Of all the fundamental statistics available for comparing stocks, the Price/Earnings ratio is the single most widely used. Although all of the other ratios involving a stock's price -- Price/Sales, PE/Growth, Price/Book, Price/Cash Flow -- have value, the P/E ratio is the most well known, most often quoted when only one statistic is given, and serves as an instant reading on the value of a stock.

At first glance, the Price/Earnings ratio seems quite simple. It is calculated, using per share data, as:

CURRENT PRICE / ANNUAL EARNINGS = P/E RATIO

Unfortunately, the P/E statistic is frequently used without clearly defining exactly what is being discussed. Although the math is easy, and the price is usually well-defined, there are options for determining the earnings number, and this is where the most confusion about P/Es occurs.

When you see that a stock has "a P/E of 22.0" the first thing you must do is determine what kind of P/E this is. This usually means determining what kind of earnings number is being used. There are several kinds of P/Es:

• Trailing P/E:
  The earnings are the most recently reported quarter, plus the previous three reported quarters. The abbreviation TTM is often used to indicate Trailing Twelve Months. However, it really means trailing four quarters of reported numbers.
• Forward P/E:
  The earnings number is the total of estimates for the current unreported quarter and estimates for the following three quarters.
• Fiscal Year P/E:
  Often you will see P/E as calculated on a particular fiscal year for the company. This may be in a chart or table with P/E's for several future fiscal years shown. When this is done, the current fiscal year may have some actual quarters and some quarters with projections. The future fiscal years are all estimated, of course. For example, in January 2001, FY02 P/E would refer to today's price divided by the total earnings estimated for fiscal year 2002. FY01 P/E might include some actual quarters and some estimated. If the current fiscal year is not complete, the numbers are usually listed as estimates, even if there is only one quarter unreported.

There are other variations as well. For example, Value Line uses a blended P/E ratio, which they define as the most recent two actual quarters added to the estimates for the upcoming two quarters. Occasionally, you will see the current quarter's estimates combined with the three trailing quarters.

In all of these variations, however, the price used is today's price, or at least, the price on or around the time of analysis. Rarely is a future (projected) price used and a historical price is never used. When you see a reference such as "Microsoft never had a P/E higher than 27 in the '80s" the price being referred to is always the price at that time. Generally, in historical reflections, the earnings will be trailing earnings. Occasionally you will see a comparison of today's Forward P/Es with Forward P/Es from the past, but the historical Forward P/E calculations will likely use the actual forward earnings, not the estimates from those times.

It is obvious, of course, that estimated earnings will vary by source. However, even actual earnings may vary by source due to calculation methods. While the total earnings of the company are always known, in deriving the per-share earnings, there are two factors to consider: 1) basic vs. diluted shares; and, 2) including or excluding extraordinary charges.

Basic earnings are calculated by taking the total reported earnings of the company and dividing by the number of shares actually issued currently. Fully diluted earnings are calculated by taking the total reported earnings and dividing by the total of the following: 1) all shares issued; 2) plus all shares subject to options; 2) plus all shares subject to warrants; 3) plus all shares subject to convertible bonds. In short, "fully diluted" calculations postulate what would happen if everyone who has a right to a share, exercised that right today. In many cases, for example options, the holder cannot actually exercise that right today. However, since many companies continually issue new options and these are converted over time, fully diluted earnings calculations help account for that future dilution.

Basic earnings per share numbers will be higher than fully diluted earnings per-share numbers when earnings are positive. But when a company is losing money, the diluted EPS will be higher as the higher share-count will produce a smaller per share loss. P/Es using basic earnings numbers will therefore be lower than P/Es using fully diluted earnings numbers when a company is profitable (when a company is losing money, a P/E cannot be calculated).

The other calculation affecting earnings is extraordinary charges. Extraordinary charges are expenses that companies classify as non-recurring, and are treated as a separate line item on financial reports. These may be costs associated with a layoff or merger, or other one-time event. Generally, most analysts exclude extraordinary charges from earnings calculations. (In fact, Wall Street routinely ignores extraordinary charges altogether even though the expense is real.) If you are making your own P/E calculations from annual reports or from screening databases, you should probably make your calculations using the earnings line which excludes extraordinary charges.

Quite simply, the P/E ratio gives you a quick comparison for determining whether a stock is "cheap" or "expensive." Stocks with low P/Es are "cheaper" than stocks with high P/Es. If the stocks are in the same industry, have the same annual earnings, but have different P/Es, the stock with the lower P/E is cheaper. That doesn't mean it is a better buy, as it may have a lower projected growth rate, or some other issue. On the other hand, it might just be cheaper. The P/E ratio is simply a good place to start when making comparisons between stocks.

With all the variations in P/Es, you must make sure you are comparing apples to apples. In any published report, Briefing.com will always identify when a variation in the P/E is being discussed. Briefing.com's Company Reports use Trailing P/Es. All companies now use diluted share-counts when calculated P/Es.

During earnings reporting season, you also have to make sure that both stocks you are looking at have the same reporting status. With Trailing P/Es, as soon as a stock reports, the new quarter is substituted for the oldest quarter's earnings. If the second stock has not yet reported, its TTM P/E will not reflect the same time period and you should take this into account.

The biggest problem, however, is comparing P/E numbers for stocks when using two different sources. When comparing an analysis from one source with an analysis from another source you need to check the P/E calculation methods before comparing P/Es. Comparing one stock's Trailing P/E with another's Forward P/E will be misleading.

In researching or analyzing stocks, Briefing frequently uses forward P/E's, or fiscal year P/Es, using estimated earnings for the coming year. These are always clearly identified as such. In no case will Briefing mix different P/E calculations for different stocks in an analysis without clearly identifying which P/Es are being used.

Which method is best? Unfortunately, no single method is clearly better than another. Each is useful. Trailing P/Es are solid numbers. Forward P/Es can be misleading if estimates turn out to be wrong. However, Forward P/E's are helpful, especially when paired with the PE/Growth ratio, as this gives a picture of how expensive the stock is relative to its expected future growth. And the future is what the stock market is all about.