Join our community of smart investors

Ideas Farm: If it ain't broke(r)

Broker forecasts can be most useful when they are wrong, while broker recommendations can be useful because they may often be right.
Ideas Farm: If it ain't broke(r)
  • Improvements to how we present broker upgrade and downgrade data
  • Broker forecasts are often at their most useful when they are “wrong”
  • Brokers recommendations may be much better job than most people think
  • Loads of updated idea-generating data

We’re keen to keep improving the data published in our Ideas Farm, and this week we have attempted to spruce up how we present information about broker upgrades and downgrades. 

The problems we’ve had with this data from the outset is that some of the large changes in forecasts are the result of a company having very piddly EPS forecasts in the first place. For example, this week the 1,683 per cent next-12 month EPS upgrade reported for Topps Tiles (TPT) needs to be seen in the context of the fact that one month ago brokers were forecasting only 0.2p of earnings over the same periods (the 12 months to 2 Dec 2021). 

The fact this represents less bad trading conditions rather than startling growth can also be seen in more context by comparing forecasts with the current share price of 56p. This is why we are also including share price date in our new version of the table. 

Ideas Farm readers* have been really helpful in making suggestions about ways to improve the table’s presentation and we’ll continue to try to improve the data further in the future. One negative is that page space in our magazine means the print version of our upgrade and downgrade tables now only contain details of changes over the last month so it can accommodate some of our data additions. However, the downloadable online version of the table, which is available for free to all registered users, continues to contain information about forecast movements over additional periods of 3 months and 1 year.

The example of Topps illustrates a quandary investors have when trying to attach significance to broker upgrades and downgrades. When earnings forecasts are very depressed, big percentage changes can result from relatively minor changes in absolute terms. Indeed, the consensus forecast for Topp’s EPS over the coming 12 months remains 24 per cent below what was expected for the same period a year ago. Meanwhile, if a company has very limited broker coverage, the simple fact that a forecast from a new broker becomes amalgamated into the consensus number can look like a major change, whereas it may just represent a new opinion on prospects.

The fact that thee signal from forecast changes is not always clear does not mean the information is not useful, though. What it does mean is that upgrades and downgrades need to inform broader research rather than being taken as a hard-and fast-signal. 

The kind of upgrade or downgrade that is really interesting for investors is the type that looks synonymous with a longer-term trend. Brokers are often consistently “wrong” with their forecasts and shares can move whenever they are forced to upgrade or downgrade next. It is important to appreciate that this “wrongness” is often for very good reasons. Afterall, we’re talking about forecasts made by very smart people with huge knowledge of the sectors and companies they cover. When forecasts are persistently off in the same direction, it can be a sign that there is a trend influencing trading that cannot be reliably factored into models. Understanding the narratives behind such trends and backing such themes when they look likely to persist can be very profitable. 

Interestingly, some recent research from Technische Universität München (highlighted this week by Liberum strategist and research sleuth Joachim Klement) suggest brokers outside the US may be cannier than many presume in with buy and sell calls, too. The research found that the most popular fifth of stocks with non-US brokers significantly outperformed the fifth of most hated. The research was based on 3.8m forecasts across 45 countries over 25 years to 2019. The annualised love/hate outperformance for UK brokers stood at an impressive 4.9 per cent. This wasn’t a patch on their Norwegian peers, though, with 14.2 per cern annualised outperformance. 

Another thing to love, not hate, about brokers.

*special thanks to Bernard Payne on twitter