There’s plenty of value to be found in producing your own internal consensus estimates of sell-side forecasts, reports Alex Jolliffe
Producing your own consensus forecasts can be painstaking. Digging through analysts’ notes and models to ensure you are aggregating comparable figures is time-consuming. But for many IROs it’s worth the effort; and new tools are helping to ease the burden.
Consensus figures are crucial to IR. When Jill Sherratt, interim head of IR at UK services company Serco, writes a paper for the board, she includes the latest consensus and compares it with the last quarter. She also writes about its future direction.
‘I talk to the CFO about this at least once a month,’ she says. ‘That conversation is absolutely vital.’
Consensus also kick-starts conversations with the sell side, Sherratt says. ‘You won’t have a single conversation with an analyst without discussing it – it’s your critical output.’
Oliver Schmidt, head of IR at Allianz, the financial services group, says consensus can show him when to explain a point to the sell side.
‘If the analyst calls me and is 10 percent above or below the consensus, we discuss it,’ he says. ‘He or she may have overlooked a one-off that we have already disclosed.’
Many IROs are dissatisfied with the accuracy of consensus estimates provided by traditional third-party aggregators, however.
The numbers they provide often do not compare like-for-like figures, because analysts produce figures differently: for instance, some include exceptional costs when predicting revenue, others do not.
Critics say third-party aggregators also include out-of-date numbers.
‘The quality of analysis can be inconsistent, in my experience,’ observes John Dawson, head of IR at National Grid and chair of the UK’s IR Society.
This has added to the impetus for IR departments to produce their own house view of consensus. This way, they can be sure the numbers are reliable; they can also be more confident in discussing consensus with the market.
Dawson has followed the same routine at his last three firms: four times a year, ahead of quarterly results, he sends out a template to analysts to fill in with their latest published data.
The results are then published about a week before results, showing the average, minimum and maximum for each figure. Dawson also publishes the number of analysts who contributed to each line.
Labor-saving devices
The benefits of producing internal consensus figures are clear, but there’s a problem in the amount of work it takes to complete the job. Whether a company sends out a template, or digs up the numbers itself from analyst research, the process involves a lot of time and effort.
A survey by the UK’s IR Society released in March this year found that time constraints often prevent companies from getting all the figures they want.
Given the difficulties, it’s unsurprising new tools are being developed to help companies with the slog involved in producing consensus estimates.
London IR firm Temple Bar Advisory has recently launched Analyst Consensus Estimates (ACE), a web-based platform IROs can use to request, receive and review analysts’ estimates at high speed.
They can also store historic forecasts and use integrated tools to interpret and evaluate the data they collect.
Separately, Vuma Consensus, a London-based technology firm, has produced a web-based system that enables analysts to input their figures online using a template devised by the company.
Transparency is added to this process by an audit trail that shows when analysts entered their figures.
James Hamilton, an analyst at broker Numis Securities, sees potential in an online system. He points out that there is particular scope for help with acquisitions, which analysts often treat differently when producing forecasts.
Another option for IROs is to receive consensus figures from boutique data providers. One example is Sweden-based Inquiry Financial, which stays in direct contact with analysts and tailors data collection for each company it covers.
Dawson says he can see the benefit of outsourcing the compilation of consensus estimates for firms that have stretched resources.
But he believes companies that can do the work internally should do, because it will enable them to retain more knowledge in-house about how different analysts work.
Once the consensus has been nailed down, the next question is what to do with it. Some just use the numbers for internal reference; others release them to the market through phone calls, via email or publish them on the IR website.
In fact, wide distribution is growing in popularity: the IR Society’s survey found a ‘significant increase’ over the previous 12 months of companies either posting consensus figures online or distributing them through a widely circulated email.
Sanofi’s solutionWhen Sébastien Martel joined Sanofi, the French pharmaceuticals company, in September 2008 as vice president and head of IR, he noticed a troublesome phenomenon. Every quarter, with Sanofi’s earnings announcement approaching, Reuters, Bloomberg and other wire services would post consensus estimates that were sometimes way off the mark.‘Some analysts were giving sales estimates, others revenue estimates, so the consensus could be an average of apples and oranges,’ Martel explains. Sanofi’s royalty revenue – which is significant – was often left out of the reckoning altogether. Martel’s solution? Every quarter, five days before earnings are announced, he sends the wire services a list of preview notes published to date (typically a lengthy list given Sanofi’s 40 or so sell-side analysts) along with official top-line and bottom-line consensus estimates. Today the wire services covering Sanofi are used to receiving the company’s own calculations, and the confusion around consensus has come to an end. |