Wednesday, November 21, 2012

Hewlett Packard and Autonomy: notes from my Santangels presentation

A few weeks ago I presented at the Santangels conference in New York. The conference was held under Chatham House rules.

My topic: how to fake your accounts from the perspective of the fraudster (and hence how as an investor to tell if a company is faking its accounts).

At the end I suggested that even if you were right and the company was a fraud you could still short the stock and lose money - someone with deep pockets might buy the fraud. The example I gave was Autonomy which was purchased by Hewlett Packard.

Yesterday Hewlett Packard admitted that Autonomy was a grotesque fraud. They wrote off most of the purchase price.

Chatham House rules don't much matter in the face of a major scandal (and Autonomy is a major scandal). So this morning (Australian time) my email was full of press asking for on-the-record comment. I commented to both Deal Book (New York Times) and Reuters.

Jeff Matthews commentary

I tested my draft Santangels presentation on Jeff Matthews - so he had my views on Autonomy. He broke silence on his blog.

Jeff is critical of Meg Whitman's comments that they "relied on the audited accounts". He observes that I relied on the same audited accounts.

That is fair comment: I could tell the accounts were not kosher from an office in Sydney. I was not alone - Jim Chanos came to a similar conclusion. So did many people in the UK including the writers of FT Alphaville who have been sceptical of Autonomy for years.

In other words my insight was not unique: many other people (including more than one blog reader) shared similar insight. I claim no special genius.

And I was not entirely right either. I knew Autonomy was problematic. I underestimated the problems: Autonomy appears worse than the worst case I modelled.

Looking at Autonomy's accounts

For the record here is how you could tell - just by looking at audited accounts - that Autonomy was not quite kosher.

Here is the P&L from the 2010 accounts:


And here is the balance sheet:


There are lots of things to note - but I will limit myself to the simple (which I put in the Santangels presentation).

Sales were $870 million.

Receiveables were $330 million - which is four and a half months of receiveables.

Deferred revenue is $177 million - just over half of receiveables.

This is really perverse for a software company. Software companies sell stuff that is barely tangible - they sell it up front and for cash. They have very few receiveables.

They do however have an obligation to service that software for a long time after they sell it - so the unearned income is relatively large (usually a multiple of receiveables).

Autonomy was booking as income lots of cash it had not received (which is why the receiveables were large) and not booking any obligation to provide future services for that income.

This is prima-facie suspect (and you could tell simply by looking at the balance sheet). All it required was basic applied accounting.

Comments on Hewlett Packard's management 

The management (and for that matter the board) of Hewlett Packard have diverse responsibilities. Most importantly they are responsible for technology - and the technological issues are broad. They are also responsible for accounting and financial management.

I have simply no opinion on who stuffed up on the technical issues. Hewlett Packard asserts there are still some (small but) valuable businesses at Autonomy. Maybe the tech is not all bad.

But I am happy to point the finger on accounting. The accounts were self-evidently suspect. Working that out is primarily the job of a finance and accounting function at Hewlett Packard - and that function was led by and continues to be led by Catherine Lesjack. Catherine Lesjack is ultimately responsible for the financial parts of HPs due diligence on Autonomy and she failed.

This really is not rocket science. Lesjack is not up to the job. She should resign.





John

PS: There has been much criticism of Meg Whitman in the blogosphere, twitter and amongst my readers. I do not think that is fair. Autonomy was a deal done by Leo Apotheker (the previous CEO) and the current CFO Catherine Lesjack - it is them who bear primary responsibility.

J

44 comments:

taxloss said...

One of the stranger things about the HP end of this deal was that apparently the due diligence peeps reported into the Chief Strategy Officer, not the CFO.
See here: http://www.theregister.co.uk/2012/11/20/hp_alleges_autonomy_fraud/

so perhaps Ms Lesjack gets a pass.

John Hempton said...

Lesjack does not get a pass because she thought she had no responsibility.

The responsibility on financial matters is hers.

The only excuse for giving Lesjack a pass would be that the board ORDERED her not to be involved or SHE OBJECTED and the board over-rode her.

Either of those - and we will find out soon enough.

John

Anonymous said...

Who was the main beneficiary from the sale of Autonomy?

Anonymous said...

Meg was on the board though and voted "Yes" on the deal, no?

Anonymous said...

I guess she knew...
http://www.businessinsider.com/hp-cfo-kathie-lesjak-tried-to-stop-117-billion-acquisition-of-autonomy-2012-5

Anonymous said...

I've always thought that Autonomy had a strong whiff of fraud about it. I never hear anyone talk about working there, their financial reporting is often shadier than Groupon's, there's a Madoff-esque figure at the top (in the sense that he seems to be treated as almost godlike), and no-one seems to use its products.

I know it powers the FT's search facility and they joke about how useless it is.

My only explanation has been that it perhaps has a heavy focus on intelligence work.

Anonymous said...

For your more junior readers, could you please explain a bit more how the account receivable and deferred revenue should look? I would guess that deferred revenue should be much larger than account receivable, is that right? If so, I still don't understand what does selling intangible things have to do receivables.

Anonymous said...

In the untimate warning two ( not one but two) of the current auditing firms who call themselves the Big Four ( I call them the "Big Fraud") signed off on this to give the corporate lackey's cover- Deloitte & Touche & KPMG.

We need an SEC invesigation as to how they signed off the due dilegance that was 88% of value was fraudulant.

Nick said...

Our short thesis was two pronged and even more simple:
1. The "product" was impossible to understand.
2. If you stripped out M&A their organic growth was negative.

Of course their balance sheet stunk to high heaven as well.
You can have a delta between earnings and cash for a while - but not for ever.

However, it was on our "do not short" list because I didn't have the stones to wait for it to implode.

As ever - luckier than smart!

Anonymous said...

A minor technical point.

"Autonomy was booking as income lots of cash it had not received (which is why the receiveables were large) and not booking any obligation to provide future services for that income."

Autonomy is not required to book an obligation to provide future services for income previously recorded. If everything was kosher, technically you cant book income until the obligation has already passed.

Deferred revenue refers to cash received, where an obligation still exists. They have not recorded any income on this cash received.

As a result, I struggle to see why a low deferred revenue balance is a red flag? A low deferred revenue balance just means customers are not paying up front.

Anonymous said...

John, is or was your fund short HP as the Reuters article states? Excellent call if you were.

NoMeanSum said...

Great post. The ex-CEO of Autonomy seemed very convincing to an uninformed person such as myself but the numbers tell a different story.

Anonymous said...

Thanks for the tips on spotting bad stuff, John.

Having been through multiple mergers/acquisitions at various technology companies, even as a low-level grunt I could see that generally the due diligence was a farce. It's irrelevant to the deal. The CEO and the board have already made up their minds by the time DD is undertaken, and nothing is going to stop them.

Anonymous said...

John,
I think it's hard to expect Lesjack to have opposed more strongly than she reportedly did (this from the Fortune piece last May, which seems well informed):

It was a stirring presentation, but it was followed by an even more dramatic moment that blindsided Apotheker. He knew that the CFO, Lesjak, opposed the deal. She had told him the price, around 11 times revenue, was too rich. Comparable companies were selling for three times revenue, according to investment bank Software Equity Group. He'd countered that Autonomy's profitability more than justified the price. The two had discussed it privately.

But then, with no warning to Apotheker, Lesjak made an impassioned case against the acquisition before the board. "I can't support it," she told the directors, according to a person who was present. "I don't think it's a good idea. I don't think we're ready. I think it's too expensive. I'm putting a line down. This is not in the best interests of the company." Directors were shaken. Lesjak was considered a voice of sobriety, and here she was on the verge of insubordination, directly resisting a key element of her boss' strategy.

Good article:
http://tech.fortune.cnn.com/2012/05/08/500-hp-apotheker/
What a mess...

john b said...

Anon 8:41: Mike Lynch, obviously.
Anon 9:57: nah, "you're paying too high a revenue multiple" isn't really her business. "the revenues aren't real" is.

Given the extent to which everyone (as a former DD guy who's now a freelance writer with no insider anything, I'm including myself in "everyone" here) knew Autonomy revenues were aggressively accounted, this is going to be an interesting moment for KPMG especially. Sensitivities related to differing accounting policies should have picked up the problem, based on John's two numbers, even if Autonomy wasn't expressly fraudulent.

Anonymous said...

John, could you post your Santangel’s presentation on this site? Thanks!

David Merkel said...

H-P’s Deal Process Bypassed CFO Pre-Whitman

http://blogs.wsj.com/cfo/2012/11/20/h-ps-deal-process-bypassed-cfo-pre-whitman/?mod=wsjpro_hps_cforeporttab/print/

'Whitman said in a conference call with analysts Tuesday that, soon after taking over as CEO, she reorganized the company’s M&A structure to put CFO Cathie Lesjak in charge of due diligence.

“At the time when I came to the company I was surprised to find that due diligence and M&A reported to strategy as opposed to the chief financial officer,” said Whitman, according to an unedited transcript of the call provided by FactSet. “I’ve never seen that before in my career and that’s the decision I made right away before I knew any of this.”'

Lesjak should have had the power to stop the deal. She did not have the power; she still tried to stop it.

Anonymous said...

Several media / blogs report Lesjack objected and was overridden, e.g.

"HP acquired Autonomy in October of last year. The acquisition was initiated in August by then-CEO Apotheker in a bid to expand HP's miniscule software business. Apotheker was fired a little more than a month later, but his replacement, current CEO Whitman -- who was a member of the board that green-lighted the acquisition -- went ahead with the Autonomy deal. HP CFO Cathie Lesjak had reportedly objected to the Autonomy deal at the time, citing the high price tag."

alkali said...

Software companies sell stuff that is barely tangible - they sell it up front and for cash. They have very few receivables.

Not sure this is right. My understanding is that SOP 97-2, a company licensing software can book revenue according to one of several revenue models: (1) recording revenue only when cash is received, (2) recording revenue on sale ("up front") even though the cash may be paid later, if various criteria are met, or (3) recording revenue on a "subscription" basis in increments over the term of the license. If the company chooses a non-cash method, then it's certainly going to have receivables.

Aaron said...

Anonymous who asked about AR and deferred revenue... I'll take a crack at your question...

What John is doing (and why I love this freaking blog!) is looking at the accounts IN THE CONTEXT of the underlying business.

Basically "normal" AR and deferred revenue ratios should reflect the nature of the business and typical billing within the industry.

Most software businesses receive payment in full up front. In that scenario you would expect AR to be small compared to revenue (maybe one month?). Yet Autonomy's accounts indicate that there are customers who purchased the software 4.5 months ago and still have open balances. Maybe they bill differently than typical software companies but basically that would be a red flag you'd want to look into because 1 possible explanation is that they are cheating and overstating the revenue. I don't know anything specific about Autonomy but, one way this could happen for example is if they took sales they expected in the next accounting period, recognized them as revenue in the current period and also recognized the receivable even though it hasn't even been billed yet. This would have the effect of making it look like revenue was growing faster than it actually was but of course would someday catch up to them.

Deferred revenue is money that they HAVE actually received but not "earned" yet. In the case of a software company, they likely sold their software with a service package for, say, 1 year. In that scenario they should be recognizing the service revenue evenly throughout the year so that it matches the costs of providing the service contract. If they wanted to cheat, they might for example change their accounting policy for deferred revenue to recognize it as earned earlier than is justified. That would push up the revenue number but of course be a problem because they still have the very real costs of servicing their software throughout the year.

Both thinking about it in the context of the underlying business and comparing the ratios to similar businesses can help you find "normal" ratios for the company and identify potential red flags. Also you want to look at not just the ratio but the trend in the ratio. In this case the deferred revenue ratio got smaller compared to the previous year and the AR ratio got larger (although to be fair, not by much)... Both moving in the direction you would expect if there was ongoing cheating going on.

By the way, really really great book if you want to learn more about this is called "Financial Shenanigans."

Anonymous said...

Read FT alphaville - there has been a multi yr coverage of autonomy. Lots of analysts had raised issues. Why had no bank picked this up in dd?

Anonymous said...

Would also like to see the "Faking Reported Income 101" presentation posted on the blog. Many of us come here to learn forensic accounting rather than for an individual short name. How about it John? If cannot on the blog maybe another route - email to bloggers?

Anonymous said...

John,

This is truly incredible. I immediately thought of your Santangel's presentation (which was excellent, by the way) when I saw this announcement.

Anonymous said...

Bottom line is: How many sales recognized in Account are Fraudulent.

Shanghaied said...

John, I would also be very interested in seeing your Santangels presentation.

Thanks
Mike

alexis said...

John, could the AR be revenue from support and maintenance? When a software license is sold for, say, $100, there is usually an annual maintenance fee of, say, $22 on top of this. Sales revenue normally includes 3 years of this, so for every $100 license, a sale of $166 would be booked. Of this, $122 is paid upfront, and $44 is paid in years 2 and 3. Apologies if I have misunderstood your point.

Anonymous said...

I used in about 2005 to work for a company next door to Autonomy in a shared building in Cambridge and I knew two of their staff, non-engineering, secretary-types.

Both said working there was desperately stressful and they both sought to leave ASAP.

I thought it odd a company apparently so successfully could be so unpleasent to work at.

(Blank Xavier)

Lior said...

John,
Thank you once again for great insights.
Would you recommend, for those of us who do not earn our living from the financial sector, a text or a book which teaches how to read such annual reports, including the income statement, the balance sheet, and the cash flow?

Cheers,
Lior

Falkor said...

Hello John,

Nice take but really half baked here. I mean, you gave one (glarinly obvious)point on how not to be defrauded like HP got into.

Please tell us more ways to identify the frauds which I guess is what your presentation was all about.

If you could make that into a series of posts over sometime, it would be great.

Thanks.

Anonymous said...

"In our balance sheet on the left side there is nothing right, and on the right side there is nothing left"

Anonymous said...

On Autonomy's balance sheet, on the left side, there was nothing right and on the right side, there was nothing left.

t said...

John,

Just for the record, are you saying Autonomy had fraudulent accounts or were just very aggressive in their accounting. I don't know whether you'd argue that fraud (i.e. "deception made for personal gain") can be done within the law or not. If you would, are you saying that Autonomy violated accounting standards?

I don't know whether you ever looked at their I.R. website when they were qouted but they had the biggest I've ever seen - answering in detail every single analyst's question about the accounting... and there were a LOT of analysts who used to ask about the accounting, especially the revenue recognition.

Conscience of a Conservative said...

Interesting comments on the accounting, but it seems incredible to me that H.P. did no independent due diligence here and that the "fraud" resulted in an over payment of 8 billion dollars. Am I missing something?

Anonymous said...

i think the founder of the Automomy has a good point: the accounting issue is referring to about 200 million of deferred revenue, yet HP is writing down 10 billion. They can't just blame previous CEO or Automomy. The board are either very stupid or somehow benefit from these kind of deals.

Conscience of a Conservative said...

The amount of the write-down befuddles the mind. Can simply not recognizing unearned income be worth 8 billion dollars? Also this sounds like rookie accounting stuff. Can't believe the auditor and HP didn't spot this.

Anonymous said...

The comment on receivables is naive.

Those receivables do not look out of whack at all for an enterprise software company. I used to be CXO one selling about $50M in orders a year, and the lag between rev. rec./invoice date (typically on acceptance, not shipping) and cash, depended on the negotiated terms - and with big telco customer it was typically a minimum of 90 days, sometime as much at 210 if they had power, plus delays due to "confusion" and God help you if it was Orascom Telecom Holding.

I agree the balance sheet stinks though - all that goodwill and intangibles...

Anonymous said...

Dear "t" (Nov 23 2:12)

Very difficult to prove fraud just from financial accounts. Red flags can be a good starting point for further investigation.

Re: Investor relations etc. Typically, investor relations/C executives will have very convincing answers for all questions (even tough ones!) regarding accounting and financial statements. It is their job after all! In my experience there is absolutely ZERO to be gained from talking to management other than to ask clarifying questions about accounting choices.

I have never heard a company rep acknowledge they are accelerating revenue recognition or the like and doubt I ever will!

Relevant Chanos quote : "In the last 25 years I can not think of a major fraud that didn't have audited financial statements that confirmed to GAAP"

JB

Anonymous said...

One of the more interesting things about Autonomy were that they did not have an internal audit group and they voluntarily delisted from the U.S. ahead of Sarbanes Oxley implementation (citing costs of compliance). Lack of an internal audit group in a high growth, acquisitive company with a market cap well over the billion dollar mark is a huge red flag in and of itself, not to mention the games they were playing with organic growth, the disconnect between receivables and deferred revenue, and the really aggressive cost capitalization they use. Talk about the perfect control environment to commit fraud...
Everyone involved on the due diligence side of this deal should be fired and Meg does not escape blame here. I don't know how much the break up fee was, but it had to be less than $8.8 billion.

And the Autonomy CEO was pretty disingenuous to suggest that $200 million was a pittance compared to the write-down when he knew full well that the only reason software companies get bought at such ridiculous prices is for the implied growth potential. If that $200 million wasn't there, Autonomy wasn't reporting growth and no one would've been interested in buying them out.

Anonymous said...

Not sure if you analysis is correct. I looked at ADBE, SAP and DOX and all 3 have similar AR/Sales and Def Revs/Sales ratios. Not sure where you got this rule of thumb?

Anonymous said...

What if a software deal has a large portion of services and payment would not be rendered until services are completed? Would they be able to recognize this as a receivable? My understanding is that almost all of their deals have services that cost as much or more as the actual license costs. How would this/should this be reflected in the books?

Anonymous said...

What if a software deal has a large portion of services and payment would not be rendered until services are completed? Would they be able to recognize this as a receivable? My understanding is that almost all of their deals have services that cost as much or more as the actual license costs. How would this/should this be reflected in the books?

Anonymous said...

Autonomy bought a company we used at one of my previous workplaces. They immediately canned the good support staff (or they left) and replaced them with crappy unresponsive support.

But yeah goodwill I'm sure if its on the balance sheet it must be for reals.

Anonymous said...

Leo has a history of ignoring advices and doing his own way, eg. cut touchpad in 6 days, selling PC division, etc...

Meg Whitman Sucks said...

Great article and accurately described.

General disclaimer

The content contained in this blog represents the opinions of Mr. Hempton. Mr. Hempton may hold either long or short positions in securities of various companies discussed in the blog based upon Mr. Hempton's recommendations. The commentary in this blog in no way constitutes a solicitation of business or investment advice. In fact, it should not be relied upon in making investment decisions, ever. It is intended solely for the entertainment of the reader, and the author.  In particular this blog is not directed for investment purposes at US Persons.