Price/earnings multiples

Ft.com
The problem with price/earnings multiples is that they are simultaneously spurious and essential.

While a measure calculated by dividing a company’s market capitalisation by its annual net income has questionable predictive power, its ubiquity as valuation shorthand makes it hard to ignore. If investors are to join the crowd in using it, however, they should bear a few rules of thumb in mind.

First, a p/e multiple is a tool for comparison. It must be placed next to peers, the market and a stock’s own growth and valuation history. So second, make sure the comparison is valid, both in terms of time period under consideration and the accounting standards used.

That latter part leads to the third rule: beware of adjustments. A company might like to present profits without reference to significant business costs such as paying for staff or technology, but this is a poor basis for analysis. Giving stock away will be a cash cost in the future, so include it and use the most conservative earnings number produced by accounting standards.

The same goes for touting an ex-cash multiple for a company hoarding cash. The point of looking at an income statement’s bottom line is to capture the effects of taxation and capital structure. Ignoring a cash pile’s low return defeats the purpose, as does comparison to stocks and the market without donning the same eye-patch.

Fourth, ignore spurious accuracy. Shiller’s cyclically adjusted p/e incorporates ten years of historic data. A prospective p/e multiple by comparison is an inherently loose number, based on estimates that themselves change on a regular basis. Introducing a decimal point awards undeserved authority and accuracy to something that should be treated as an educated guess. Treat a p/e as shorthand for what the market might be thinking, not as a short cut to serious investment decisions.