Skip to main content
strategy lab

Norman Rothery is the value investor for Globe Investor's Strategy Lab. Follow his contributions here and view his model portfolio here.

There is something special about enjoying an ice cream next to the warm waters of Lake Erie on a hot summer day. My favourite purveyor of cones offers a load of flavours to choose from and seems to do quite well for itself.

But, like many seasonal businesses, its profits rise in the summer and ebb in the winter. It's a pattern seen by most companies over longer periods as the economy as a whole crests and crashes.

Sensible owners take such changes into account when evaluating their businesses. It doesn't make much sense to measure an ice cream stand's worth based only on its results over the winter months. Instead, it's better to see how it fares over at least an entire year, and preferably several.

It's why Professor Robert Shiller likes to use cyclically adjusted earnings when calculating the market's price-to-earnings ratio (P/E).

Instead of just focusing on results over the past year, he uses average earnings over the past decade to get a better sense of the market's potential.

However, the idea of using long-term earnings when evaluating companies is hardly a new one. Benjamin Graham, the father of value investing and a highly successful money manager in his own right, suggested employing a similar technique when studying individual companies.

While most market studies calculate P/E based on earnings over the past year, GMO's James Montier reported on the efficacy of using longer-term earnings in his book Value Investing.

He started with the usual P/E formulation based on trailing 12-month earnings and determined the performance of low ratio stocks using a global developed-markets database. He found that stocks with the lowest ratios beat the market by roughly two to three percentage points annually from 1985 to 2008.

The results were even better when he used more earnings data in the calculations. Stocks with the lowest ratios based on earnings over the prior decade went on to beat the market by a whopping 5 per cent a year on average.

The idea of taking such a long view is particularly attractive these days because it is easy to be fooled into thinking that the growth seen since the lows of 2009 is likely to persist. In many cases, companies are just getting back to profitability levels seen from before the crash.

The long-term view also fits well in the Canadian context because our markets contain a large number of resource-based companies that tend to be highly cyclical.

On the other hand, firms in terminal decline will often trade at low P/Es before joining Detroit in bankruptcy. It's why I tend to hold off on low-ratio stocks that have declined significantly over the past year.

One low-P/E candidate that has improved significantly over the past year is Toronto-based Norbord (NBD). The firm makes oriented strand board for the housing market and it falls into the heavily cyclical category.

The firm did really well before the U.S. housing crash of 2008, but it was smashed in the downturn. Matters were so bad that Toronto-based Brookfield Asset Management (BAM.A), which owns most of Norbord, injected more money into it via a highly dilutive rights offering.

Since then Norbord has returned to profitability, but it still trades at only seven times its earnings over the past decade, according to S&P Capital IQ. It also recently started to pay a dividend again and offers a yield near 7.7 per cent.

It's an interesting situation, but I have to admit that I liked Norbord much more when I bought a few shares in 2011. Nonetheless, it should do reasonably well, provided the U.S. housing market continues to recover.

It's something to ponder over an ice cream this summer.