Hewlett Packard's surprising announcement of accounting irregularities at Autonomy caught the market by surprise on Tuesday and led to a nearly 12 percent decline in the company's stock. But Autonomy's accounting had been questioned by analysts years ago.
Paul Morland, technology research analyst at broking and advisory house Peel Hunt, told CNBC that he had noticed three red flags in Autonomy's accounts in the years leading up to the HP acquisition: poor cash conversion, an inflated organic growth rate, and the categorizing of hardware sales as software.
Indeed, Morland said that in the six reports he had produced since 2008 in which he had mentioned Autonomy, the U.K.-based maker of data analysis software, he had mostly recommended selling the stock.
"There were periods when I wasn't a seller," he told CNBC on Wednesday, saying that his work as an analyst meant he had to be mindful of what the share price was discounting at the particular time of analysis — but his opinion changed in 2008.
"Sometime in 2009, I began to find out about the things we've been talking about and I moved towards a more negative stance. … I had a 'sell' recommendation on the stock for most of the three years leading up to the deal."
Mike Lynch, the former Autonomy CEO, declined to comment, but his spokesperson pointed to his previous comments on the issue. Lynch told CNBC on Tuesday that he was shocked by the allegations and blamed HP for mismanagement.
In an analyst note published by Morland in June 2009 titled "Accounting Red Flags," when he worked at Astaire Securities, he wrote: "Although investors do not have access to the same detailed information as auditors, there are plenty of analytical techniques that can be used to help identify when a company's performance might not be quite as good as it seems."
Morland said he had specifically questioned the company's cash conversion ratio (the ratio of cashflow from operations and EBITDA), which seemed lower than other software companies.
"[Autonomy] countered that by saying that they were high growth, which can absorb working capital and make conversion worse … my model suggested that it was still lower than it should have been even though they were growing higher," he said.
"And now we know — we think we know — that they weren't growing as fast as they said they were, and therefore the cash conversion should have been even better than I thought it should have been at the time," Morland added.
"There were only a few of us writing this sort of stuff on the accounts," he said, noting that JPMorgan's IT services analyst Daud Khan had also questioned the company's accounts.
Morland told CNBC some investors were paying attention to his reports before the company's $11.5 billion acquisition by HP in 2011.
"People were listening. The reason I know that is because, at the point of acquisition, the shareholder base changed significantly during the course of the previous two and a half years," he said. "A lot of the U.K. institutions had sold out, who were my primary audience and they'd been replaced largely by U.S.-based shareholders who perhaps weren't getting access to the same sort of research."
Noted short-seller Jim Chanos, president of Kynikos Associates, said Autonomy's accounting problems were apparent and there were many sell-side analysts who had been skeptical. "There were all sorts of cookie-jar accounting ... that appeared to be going on at Autonomy," he said. (Read More: How Jim Chanos Spotted the HP Scandal)
Morland insisted that he was not accusing Autonomy of fraud.
"Unless you're an accountant it's difficult to spot most of the things that were going on, quite frankly," he said, adding that those who got it "right" hadn't necessarily profited from their analysis.
"It's frustrating that the people who got this right didn't make any money out if it. Investors who were in Autonomy made a lot of money because HP paid a massive premium — but they got it wrong," he said. "The hedge funds that shorted it were right to short it, but they lost a lot of money when the price went up [around] 78 percent when the deal was announced."
Among those who lost money was Chanos, who told CNBC the firm's European fund had been short Autonomy's stock and that he "watched in horror" as the firm was bought by HP at a premium.
But Morland said he has learned some lessons from his experience covering Autonomy.
"Stick with your beliefs and what your analysis is telling you," he said. "A lot of other analysts must have had their suspicions but didn't really speak out about it. I think it's right to speak out about it."
—By CNBC's Holly Ellyatt and Deepanshu Bagchee @DeepCNBC