From
Common Stocks and Uncommon Profits:
The fact that a stock has or has not risen in the last several years is of no significance whatsoever in determining whether it should be bought now. What does matter is whether enough improvement has taken place or is likely to take place in the future to justify importantly higher prices than those now prevailing.
Similarly, many investors will give heavy weight to the per-share earnings of the past five years in trying to decide whether a stock should be bought. To look at the per-share earnings by themselves and give the earnings of four or five years ago any significance is like trying to get useful work from an engine which is unconnected to any device to which that engine's power is supposed to be applied. Just knowing, by itself, that four or five years ago a company's per-share earnings were either four times or a quarter of this year's earnings has almost no significance in indicating whether a particular stock should be bought or sold. Again, what counts is knowledge of background conditions. An understanding of what probably will happen over the next several years is of overriding importance.
The investor is constantly being fed a diet of reports and so-called analyses largely centered around these price figures for the past five years. He should keep in mind that it is the next five years' earnings, not those of the past five years, that now matter to him. One reason he is fed such a diet of back statistics is that if this type of material is put in a report it is not hard to be sure it is correct. If more important matters are gone into, subsequent events may make the report look quite silly. Therefore, there is a strong temptation to fill up as much space as possible with indisputable facts, whether or not the facts are significant. However, many people in the financial community place emphasis on this type of prior years' statistics for a different set of reasons. They seem to be unable to grasp how great can be the change in just a few years' time in the real value of certain types of modern corporations. Therefore they emphasize these past earnings records in a sincere belief that detailed accounting descriptions of what happened last year will give a true picture of what will happen next year. This may be true for certain classes of regulated companies such as public utilities. For the type of enterprise which I believe should interest an investor desiring the best results for his money, it can be completely false.
