Dividends are wonderful yet odd things. While it's often great to hold stocks that pay you handsomely for keeping them around, many are like your old friend from the neighborhood who only shows up when he needs money. Of course, he promises to pay you back, with interest, once he "hits a big score." But you know it won't happen. You just hand over the cash because, well, that's what buddies do.

A lousy dividend payer is exactly like your old pal. It'll entice you with high yields but underneath it lacks the cash to pay you what's promised. And so its stock, and your portfolio, gets deep-sixed. Fortunately, it's pretty easy to find these greaseballs. Fairpoint Communications (NYSE:FRP) comes to mind. Its meaty 14.2% yield can't be funded through existing cash flow. No wonder the shares are down more than 33% since February, when the stock began trading.

Four ways to find the good ones
A little work will help you avoid this scenario. And I do mean a little. Mathew Emmert, chief analyst for Motley Fool Income Investor, has come up with four criteria for picking dividend winners that you should steal for your own investing. They are:

  1. Enough size and financial strength to create growth and pay a substantial dividend. Alternatively, a company offering a deep value.
  2. A market capitalization of $1 billion or more and debt less than 60% of capital.
  3. Return on equity in excess of 10% and return on assets approaching 2% for financials, 5% for property real estate investment trusts (REITs), and 7% for others.
  4. A dividend yield approaching 3% that's fully funded through free cash flow.

Now, ready to do this for yourself? Good. I've chosen -- gulp -- General Motors (NYSE:GM) as our guinea pig. Let's see if it passes Mathew's tests, starting at the bottom:

4. A dividend yield approaching 3% that's fully funded through free cash flow: At first blush, GM's 10.5% dividend yield looks juicier than a well-cooked filet mignon. And that's the way it should be. After all, the S&P 500 has returned roughly 7% during 2005, and many prominent researchers suggest that 6% is the best most investors can expect to see from the stock market long term -- at least until price-to-earnings (P/E)ratios return to historically average levels. In such an environment, GM's yield alone could be enough to deliver market-crushing returns.

But that assumes GM's payout isn't at risk. I believe it is. Cash flow, you see, is a major problem for the automaker. According to the most recent quarterly filing, GM's capital expenditures exceeded cash flow by close to $1.4 billion over the past nine months, and by more than $3.4 billion over the prior 52 weeks. And that was before dividend payments. Shareholders have received $1.15 billion in dividends over the past year.

No doubt GM's $53 billion cushion of cash and investments will help it sustain the payout, service debt, and maintain pension obligations for a while. Yet just a year ago the company's treasure chest was $4.6 billion richer. Eliminating the dividend might have cut the deficit by 25% or more. You don't really think management didn't notice, do you?

Grade: Fail.

3. Generous returns on equity and assets: Mathew says he likes to see management that deploys capital effectively. There are several ways to measure this, but he prefers to focus on return on equity (ROE) and return on assets (ROA) in his initial assessment. (ROE measures the effectiveness of reinvesting stockholder money to generate higher profits. ROA is similar, except that it measures the effectiveness of investing in capital assets such as property.)

It's simply no contest here. First, stockholders' equity declined by roughly $5 billion over the past year. And, second, GM lost $3.9 billion over the same period. That means both ROE and ROA slipped into negative territory, a terrible sign.

Grade: Fail.

2. A market cap of $1 billion or more and debt less than 60% of capital: GM is certainly tall enough for Mathew's screen, with a market capitalization of $10.6 billion. But decades of candy and ice cream have made the old automaker fat and lazy. Its $278.2 billion in debt is more than 25 times total capital. And I thought Ford (NYSE:F) looked bad with debt approaching 10 times capital. Sheesh.

Grade: Fail.

1. Compelling growth and/or a deep value: There isn't much good news here, either. Indeed, according to Yahoo! Finance, analysts believe that GM will earn $1.06 per stub next year. That's a forward P/E ratio of 17.7 as of this writing.

Don't think that's cheap. First, it's a premium to the Dow, which is trading for just 14.7 times projected 2006 profits. Second, the Value Line Investment Survey shows that in the 16 years between 1989 and 2004, GM's average annual P/E ratio reached double digits only five times. And, of those, two were during the bubble years of 2000 and 2001.

Grade: Fail.

Leave this stock on the lot
By far, GM boasts the worst stock of the automakers, even if its dividend yield wallops the 1.2% Toyota (NYSE:TM) offers. Or the 3.8% that DaimlerChrysler (NYSE:DCX) pays.

But my intent wasn't to lampoon GM. Instead, it was to show you how Mathew goes out about eliminating risk in seeking big investing rewards. It's about a lot more than just the yield. Which is why Income Investor, which you can try for free, is still beating the market as I write today. So take heed. And the next time a used-stock salesman pitches you the promise of a market-trouncing dividend, check under the hood before you drive off the lot. I guarantee both you and your portfolio will be glad you did.

Go on, take the money and run. Take a risk-free trial to Motley Fool Income Investor today and you'll get access to all of the picks and research that have helped Mathew beat the market since the service's inception. And there's never, ever an obligation to buy. (Though if you do, you'll getStocks 2006, our analysts' best picks for the year ahead, free. And the service is anchored by our money-back guarantee, no questions asked.)

This article was originally published on July 13, 2005. It has been updated.

Fool contributor Tim Beyers appreciates good yields like any common Fool. But he dodges the dangerous ones. Tim didn't own stock in any of the companies mentioned in this story at the time of publication. You can find out what is in his portfolio by checking his Fool profile. The Motley Fool has an ironclad disclosure policy.