HP, Autonomy, allegations of fraud … and terabytes of analysis

23 November 2012 – So Hewlett-Packard (HP) has written down $8.8 billion of the value of Autonomy, the British software company it bought last year, after discovering that Autonomy misrepresented its finances, HP stating that it had uncovered “serious accounting improprieties” at the company including “outright misrepresentations” that occurred before HP acquired the company. The markets were not amused.  HP stock plunged and has lost nearly half of its value this year.  And now the FBI in the U.S. and the Serious Fraud Office in the UK are launching investigations.  Lots of yammering and stammering by Leo Apotheker (the ex-HP CEO who agreed the deal) and Mike Lynch (Autonomy founder).

Even the BBC has been dragged in (as if Jimmy Savile wasn’t enough) because Lynch is a non-executive director which caused the BBC to issue a statement that “we will be discussing these reports with Dr Lynch imminently”.

And HP has been struggling to find its Happy Place.  The Autonomy write-off was the second huge one in successive quarters: in July HP wrote down the value of EDS, the IT services company it had bought in 2008 for $13.9 billion, by $8 billion. In the fourth quarter it reported year-on-year declines in revenues from PCs, printers, services, servers — just about everything, in fact, except software, where revenue was up by 14%.

But let’s be honest.  If HP failed to do adequate due diligence on its $11bn acquisition of Autonomy, it is not because too few experts were hired. Two auditors, Deloitte on the Autonomy side and KPMG on the HP side, worked on the original deal. PwC is doing an autopsy; Ernst & Young is, presumably, sulking and waiting by the phone. Then there are the banks, six on Autonomy’s side (Qatalyst, Goldman Sachs, Citigroup, UBS, Bank of America, and JPMorgan) and two on HP’s (Perella Weinberg and Barclays).

In an interview on Wednesday, Lynch conceded one of HP’s main charges, that his company had sold computer hardware at a loss to some of its customers and treated part of the cost of these sales as a marketing expense in its accounts. HP has accused Autonomy of using the sales to inflate its revenues, while burying the costs in order to boost its gross profit margin and give the appearance of having made high-margin software sales. However, Lynch said the hardware sales had been far smaller than claimed by HP and that only around 2 per cent of Autonomy’s revenues were related to the unprofitable sales that make up the most controversial part of the hardware business, rather than the 10-15 per cent alleged. He also dismissed the claim that he had hidden the fact that Autonomy was selling hardware and said the accounting treatment of the costs had been approved by his former company’s auditors, since they amounted to a marketing cost to support other Autonomy business.

I have some links below that discuss the accounting issues and the smoke and mirrors involved.  As one analyst said “it is relatively simple to inflate the value of software companies with sharp accounting. There is more scope for abuse because you can press a button and record a sale at almost zero cost.  In other businesses there tends to be more involved, like making and delivering something.”

Side note:  one cannot miss the irony here.  When you look at the numbers, the process, you are in the magical world of what the pundits call the “new normal” : legalized accounting fraud.  You simply game the tax code, and with that old reliable, the “kitchen sink” write-off method, you cleaning up bad business decisions/practices. Volia!  Meanwhile, these corporate “personhoods” are swimming in trillions of dollars in cash on their balance sheets. While over there on the “fiscal cliff” of national bankruptcy … 

At the time, the Autonomy grab was widely criticized by many technology analysts as too expensive, given the accounting seemed to be pretty fuzzy numbers … some finding it all “vaporware wrapped up in fancy Cambridge talk” with the kind of accounting tricks managers have engaged in since the dawn of publicly traded stock.

Shortly after the acquisition last year I had a chat with Nick Patience about the move by H-P … Nick is now with Recommind but then was with 451 Research which he co-founded … and if you are a follower of the 451 Research band of analysts you know that he and others had watched Autonomy very closely and the numbers didn’t add up.  As Nick said, there had to be “a willing suspension of disbelief”.  As an example, here is something Nick wrote in July 2010 on an  Autonomy quarter (click here).

To me it was just another story of older tech giants spending billions at supposedly “game-changing” acquisitions.  The phrase I remember most from the Enterprise Search Summit in London last fall after the acquisition was that of an analyst who called it a “structural panic buy”.

Lost in all the analysis this week was “can Autonomy products do the job?” To many software analysts Autonomy’s core enterprise search and data analysis technology is viable but very complicated and expensive.  Several analysts have written about an October 2012 HP demonstration of the Autonomy product used in an application to help figure out what to pay for Web ads.  It initially failed to work, then delivered results that appeared similar to those of several other search products. I saw some of the software in action when I was involved with the SocGen/Kerviel trial a few years ago.  I saw how SocGen had used Autonomy software in that massive bank fraud case to work out what had been going on with Kerviel and his trades, as well as his communications — the same technology used for early case assessment, document review software, and social media analysis. It was interesting software and it afforded me the opportunity to dig deep into the enterprise software market and to see how Autonomy sought to commercialize statistical techniques (developed by Lynch at Cambridge University) based on Bayesian inference, a mathematical technique that can estimate the probability of potential outcomes based on previous evidence. And again a big hat tip to the 451 Research folks who allowed me access to much of their analysis of the enterprise market.

Plus last summer I had been working on my e-book “A Culture of Analytics” which profiles IBM and EMC so I had been somewhat steeped in material on EMC, H-P, IBM, Oracle, etc. I had studied the success of EMC who paid $625 million in 2003 for VMWare (which makes some of the critical technology in cloud computing) which today … after taking some of the company public … has pushed EMC’s stake to $30 billion. It has been an excellent job of analysis and integration. It was a well thought out strategy.  EMC did not abandon its dedication to its bread-and-butter storage hardware business, or slow its development. EMC knew its IT customers value reliability and performance above all other features, especially with storage, so it kept its eye on the ball. And a big, big difference: EMC had been working with the VMware software for a long time and knew it well.

And as for H-P, as I wrote in my blog at the time, I was unconvinced that H-P’s ex-CEO Leo Apotheker could pull off this “Big Bang” for a number of reasons, not least of which was H-P’s poor track record at simple business execution and integrating acquisitions.

The irony was that HP wanted to make itself more like IBM, that Autonomy’s number-crunching technology could take on I.B.M. in data analysis.  But H-P executed in all the wrong ways, with the lack of cooperation between the two companies quickly became evident. The web has scores of “behind the scenes” stories.  At IBM, Gerstner refused to act precipitously after taking over in 1993, not dismembering Big Blue as he was urged to do by IBM insiders and industry pundits, and go a different way.  Yes, he had an alternative vision but he made damn sure it worked in practice before  making final decisions. For my blog post last summer click here.

The long-term effect of all this is uncertain although Law Technology News has noted that the 17 percent of the e-discovery market Autonomy held among Am Law 100 firms last year has since fallen to just 4 percent per the most recent technology survey by The American Lawyer. And I am sure Meg Whitman’s tepid support is no help. When asked about the role of Autonomy in HP she said “It’s [the writedowns] are very disappointing … [but] we’ve integrated the technology into a couple of places; we will integrate it into more.”

Well, perhaps two things are certain.  As Craig Ball writes ”it can be safely concluded that the acquisition will indeed pump millions into the e-discovery economy as HP sues everyone in sight”.  And my God, the legal fees to come.

There have been terabytes of comment and analysis on the H-P/Autonomy imbroglio this week.  Here are links to some of the “what this is all about” analysis:

Hewlett’s Loss: A Folly Unfolds, by the Numbers (from the New York Times)

With Autonomy, H-P Bought An Old-Fashioned Accounting Scandal. Here’s How It Worked. (from Forbes)

HP Autonomy Accounting Scandal Explained (from Business Insider)

HP Is In A Whole Mess Of Trouble Right Now [VIDEO]  (from Business Insider)

Oracle’s Ellison Vindicated in Autonomy Flap (and about that infamous Powerpoint …) (from Wall Street Journal)

Why EMC Won With VMware and HP Failed With Autonomy (from Forbes)

It’s Not Just HP And Autonomy, The Enterprise Software Space Is A Giant Stinking Mess (from TechCrunch)

Deal Lawyers Mostly Mum on HP’s Ill-Fated Autonomy Acquisition (from Law Technology News)

H.P. Claim Highlights a Gray Area in Software Sales (from the New York Times)

Founder/CEO, The Posse List