What more is there to tell?
I first became interested in investing my own money in the stock market over ten years ago. Bored by a comfortable but limited existence at Deutsche Bank, I took a year out to do an MSc in Investment Analysis. While I was looking round for a dissertation topic I read Graham and Dodd’s 1934 classic Security Analysis. One of their suggestions was that you should not judge a company’s earnings potential over just the last year, but take a longer term view of 7-10 years to allow for fluctuations in the economy as a whole. This seemed such a common-sense suggestion, and yet after searching through dozens of academic papers, as far as I could tell no-one had thought to test Graham and Dodd’s assertion in the 70 years since. Being a new idea to the academic world the idea took a bit of explaining to my supervisor, but the results from the limited tests I was then able to do were promising.
Since then I have gone on to do a PhD in how to improve the P/E ratio as a predictor of investor returns, some of which is updated and summarised here.
The long-term P/E and decomposing the P/E are, I feel, surprisingly rich and complex stories. I have also taught much of this book’s contents to some thousands of undergraduate and Masters students at Durham University and the University of York, a few of whom I hope have shared my enthusiasm.
But the story of the P/E is not a dry academic investigation. It is a practical guide to how any stock market investor can put an improved P/E to use.
The faults of the P/E
Even a novice private investor should know that the P/E has some glaring faults. A low P/E may mean a company that has been marked down for no readily apparent reason, and thus an attractive value investment for those with the patience to wait while the market re-values it. However, the P/E is a backward-looking measure, based on one large number (sales last year) minus another large number (total costs last year). Just because the company earned £1 per share last year doesn’t necessarily mean it will earn anything like that for the foreseeable future. A low P/E can also easily mean a company that is deservedly cheap because it is in financial difficulty, and that the P/E is likely to become cheaper yet or the company even go into administration.
From a different point of view, in the world of academic finance the P/E has been largely forgotten as an investment ratio for researchers since the early 1990s. Its power to identify cheap shares that will perform was first shown in 1960, but in the last 15 years few papers have been written on it. The P/E has simply shown itself not to be as powerful an indicator of a value share as several other similar measures available. In particular the P/E was left out of Fama and French’s three-factor model of stock returns (which I cover in Part II). This model has, in the academic world at least, become the most popular way of explaining the returns that can be expected from holding the shares of various types of company.
Putting right the faults in the P/E
It is these failings that I seek to put right in Parts III and IV. Comparing a company’s share price to its earnings is still a major basis of analysis for many value investors, but the pitfalls are numerous. This book explains how the P/E’s weaknesses can be overcome.
The short-term bias of the traditional P/E is easily overcome. The fact that the P/E is largely conditioned by the overall market P/E, the company’s size and the sector in which it operates, takes some rather more complicated adjustment but can still be done using freely available information. These adjustments put in your hands a tool that is much more finely tuned to revealing shares that are good value.
The other important thing the P/E does not tell you is whether a company is in trouble. As the history of the Naked P/E (explained later in the book) attests, building your portfolio on the basis of one statistic, however clever it may be, can lead to catastrophe during a market crash. Luckily, most of this risk can be avoided if you are prepared to step outside the P/E and use complementary statistics that steer you clear of the riskiest stocks. Combining the improved P/E with a measure of a company’s current financial strength provides a simple yet powerful filter. The final result is a set of stocks that are both underpriced in terms of their earnings power, and also solvent and likely to remain so.
PART I. The P/E Calculation
This Part starts with the story of how investors came to use the P/E. It has been in use for less than a century, but during that time it has become the most important investment ratio as far as practical investors are concerned.
A chapter looks at the more complicated component of the P/E: earnings. How the market decides share prices minute-by-minute is extremely complicated, but viewed from the level of the P/E calculation, the price used is quite straightforward. Earnings, however, are much more complex. There are many possible definitions of which parts of the profit and loss account are really ‘last year’s earnings’, and I cover them in detail here.
Having defined earnings, the share price and then the price–earnings ratio itself are covered in the following chapter. As with earnings, there are several different ways of calculating the P/E.
All the theory is brought to life in this Part’s final chapter, with a detailed example of how the P/E works in practice using Haynes Publishing’s annual report.
Chapter 1. History of the P/E
The P/E ratio is today the most commonly used valuation metric in the world.
Prof. Janette Rutterford, Open University, 2004
The P/E has a long history, but it has not always been the most popular way to value shares. Since the invention of stock markets up until less than 100 years ago, the dividend yield (DY) was the main figure every investor was interested in. The asset backing behind a company was also important. The P/E is, compared to these two, a relatively recent invention. Even the phrase ‘price–earnings ratio’ only became popular in the 1920s in the US. In the UK dividends were still what mattered up until the mid-1960s.
The world of investment to 1914: dividend yield is king
For centuries, at least since the time of the South Sea Bubble in 1720, dividend yield (DY) was the main ratio investors used to value their stocks. This is simply the ratio of the dividends paid out last year to the price of the stock. So if a stock costing £1 paid out 5p in dividends last year, the dividend yield of 5% compares favourably to the return available from a deposit account (in 2011), but of course involving the risk of capital loss should the share price fall.
In the US, bond issues during the 1800s and early 1900s outweighed stock issues three to one. The stock market consisted largely of railway stocks, with utilities and then industrials only becoming more important by 1900. In these circumstances it is hardly surprising that dividend yield was the favoured method of deciding whether a stock was cheap or expensive, because dividend yield could be directly compared to the yield on a bond.
In the UK, the stock market was more developed, even though the railway boom was equally active. One feature was how internationalised the stock market was: the majority of new issues was of foreign stocks and bonds. Another feature strange to a modern investor was the preponderance of debenture and preferred stocks. In a new issue from an established investment-grade company, these were often the only type of stock an investor could buy. The founders would keep control of the ordinary stock. It was speculative companies that issued ordinary shares.
Even in the UK though, DY was the main
