Debunking those excuses for not publishing analyst consensus

Every listed company monitors analyst consensus internally. This is in line with the IR Society best practice guidance that companies should establish internally which line items need to be tracked for company-compiled consensus. This may depend on factors such as internal KPIs, market guidance and metrics of interest to the buy and sell-side. To do this, IROs should request to see copies of analyst models for review after any change of forecasts is published, in order to maintain an up to date view.

The IR Society goes further than this purely internal monitoring, and states that best-practice is to publish consensus on the website. I recently read this comment on Linked-in by veteran IRO, Andrew Ripper – he believes that only 40% of FTSE100/250 companies publish consensus on their websites.

As an Investor Relations (IR) professional, ex sell side analyst and private investor, I’ve spent a lot of time on the IR pages of corporate websites over the years. They are an important source of information, being where investors/analysts go to learn about the business, get access to reports, make contact and understand the investment case. Most cover the basics well. 

Investors also need access to consensus estimates, to understand market expectations and valuation. This is particularly important in an environment where corporates may be understandably more reluctant to provide guidance (given COVID), there is less general access to broker estimates (due to MiFID) and consensus on wire services can be hit and miss.

Considering a broad range of 100 corporates, I estimate that only around 40% of the FTSE100 and FTSE250 include company-compiled consensus on their websites, of which only a third (~15% of the total) include cash flow and/or net debt estimates, important metrics. This includes some companies which do it well, such as Rio Tinto, Vodaphone and Serco, but there’s scope for many more to join them. Putting consensus on the website should be best practice – investors will thank you for it.

It got me thinking. Consensus is such a basic tool for investors that putting it on your company website should be an easy best-practice win so I struggle to understand why you wouldn’t. Yet many companies don’t. Over the years, I’ve heard numerous reasons/ excuses, and I wanted to take this opportunity to debunk some of them.

• We only have a few analysts covering us so it’s not meaningful. True, the fewer contributors, the less of an average it is, but it is still relevant. Institutional investors may have access via Bloomberg/ other third party providers, but the retail investor doesn’t. If you’re really worried about the number of analysts that contribute to consensus, just put a footnote to say “This is based on forecasts from [x] analysts.”
• It’s too complicated. Apart from being a bit rude about your investor base (!), it’s the job of an IRO to make the investor information accessible to their audience. That could mean focusing on KPIs, adding relevant industry specific metrics/ ratios, or adjusting numbers to put them on a consistent, meaningful, basis. In fact, it’s arguably more important that you disclose a company-compiled consensus in this situation as the third-party providers may not be generating a consensus that is accurate/ particularly helpful. Don’t forget to footnote what you’ve done for transparency.
• The analyst models are out of date. Again, it’s true that some analyst models are not updated as often as IR would like, but they are still out in the market so if this is the case, I would argue it is helpful to investors clarify the issue. I’d generally suggest that numbers published over a year ago can legitimately be excluded from the consensus calculation. If you’re doing this, it’s best to set a clear policy for calculating consensus, disclose it, and apply it consistently. Then there is no risk of being accused of manipulating the numbers to generate a particular outcome.
• Outliers. This is probably the trickiest issue to handle. Often an analyst has intentionally taken an extreme stance, and stands by their number, so (whether you like it or not) it is not necessarily appropriate to exclude it from a company-compiled consensus. One solution in this situation would be to give a range, and maybe also show mean, mode, and median.

The only other thing I’d add, and I’m sure it’s obvious, is that once you publish consensus, you need to keep it up to date.

I’d love to hear your thoughts on publishing consensus.