At this time of year, it is only natural for investors to try and make forecasts about what will happen over the next 12 months – just as it is only natural for us, as contrarian investors who prefer an outlook that is five or even 10 times longer, to do the complete opposite. So here we go with our now traditional annual anti-forecast for the year ahead.
For our 2017 vintage, we take as our starting point an article in The Economist that puts forward the not wholly startling argument broker share analysis can be notoriously optimistic. Neatly entitled ‘Discounting the bull’, it quotes FactSet data showing 49% of firms in the S&P 500 index rated as ‘buys’, 45% as ‘holds’ and just 6% as ‘sells’. Over the past year, it adds, 30% of S&P 500 businesses yielded negative returns.
The article goes on to suggest a number of reasons why the inhabitants of the sell-side sector, the brokers and investment banks selling research, tend to be so wide of the mark – for example, that they can work backwards from their analysis of similar companies; that few people like to go against the crowd so they rarely stray too far from their peers’ calls; and that they can often grow a little too close to the companies they are covering.
None of which will be much of a surprise to regular visitors to The Value Perspective, who will have come across similar arguments – and similarly unflattering conclusions – on this site. Indeed, the Morgan Stanley graph The Economist uses to show how many iterations Wall Street earnings forecasts go through from first appearance to final revision is an updated version of one we ran in this article back in 2012.
The Economist goes on to suggest the very predictability of the errors in analysts’ forecasts could be informative in itself and, while that struck us as perhaps one leap too far, our attention was caught by a strand of the article’s analysis. The numbers it crunches cover a respectable timeframe – from 1985 to 2015 – and involve an assumption as to the market’s long-term growth rate.
The article notes: “Simply taking the market’s earnings figures from the previous year and multiplying by 1.07 (corresponding with the stockmarket’s long-run growth rate) can be expected to yield a more accurate forecast of profits more than a year in the future.” Or, to put it another way, a study of 30 years suggests a simple multiplier does a better job of forecasting future earnings than Wall Street’s finest.
Now that is all quite interesting but value investors have their own 30 years of data – plus another 100 years more on top of that – to demonstrate that value has outperformed on a long-term basis. In other words, from 130 years of history, not only do we know the market tends to make 6% or 7% a year on average, we also know value tends to make around 2% a year on top of that. Of course this performance cannot be guaranteed in future.
We repeat – value investors do need to take a long-term view of at least five and, preferably, 10 years. And yet, once that caveat is made, the above ‘value premium’ does rather beg an important question: just how many investors really need to worry themselves about what analysts and other market-watchers reckon may – or, just as likely, may not – happen over the coming 12 months?
As a postscript, we should point out that, with the exception of ‘herd mentality’, we have barely even touched on the maelstrom of behavioural biases – the ‘narrative fallacy’ we discussed in Three important investment lessons from 2016, overconfidence and so forth – that always play into people’s outlooks for the year ahead. One for the 2018 anti-forecast perhaps …