Ideas Farm: Profit from predictable mistakes

A great illustration of how off-the-mark forecasts commonly are comes from a study by the great contrarian investor David Dreman, which was subsequently updated in 2008 by investment firm Star Capital (a screen based on Mr Dreman’s deep-value investment approach can be found in this week’s stock screening column). The joint Dreman/Star study covered 1.5m forecasts spanning 33 years. It found that over the course of 12 months the average forecast was off by 30 per cent - ie very wrong.

At this point, it may be tempting to conclude that broker forecasts are useless. Indeed, this year’s coronavirus crisis illustrates the futility of trying to predict the future; especially so for those employed to very precisely quantify what sales and profits will be achieved by individual companies in very specific time frames. 

Many events that are much less high-profile than coronavirus but are similarly unpredictable tend to jolt individual company forecasts up and down over the course of any 12-month period. 

However, as well as what Donald Rumsfeld famously referred to as “unknown unknowns” – events no one saw coming and no one could quantify – brokers also have to grapple with his “known unknown” – events and trends that can be seen coming but are still impossible to quantify. It is the efforts of brokers (generally extremely intelligent, considered and experienced people) to quantify the “known unknowns”, and the pressures on them to underestimate their significance, that can be regarded as of real value to investors. 

What behavioural psychologists term “anchoring”, often causes forecasts to remain too wed to the status quo. But not only is it hard to imagine a future that is significantly different to the present, it is also dangerous for brokers to predict it. For those analysts bold enough to make predictions that stand out from consensus forecasts, there can be genuine career risk. 

Bold forecasts are likely to attract ridicule at the time they are made and derision if they are not reasonably accurate – the Dreman/Star analysis shows inaccuracy is always a strong possibility. In the words of the great economist and investor John Meynard Keynes: “It is better for reputation to fail conventionally than to succeed unconventionally.”

The consequence of this is that incremental changes to broker forecasts can often resemble more of an upward or downward trend than a random walk. This can reveal a positive or negative story unfolding and sometimes can be far more indicative of a long-term trend than the consensus view. 

The table we publish in the Ideas Farm of consensus-broker upgrades and downgrades hopes to highlight possible trends that are either emerging or are already established. As always with the information on the Ideas Farm pages, the data only provides an indication that something interesting may be happening, and only research can really suggest if a decent investment case exists.

Three of the shares examined in our Tips section this week have shown noteworthy recent forecast trends. Earnings forecasts for bike and car parts retailer Halfords (HFD) have moved sharply higher after many years of decline, and there are some reasons to think this could prove the start of a positive trend. Meanwhile, downgrades at TI Fluid Systems (TIFS) and Wood Group (WG.) – a share that also appears in our table of most shorted stocks – may be a genuine cause for concern. 

Click here to download the tables in PDF form

Source: S&P CapitalIQ

Source: S&P CapitalIQ

Source: FCA

Source: S&P CapitalIQ

Source: S&P CapitalIQ

Click here to download the tables in PDF form