’s.
For those of you on the International Financial Reporting Standards (IFRS) bandwagon, consider this: It would have been much harder for Autonomy to have mischaracterized revenues under US-GAAP. Indeed, one of the sticking points in converging IFRS and US-GAAP is the subject of revenue recognition. As the result of multiple software company accounting scandals, the U.S. has the most specific rules governing the timing and treatment of sales revenues. Being more principles-based, IFRS is more concerned with the overall appearance. In theory, that shouldn’t make a difference. However, U.S.-based auditors are more likely to flyspeck revenue treatment decisions because of their very precise revenue recognition rules.
I suppose, on the bright side, HP will have a success story of how it was able to use its own eDiscovery software (which came with the Autonomy acquisition) to digest tons of email and documents for evidence of fraud.
For now, the bottom line for me is that the Autonomy accounting irregularities could be very real but are at the same time a fig leaf for the write-offs associated with a bad acquisition. Internal and public auditors are bound to learn a host of lessons. And the affair is bound to give those on the SEC and the Financial Accounting Standards Board evidence for keeping US-GAAP separate from IFRS, even as the two systems converge on most accounting matters.
Robert Kugel – SVP Research