Foggy metrics obscure value of cloud companies

Cloud computing companies have been in hot demand on Wall Street this year, winning over investors with their rapid growth and dependable, subscription-based, business models.

The BVP Emerging Cloud Index, led by companies such as Adobe, Salesforce and Zoom, is up 84 per cent this year, compared with a 36 per cent rise in the Nasdaq and a 13 per cent increase in the S&P 500.

But investors and other financial experts are warning that the seemingly bulletproof revenue streams and future order books of cloud companies are not always what they seem.

A new revenue accounting regime has made it hard to assess the performance of the sector and key yardsticks that are widely used to value cloud businesses are based on measures that leave plenty of room for companies to pick the most favourable treatment.

The difficulties stem from adapting traditional accounting rules to a new generation of companies that sell services over contracts that are several years long, said Tien Tzuo, chief executive of Zuora, a cloud company that supplies software for running subscription-based businesses.

He and other cloud software experts described the resulting discrepancies as teething problems, rather than signs of deeper financial issues. But with many cloud companies now trading off sky-high valuations, there is little room for error.

Recognising future earnings

At the heart of the problem is working out how to account for the revenue generated through cloud software contracts, and to assess how much future revenue investors can rely on.

Cloud companies usually sell bundles that include upfront work, such as implementing a new software system for a customer and training staff, alongside multiyear subscriptions to their products.

Investors have rewarded businesses that can show dependable future revenues, incentivising companies to characterise as much revenue as possible as repeat business, noted Ravi Kumar, a partner at Connor Group who advises investors on accounting issues. 

“People are trying to pull every part of the value proposition into software,” said Byron Deeter, a partner at Bessemer Venture Partners, a venture capital firm that invests in cloud start-ups. VC investors are usually able to dig deep into a company’s contracts and can make their own assessments, he said, though stock market investors get less disclosure.

When to book earnings?

A related problem stems from determining what part of a contract has been fulfilled in a given period — meaning revenue can be booked, rather than deferred to a future period. Companies selling three-year contracts may split the revenue equally, or vary the amount based on the value they think has been delivered in a given year. Newer cloud companies often lack the procedures to precisely assess how to allocate revenue from complex contracts, said Mr Kumar — a problem that could spill over on to Wall Street as the high demand for cloud stocks leads companies to go public at an earlier stage. 

The timing of revenue has become particularly acute for software companies with mixed businesses, selling both cloud services and traditional “on premise” software. A new accounting rule that came into force for public companies two years ago, ASC 606, forces companies to book all on-premise sales immediately, even if the business relationships extend long into the future.

That has penalised companies with more on-premise business and could produce distortions as they seek to channel more of their sales into the newer, more highly valued, cloud segment, said Tomasz Tunguz, at Redpoint Ventures.

This marks a complete reversal from the way a previous generation of software companies was judged. Accounting scandals of the past — like the one at Computer Associates, which led to a 12-year prison sentence for its CEO — often stemmed from efforts to report revenue immediately, when it should have been spread into future periods.

“In the old days, the regulators wanted to make sure you weren’t booking too much revenue up front: now, they’ve completely flipped,” said Jason Child, a former head of finance for Amazon’s international operations and now chief financial officer at Splunk, an analytics software company.

This has yet to lead to high-profile scandals, though some complaints have spilled over into public view. Oracle recently asked a judge to dismiss a lawsuit that was seeking class action status over claims that it had required customers buying its traditional on-premise software to also take out cloud contracts, in order to exaggerate that part of its business.

Informal measures of success

Meanwhile, the focus on long-term revenues has led Wall Street to focus on a set of informal measures, outside of accounting principles, which are applied “extremely differently” and create inconsistencies “when you’re comparing one company versus another”, said Mr Kumar.

One of these new metrics, known as net dollar retention, seeks to measure how much business a company generates in a given year from customers it had the year before. For instance, a business that lost customers that accounted for 5 per cent of its revenue, but generated 15 per cent in additional sales from those it retained, could claim a net dollar retention rate of 110 per cent — a sign of a healthy, expanding business.

Differences in defining who should count as a “new” customer, however, makes this calculation far less clear-cut than it might appear, said Mr Deeter. Some software companies count new sales to different parts of a large group as being to the same customer, while others treat different subsidiaries, or different geographies, separately.

Some customers are also left out of the calculations. Box and Zendesk, for example, exclude their smallest customers. Zoom last week reported a strong 130 per cent net dollar retention rate for its largest customers. But it excluded smaller customers that now represent 38 per cent of its revenue.

Calculating the churn rate

Another imprecise measure is the level of churn, or the rate at which customers fail to renew contracts when they expire. Most cloud companies do not disclose this figure as a matter of course, though it is commonly reported at the time of an IPO.

How the figure is calculated can make a big difference. For instance, a company that started the year with revenue of $50m and lost customers worth $10m, but added enough new business to get to $100m, could report a churn rate of either 20 per cent (the proportion of its original business it lost) or 10 per cent. 

“If you’re a high-growth company, you can probably hide attrition for a while,” said Mr Child at Splunk. Last Thursday, Splunk’s shares fell by almost 26 per cent after it reported lower than expected revenues.

Is revenue really recurring?

Wall Street’s most popular new yardstick for assessing cloud companies, known as annual recurring revenue, also leaves room for interpretation. A measure of how much revenue a company can count on each year, this has become the gold standard for subscription businesses.

But it is not always clear what should count as “recurring”. If customers have the right to terminate a three-year contract after the first 12 months, for instance, it can lead to differences of opinion, said Mr Kumar.

This has led some companies to amend their contracts, he added, “bundling and repackaging” the same business differently to produce a different financial treatment, without making any underlying changes to their business model.

The lure of ARR has spread well beyond the software world. Companies from other sectors that can claim a high level of guaranteed repeat business have also started to adopt the term.

They include online pharmacies such as Alto Pharmacy, which can point to customers with long-term health conditions that led to repeat prescriptions. The start-up, whose investors include Japanese tech investor SoftBank, boasts that it is “on track to achieve $1bn in ARR by 2021”. In theory, though, customers of businesses like this could take their prescriptions elsewhere at any time, said Mr Kumar.

Despite the pitfalls, investors like Mr Deeter say that Wall Street’s enthusiasm for cloud software businesses is well founded. Their subscription businesses are inherently more stable than the sales of one-off licences that software companies used to rely on, and the new metrics provide a better way of judging them than what came before.

With stocks in the sector priced for perfection, however, that still leaves plenty of room for disappointment.

Source link

Related Articles

Back to top button