All companies are “average”

Geoff Blount, Managing Director of Cannon Asset Managers, argues that over the long term, company performance tends to be average and is driven more by economics than other factors.

newspaperAs humans, we generally like to think we are above average, especially corporate management teams. So stating that all companies are “average” over the longer term might raise a few eyebrows among company owners and their management teams. To give this comment some context, we are referring to the ability of companies to sustain above-average rates of earnings growth over time. Indeed, the evidence suggests that this is an incredibly difficult task, and company management may have an uphill battle in overcoming the odds.

Cannon Asset Managers conducted research into the trajectory of listed company earnings in South Africa on rolling ten-year periods, from 1996 to 2013, with interesting results. At the start of each ten-year period surveyed, we broke down the market into ten baskets (deciles) of shares categorised by their last three years’ annualised earnings growth in real terms (after inflation), from lowest to highest. The chart below depicts our results.


Click to enlarge

For example, decile one (in red) contains the 10% of companies that had the worst earnings growth in the previous three years and decile ten (in green) is the basket of shares that had the best earnings growth over the preceding three years.

We then tracked the average rolling three-year real earnings growth of each decile over the next ten years. The companies in decile ten typically enjoy an average earnings growth of 65% in the previous three years but subsequently trended down rather quickly before “normalising” after three years. On the other side of the spectrum, companies in decile one start with earnings growth of –30% in the previous three years, but then their earnings swiftly rose for the next three years and then remained elevated for a number of years before trending back to average.

From this chart it is apparent that companies that have enjoyed strong recent earnings growth revert to the mean, and equally, the recent laggards experience a positive and sustained upwards earnings recovery, before reverting to the mean. This picture plays itself out over all the 10-year periods we surveyed.

But if companies generate earnings growth that is similar over the long term, it begs the question “how much should investors pay for these earnings?” Typically, investors pay more for decile nine and ten shares because of their recent earnings experience (and they are ultimately disappointed), and investors shy away from decile one and two stocks for the same reason, driving their prices down (and they are ultimately pleasantly surprised by their “unexpected” good performance).

What drives this phenomenon?

And what drives this phenomenon of earnings mean-reversion? Counter-intuitively, it is not strategy or market share, product innovation or neat synergies that ultimately drive earnings. Rather, for the vast majority of companies, the environment in which a company operates dictates how they perform. Across the 25 countries that Cannon has analysed, GDP growth has a strong correlation (0.70 or higher) with earnings growth. Over time the economic result causes the earnings result, and not vice versa. This is even more the case as companies grow larger.

However, sometimes mean-reversion takes longer than expected – we are in one such phase where many decile nine and ten stocks are enjoying extended growth and even stronger price momentum, pushing them to increasingly expensive multiples. However, to our mind, this is unsustainable, although it is easy to find many reasons which justify owning these popular yet expensive stocks at the moment.

Conversely, it is also easy to find reasons to avoid the unloved decile one and two stocks, whose recent performance continues to disappoint.

However, past history would indicate that earnings do mean-revert, and that the words “this time it’s different” are seldom justified.

Leave a Reply

Your email address will not be published. Required fields are marked *