How did HP lose $8.8bn?

Tim Bennett looks at what went wrong with Hewlett-Packard's $11bn purchase of software company Autonomy, and lists the four key lessons for investors.

Almost from the day it was founded in 1996, Mike Lynch's software company Autonomy was hailed as a model British technology firm. It quickly made a name for itself, creating software for searching through data from texts to voice mails to video. When American giant Hewlett-Packard (HP) bought it for $11bn in October last year, it seemed that this British success story had come to a very happy ending indeed.

But now HP claims in accusations that are firmly denied by Autonomy that the firm made "outright misrepresentations" to inflate its financial results, leading HP to overpay. HP has taken an $8.8bn write down on its investment, blaming "Autonomy's misleading accounting" for around $5bn.

When is a sale a sale?

So what went wrong? Many of HP's allegations centre on revenue recognition'. Let's take a simple transaction, such as buying a Christmas present for £50 online. The seller will need to record the sale but when? Should it be in its profit and loss account when the order is received? That's probably too early you might change your mind and cancel. What about when the goods are shipped and paid for? Sounds fine but what if the goods don't show up, or you return them? You could wait until the customer had lost all reasonable right to return but that could take months.

Subscribe to MoneyWeek

Subscribe to MoneyWeek today and get your first six magazine issues absolutely FREE

Get 6 issues free

Sign up to Money Morning

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Sign up

Each method has its pros and cons. The point is that accounting rules don't specify exactly when you should record (or book') a sale. Many firms would choose the invoice date. But others could book it sooner (aggressive' accounting) or later than that, and still act within the rules. In this case, HP is accusing Autonomy of highly aggressive booking of income on software contracts, which in turn would have exaggerated the firm's growth rate, convincing HP to pay up for the company.

Lessons in aggressive accounting

One deal being examined, according to The Wall Street Journal, involved Tikit Group, a supplier of software to accountants and law firms. It seems that Tikit bought around £4m of stock from Autonomy. Autonomy recognised the whole lot as sales, even though Tikit would only actually pay for the software once it had been billed to one of its clients. Autonomy argues that accounting rules allow it to book a sale once the stock has been shifted to a re-seller rather than when it is sold to an end client. The issue will no doubt be debated in court.

Another area of contention is whether Autonomy tried to mask hardware sales as software sales. For example, says Daniel Fisher in Forbes, say a customer at one of the data storage firms Autonomy took over had $5m of cost and four years left on a data storage contract. Autonomy would offer the same deal over the same period for $4m, but structured as a $3m purchase of software, paid upfront, and $1m of storage. The $3m would be booked as software sales immediately.

A less aggressive approach would be to spread the $4m over four years instead. Autonomy is also said to have preferred to "capitalise" rather than "expense" software development costs. Say the firm spent $100m on software development costs this would be added to an asset in the balance sheet rather than charged against profits immediately. The logic is that those capitalised costs would eventually hit the profit and loss account as future revenue came through. This had the benefit of making Autonomy's balance sheet look bigger than perhaps it really was. A less aggressive approach would be to expense a much higher proportion of those costs as they were incurred.

Autonomy will fight back

Lynch claims that HP has got it "completely and utterly wrong" and that his firm simply obeyed the accounting rules. He says HP's case is based on "a lot of nitty-gritty about small amounts of revenue on certain deals". He also notes the auditors signed off on the accounts, although this may simply mean the auditor has a case to answer separately. Lynch has written an open letter to HP's board, rejecting "all allegations of impropriety" and demanding a breakdown of the $5bn write-down. Others argue that it was clear at the time that HP had overpaid the original deal involved $6.9bn in goodwill', which is the difference between the firm's tangible asset value and the actual price that HP paid. Whatever the truth of the case, there's no doubt the deal has been a disaster for HP shareholders. I look at how to avoid making similar mistakes below.

Four key lessons from the HP debacle

What can you learn from this debacle?

Follow the cash, especially when looking at fast-growing firms. Revenue can be fudged, but cash can't so easily. So when operating profits and operating cash flows part company, watch out.

If you can find it, look at the note to the accounts called accounting policies', specifically the bit titled revenue recognition'. By comparing a firm to its peers you can get a feel for whether they are being too aggressive or not.

Don't put your full faith in the auditors, particularly when it comes to riskier, fast-growing firms. The best investors will take a quick glance at the audit report, but then do their own homework, something it seems HP may have overlooked.

Only buy shares in firms you understand. To this day no one has been able adequately to explain to me what Autonomy's software does that Google's can't which is why firms such as Autonomy are not for me.

Tim graduated with a history degree from Cambridge University in 1989 and, after a year of travelling, joined the financial services firm Ernst and Young in 1990, qualifying as a chartered accountant in 1994.

He then moved into financial markets training, designing and running a variety of courses at graduate level and beyond for a range of organisations including the Securities and Investment Institute and UBS. He joined MoneyWeek in 2007.