Ten years ago, the firm discovered that analysts were on average much too bullish about corporate prospects, slow to revise their forecasts to reflect new economic conditions, and prone to making increasingly inaccurate forecasts when economic growth declined. Three of its researchers – Marc H Goedhart, Rishi Raj and Abhishek Saxena – have just updated that work. Despite all the rewriting of the rules after the dotcom bubble burst and Sarbanes-Oxley, they found that nothing has really changed.
In fact, analysts have been over-optimistic for the past 25 years, with profits growth forecasts ranging from ten to 12 per cent a year, compared with actual growth of 6 per cent, McKinsey calculates. Actual earnings growth was higher than predicted in only two instances, both during the earnings recovery following a recession (when improvements are always better than battered and dejected economists predict). On average, earnings forecasts have been almost 100 per cent too high, a ludicrously bad record.
The only times the actual earnings of S&P 500 companies actually coincided with analysts’ forecasts was in 1988, from 1994 to 1997, and from 2003 to 2006. That’s just seven out of 25 years. There is some good news: capital markets are not as exuberant as analysts (in other words, those who buy and sell equities don’t believe everything the analysts tell them).
Except during 1999–2001, actual price-to-earnings ratios have been 25 percent lower than implied p/e ratios based on analyst forecasts. Why analysts are so exuberant is unclear – in some cases, they may be indirectly under pressure but most times they appear to fall prey to collective delusions. I don’t like pure psychological explanations, however: bubbles are driven by economic forces, primarily cheap money, so there must be something of that nature at play here too.
There are caveats: the research looks at the S&P 500, not the FTSE 100. It may be that London-based analysts aren’t as over-optimistic – though even if this is true, they are almost certainly pretty over-exuberant too. The reality, which is often forgotten, is that not every company can grow faster than nominal GDP for ever – and that means a broad measure of the market ought to grow by about 5 per cent a year on average over the long term, excluding growth derived from overseas. In good years or after recessions, of course the market (and most of its components) will massively over-perform – but such years are rare. Yet McKinsey’s analysis of five-year earnings growth (March 2005 to 2010) suggests that analysts were forecasting annual growth of more than 10 percent for 70 per cent of S&P 500 companies. There was never really any chance of that.
The message is clear: ladies and gentlemen analysts, it is time to get real. Don’t be too optimistic, be realistic – for your own sake as much as that of those who consume your research and take crucial decisions on the back of it. The analysis and number-crunching of corporate results is essential for the good functioning of markets and the proper allocation of capital; it is also key for buy-side firms, such as fund managers. I’d be keen to hear readers’ views (in confidence) on how to improve the system.