In the last lesson, we learned about how dividends can establish a firm intrinsic value for a stock and act as a check on management's capital-allocation practices. In this lesson, we'll focus in more detail on how to indentify high-quality stocks with good total return prospects.

Breaking total return into current yield and expected dividend growth, we should also sort the growth potential into two buckets--growth in the company's core business (assuming it's profitable growth, that is, or all bets are off) and the growth funded by any remaining free cash flows. We'll call this three-part process the Dividend Drill.

Consider the Current Dividend

If we can establish that a stock's current dividend is sustainable long term, we can take the stock's current yield and, voila, one chunk of our total return is accounted for. Taking a dividend for granted means establishing long-term sustainability. Nothing lasts forever--just ask the shareholders of once-venerable Goodyear Tire GT--although a few stocks, such as General Electric GE, have dividend records that come awfully close to immortality.

What establishes a secure dividend? Look for manageable debt levels. Remember, bondholders and banks are ahead of stockholders in the pay line. Next look for a reasonable payout ratio, or dividends as a percentage of profits. A payout ratio less than 80% is a good rule of thumb. Finally, look for steady cash flows. Also demand an economic moat: No-moat companies tend to be cyclical (think autos and chemicals) and lack the pricing power to maintain earnings during the inevitable industry downturns.

Coca-Cola KO is a good example. In mid-2005, the shares were changing hands at about $45 while paying a $1.10 annual dividend. At that time, the payout ratio was reasonable (52% over the previous 12 months), cash actually exceeded debt (no debt worries), and operating cash flows were consistent. Best of all, the firm's moat is very wide--Coke is arguably the most valuable brand name on earth, quite the achievement for what is, after all, caramel-colored sugar water.

Coke's yield at that point was 2.4% ($1.10 / $45), giving us the first building block of prospective total return. And based on current earnings power of roughly $2.00 per share, we'll have $0.90 in retained earnings to fund dividend growth, which, as noted earlier, takes two forms.

Assess the Company's Core Growth Potential

One key to this analysis is understanding how much investment is required to fund this growth. Few areas of the market are bursting at the seams, but most companies and industries have at least some growth potential over time as the U.S. economy expands (figure 3%-4% per year plus inflation) and emerging markets open up. Inflation can be a tailwind, too--though taking price increases for granted with manufacturing-oriented firms is not necessarily a good idea. Fortunately for most mature businesses, supporting this baseline level of growth is relatively inexpensive, and therefore high return.

Another and often simpler way to think about the cost of growth is to look at the company's free cash flow as a percent of net income. Since free cash flow includes the cost of capital investments that support growth initiatives, the difference between earnings and free cash gives us a sense of the cost of growth.

For example, let's say free cash flow consistently totals about 60% of net income, while sales and profit growth run about 6%. This suggests that only 40% of earnings will support this growth, leaving the other 60% of net income available for dividends, debt reduction, share buybacks, and other noncore investments.

This core growth gives us the second chunk of our total return equation. For Coca-Cola, let's assume 5.2% growth in operating income over the next five years, and that Coke's growth will fall significantly below that figure thereafter. Assuming that management maintains the current payout ratio, the firm's total dividend payout should rise at a similar clip. So we bolt on this 5.2% growth to our prospective total return, bringing our expectations (including the 2.4% yield noted above) to 7.6%.

But we've got one more task before moving on to the third and final step--how much will achieving this 5.2% growth cost? One of the simplest angles is to take the growth we expect (5.2%) and divide that by a representative return on equity (a nifty 30.8% for Coke in the past five years). The resulting ratio--call it "R-cubed" for "required retention ratio"--is the proportion of earnings used to fund core growth. For Coke, the R3 is 17% of income, or $0.34 per share.

Aftertax return on invested capital is also worth a look. ROIC is actually the purest way of analyzing the incremental cost of growth; in our formula ROIC replaces return on equity in the calculation of R3. However, ROIC is more complex to use, and it leaves out the company's capital structure (mix of additional borrowings and retained earnings) that is reflected in ROE. If the capital structure is stable and returns on equity are consistent--Coke checks out here on both counts--ROE is a good metric to use.

We'll stick with ROE R3, and estimate 5.2% annual growth will cost Coke $0.34 per share. Over time the absolute number will grow, but the proportion (17%) will remain the same as long as its two factors--growth and return on equity--stay the same.

Two thirds of the way through our analysis, we're up to a 7.6% return, and we still have $0.56 per share to spare ($2.00 in earnings less $1.10 for the dividend and $0.34 to fund core growth). So what's the final $0.56 per share worth?

Evaluate the "Excess" Earnings

After paying dividends and funding core growth, a company may have cash left over. It could opt to pay down debt, which would reduce interest expense and thus increase earnings. It might make an acquisition or some other investment, though the returns here could be spotty. Finally, it might opt to buy back stock.

Whatever the company decides to do with these excess funds, we put the result into the growth bucket of our prospective total return. In other words, we assume that any cash not used for a dividend is employed to create earnings and dividend growth. To get a proxy for the added growth potential of remaining earnings, we'll make an additional assumption that the path of least resistance is a share buyback.

This assumption is meant to err on the side of conservatism. The earnings yield (the inverse of P/E) on most stocks is generally much less than a company's return on equity, so we're not projecting much bang for this last slice of our buck. And acquisitions--returns of cash to someone else's shareholders--tend not to be priced for returns equal to existing investments.

Share buybacks boost earnings growth--EPS grows not only when the numerator (profit) expands, but also when the denominator (shares outstanding) shrinks. Dividing the excess earnings into the stock price gives us an "excess earnings yield," the third component of our total return calculation. So if Coke uses the last $0.56 of per-share earnings to repurchase stock, it will be able to retire 1.2% of its shares in the first year ($0.56 divided by a $45 share price). That, in turn, gives next year's earnings per share a 1.2% tailwind--even if earnings are flat, fewer shares outstanding mean higher earnings per share.

So What's It Worth?

Totaling Coke's yield (2.4%), profit growth (5.2%), and excess earnings yield (1.2%) produces an expected total return of 8.8%. It's important to note that this total return projection is contingent on the current stock price--we can expect an 8.8% annual return from Coke only if we acquire the shares at $45. If we pay less, our total return will be higher, and vice versa.

For example, let's say the market hits the proverbial banana peel, and Coke is offered at $35. Meanwhile our expectations (current earnings, dividend rate, future growth) haven't changed. Our core growth projection (5.2%) remains, but our two other factors are contingent on the stock price: At $35 the stock will yield 3.1% and our excess earnings quotient will rise to 1.6%. Our expected total return is now 9.9%, more than a full point higher. Conversely, if we wind up paying $55, our total return prospects are substantially reduced. Coke's yield will fall to 2%, the excess earnings quotient to 1.1%, and our expected return to 8.3%.

This analysis essentially calculates fair value in reverse--instead of using a required rate of return to yield a fair price for the stock, we use the stock price to calculate the shares' total return. Coke's fair value is the price at which its total return is equal to the return we would require for any stock of similar risk characteristics. Morningstar's fair value estimate in mid-2005 for Coke was $54, which was calculated using an 8.5% cost of equity--a return virtually identical to our total return projection if we use $54 as the stock's price.

What's the "right" required rate of return? Unfortunately there's more art than science to this, but we have two observations. First, over a very long period of time (200 years), the market has managed to return something around 10%. Lower-risk stocks would offer less, while higher-risk situations should require more. But most established, dividend-paying companies would fall in a range between 8% and 12%. Whatever you determine a "fair" return to be, demand more. This way you have a margin of safety between your assumptions and subsequent realities.

The Bottom Line

This analysis is not suited to every stock or situation. For one thing, even with the surge in the popularity of dividends in recent years, less than half of U.S. stocks pay a dividend. It's also not particularly well suited to deeply cyclical firms, whose earnings power and even dividend rates will vary widely from year to year. It's also not suited for emerging-growth stories. But for the ranks of relatively consistent, mature, moat-protected stocks--of which there are hundreds, if not thousands to pick from--we can use the dividend as a critical selection tool.