Autonomy/HP analysis: The accounting argument

THE FAILED TAKEOVER of software outfit Autonomy by technology giant Hewlett-Packard has loomed large in the headlines recently. And despite an abundance of commentary on the doomed deal, no-one it seems wants to speak out on the record.

What we do know is this. On 19 October 2011, HP purchased Autonomy for $11.1bn (£6.7bn) and on 20 November 2012 announced it would write down the goodwill on that acquisition totalling $8.8bn. The largely converged business combinations literature on both sides of the Atlantic simplifies the task of unpicking HP’s move. M&As are accounted for under IFRS 3 Business Combinations in the European Union and FASB ASC 805 Business Combinations in the United States.

Both standards require businesses to apply the so-called acquisition method. The assumption is that when an acquiror pays a premium price, it does so because it can see added value in the business. This premium is treated as goodwill and recognised as an asset. It is this asset that HP has written down.

HP said in its statement that “the majority of this impairment charge is linked to serious accounting improprieties, disclosure failures, and outright misrepresentation at Autonomy.” Those improprieties allegedly include inappropriate recognition of revenue on software sales to value added resellers.

Strong stuff. Then, on 18 December 2012, HP shareholders filed a verified shareholder derivative complaint. This action names HP’s management, among them CEO Meg Whitman, and advisers as co-defendants, and alleges that “proper due diligence” would have flagged up the deal as a dud.

The deal is now the subject of separate investigations on both sides of the pond, by the US Department of Justice and the Serious Fraud Office in London. Separately, an investigation by the US Air Force into the affair led to a call from USAF to ban Autonomy’s founder and CEO, Mike Lynch, and others from contracting with the U.S. government.

Autonomy’s former managers vigorously deny any impropriety and now protest their innocence on a website. In an open letter to HP, they argue that their policies were “[c]onsistent with IFRS and IAS 18.”

It is worth noting that there is no explicit prohibition in IAS 18 on the recognition of revenue immediately on a sale from a manufacturer to a reseller. IFRS is much less detailed and prescriptive than US GAAP, which has specific guidance on software sales – SOP 97-2. Indeed, it is these differences that the IASB’s joint revenue recognition project with the FASB sets out to fix.

Put crudely, many of the differences between IFRS and US GAAP amount to the critical events that trigger revenue recognition. IFRS requires preparers to exercise judgement around the question of whether or not the risks and rewards under a contract have transferred, and whether the inflow of benefits is probable before a reseller has an onward sale of the product.

The landscape is complicated by the fact that some technology-sector preparers fall back on US practice when developing their accounting policies. Equally, reseller arrangements vary and each must be assessed against the criteria in IAS 18.

How this plays out in the real world is perhaps best illustrated with Alcatel-Lucent’s latest 20-F filing. Assuming the relevant conditions are met, Lucent books revenue “on shipment to the distribution channel if such sales are not contingent on the distributor reselling the product to a third party and the distribution channel contains no right of return.” Otherwise it recognises revenue “when the reseller or distributor sells the product to a third party.”

Time, and quite possibly the courts, will tell whether HP was short-changed or the victim of its own hubris.

Stephen Bouvier is a freelance journalist

Related reading


  • faro [london]

    sorry me cofuzed as i always knew from my 1970’s accountancy studies’ hp’ means hire purchase, omg, * not ‘hubris power’…

  • Verity

    Tell me, what do the regs say about a company which claims to be wholly a software provider then sells hardware (allegedly 90m+ of HW in one year) and fails to tell its shareholders its selling hardware under the auspices that the hardware is a marketing expense, and to boot gets the auditor to sign-off on it?

    • Bob

      1. That’s an $8.8bn writedown! $90m is a drop in the ocean.
      2. Should have been picked up in an audit.
      Serial mismanagement by HP and I cant believe Whitman is still involved. The last big three acquisitions have been EDS, 3Par and Autonomy. At last counts EDS was worth around 20% of the $14bn paid for it.
      The last sensible decision taken by a CEO was Apotheker who suggested to break the company up. A few years lost since then but its only a matter of time.

      • Leopold Stotch

        Which audit?

        The audit of Autonomy, in which there was no hardware / software split in the financial statements therefore no relevant disclosure covered by the audit report?

        Or the audit of HP in which the company’s board signed off on a deal in which management negotiated a price and conducted extensive due diligence to support their assumptions?

        If the latter I’d love to be a fly on the wall of that audit meeting: “No, the carrying value isn’t supported by the fact HP have paid that exact price three months ago based on a commercially negotiated agreement. I don’t fancy your valuation model. Write it off.”

  • Nobbie

    I find it
    impossible to believe that after « months » of due diligence costing TENS of millions of dollars carried out by top five accountancy firms and mandated facilitating banks, not to mention HP’s own bean counters who work in a top tier global enterprise, failed all the same to highlight basic differences between revenue accounting procedures in the UK and the USA. Further, I find it absurd that this these “obvious” differences, should lead to an $8 billion i.e. majority write down on the value
    of what they were buying.

    At the end of the day, at the point they were buying the company, it was valued on the UK stock exchange at half the transaction price. Using a simple DFC model to value the company, at its then rich market price, it would have had to grow profits at 23% over the next ten years to justify it’s pricing. HP paid double that price.

    I don’t know about you, but you’d have had to be pretty certain of your FACTS, to be
    taking that kind of risk, either that or completely incompetent.