In the midst of a bull market, a lot of people begin thinking that price isn't very important -- earnings growth is the only thing that matters to a stock price. I strongly disagree. While concurring that earnings growth is a powerful driver of company value, the price you pay for that growth and the price received when a stock is disposed of are also important components of ultimate shareholder returns. Today I'm going to take a look at how prices, measured by the price-to-earnings ratio (P/E), interact with earnings to create returns.

To understand what I'm talking about, it's helpful to break the price of a stock into two of its key components: earnings per share (EPS) and the P/E ratio. The stock price of any company making a profit can be broken down into the these components using the following formula:

Share price = EPS * P/E ratio

As you can see, share price doesn't directly corollate with EPS. The P/E ratio -- the premium investors are willing to pay for those earnings -- plays an integral role in determining how much a stock will cost.

Some Examples
Let's look at a couple of examples showing how the interplay of P/E ratios and EPS has worked at three companies over the past five years: Sun Microsystems (Nasdaq: SUNW), Intel (Nasdaq: INTC), and CBRL Group (Nasdaq: CBRL). The tables below show trailing 12-month earnings, P/E ratios, and stock prices for the three companies at the end of 1994 and the end of 1999.

```      Trailing    Price/    Stock
Earnings   Earnings   Price
10/94      12/94     12/94
SUNW   \$0.14      15.6x    \$ 2.19
INTC   \$0.72      11.1x    \$ 7.98
CBRL   \$0.98      18.9x    \$18.50

Trailing    Price/    Stock
Earnings   Earnings   Price
10/99      12/99     12/99
SUNW   \$0.75     103.3x    \$77.44
INTC   \$2.22      37.1x    \$82.31
CBRL   \$0.99       9.8x    \$ 9.70

Percentage Change 1994-1999

Trailing    Price/   Stock
Earnings   Earnings  Price
SUNW    428%       562%    3,390%
INTC    210%       234%      931%
CBRL      1%       -48%      -48%
```
In each case, changes in P/E ratios had a more substantial impact on stock price than reported earnings growth. If investors still paid only 15.6x Sun's earnings, as they did at the end of 1994, its stock price would have "only" increased 428%, about an eighth of the reported 3,390% rise. Similarly, if investors hadn't been willing to pay more for Intel's earnings, its stock price would have moved up 210%, roughly one-fourth the actual 931% increase. On the other hand, CBRL saw its stock price halved between 1994 and 1999, even though earnings were basically flat during the period. That's representative of the pain endured by investors when P/E multiples contract.

The P/E ratios of Sun, Intel, and CBRL have changed over the past few years for good reasons. Investors have become increasingly confident that Sun and Intel are going to continue posting strong earnings growth in the future. In 1994, many people considered these stocks to be cyclical, subject to the vagaries of the technology cycle, where periods of strong earnings would be followed by slowdowns. Due to the continued technology explosion, however, perception has shifted and these industry leaders are considered poised for unabated growth. For CBRL, dismal results over the past two years have caused investors to doubt that the company's Cracker Barrel restaurant unit will show strong, steady growth as previously expected.

Implications
As investors, we need to be aware of the important role "price" plays in moving stocks. For a repeat of the stock market's performance over the past five years, we'll not only have to see earnings growth continue at its current record pace, but investors will have to take the P/E ratios they're willing to pay even higher into record territory.

For Intel to show the same stock performance between 1999 and 2004 that it did over the past five years (59% per year), a couple of things would have to happen. First, the company would need to continue posting earnings gains of 25% a year. Second, investors would have to be willing to pay 124x those earnings by the end of the year 2004. If the earnings multiple doesn't expand, Intel's stock price would increase commensurate with earnings growth.

A reduction in the P/E investors are willing to pay would be less pleasant. If five years from now, the multiple decreases to 20x from the 37.1x at December 1999, the stock price would only rise 10% a year, even though earnings grew at a 25% annual pace. A slowdown in earnings growth below 25% (traditionally optimistic analysts project only 20% long-term growth) would reduce these returns further.

The returns of Intel and Sun (and most other "hot" companies over the past few years) demonstrate the benefits of investing in companies with growing earnings and expanding multiples. We should always seek situations where it is likely that both of these inputs will be expanding -- it's one of the most powerful ways to make money in the stock market.

On the other hand, CBRL demonstrated the pitfalls of putting money into a company where P/Es contract. Even if earnings grow, stock prices are subject to staying flat or falling substantially. At a time when overall market multiples are at near-record levels, investors should be diligent in their evaluation to ensure that the interplay of earnings growth and P/E ratios will result in positive returns.