Here's a healthy reminder not to rely too heavily on expected earnings figures.
If we can't take company accounts at face value how can we be expected to trust earnings forecasts?
Just 14% of forecasts are accurate
A few years ago, Parson Consulting (now Ineum Consulting) did a study on the accuracy of earnings forecasts. It reckoned that just 14% of the FTSE 350 companies' results were within 1% of the consensus annual forecasts in 2005. What's more, only 11% of forecasts were this accurate in 2004.
You may think that +/-1% is being a bit tough on the analysts that make these forecasts. So let's widen the net to +/-10%. Over a third, 34%, of forecasts were out by more than this amount.
Perhaps somewhat surprisingly, the report found that annual estimates were typically too conservative, with 55% of analysts underestimating results. But then, in the happy days of 2005, we were in the early stages of a bull market. Perhaps analysts hadn't quite adjusted to this yet, and were still being a too gloomy about the speed of the recovery. Or maybe the guidance they were getting from the finance directors of the companies concerned was on the cautious side.
However, the report also revealed that inconsistencies varied from one year to the next, so it was quite common for forecasts for a particular company to be too high one year and too low the year after.
Interim estimates are even worse
The study also observed that interim estimates tended to be far more inaccurate than annual ones. Seasonal variations could be a factor here and the fact that many businesses have inherently 'lumpy' revenue streams. Over the course of a year, the lumps may even themselves out but quarterly results may not be so smooth.
AstraZeneca (LSE: AZN) provided a recent example of this phenomenon, having beaten expectations for the first quarter of 2009 by 12%. That's a lot for one quarter. It shows there's a great deal that analysts can't predict. In this instance, they failed to see the effect exchange rates would have, and that rivals would struggle to copy a couple of AstraZeneca's drugs.
Another recent example is that analysts failed to see how badly HBOS would impact profits at Lloyds Banking Group (LSE: LLOY) in 2008, even though it was only brought onto its books for the last few months of the year. But then Lloyds was rather taken aback as well and this was a period of exceptional turmoil in the markets.
Nothing new: do your own research
All the companies in the FTSE 350 are big firms that are followed by a lot of analysts, yet the consensus forecasts are still pretty unreliable. It emphasises once again the importance of doing your own research, or rather that you should not rely too heavily on the research of others.
Of course these analysts work outside the companies they are making forecasts for. So they rely heavily on guesswork, assumptions and steers from the company itself. As we all know, even those closest to a company's operations won't always be 100% accurate.
The economy can change rapidly too. While some companies have relatively secure revenues that are based on long-term contracts, other businesses, high street shops for example, can see demand melt away very quickly indeed.
It's also worth bearing in mind that the profits of smaller companies tend to be a lot more volatile than those companies that populate the FTSE 350. For example, small companies are often overly reliant on a few customers. If one of them takes its business elsewhere, the effect on profits can be significant.
Most people consider it a necessary evil to take earnings forecasts into account, which is why the forward P/E ratio is such a popular valuation measure. But always remember it's little more than a best guess.
More on this subject:
