With the new year upon us, it's time to dust off the Dogs of the Dow theory again. For those of you not familiar with the premise, it's pretty simple. Of the thirty stocks that make up the Dow Jones Industrial Average, the dogs are the ten stocks in the Dow that are considered to be the most undervalued. The theory is that those ten most undervalued stocks at the end of one year are poised to beat the Dow Average the following year. Does it work? Surprisingly, it does work more often than not. Let's take a look.

How do we determine "undervalued" for the purposes of this theory? In this particular case, we look at the dividend yield. While it's not necessarily the perfect measure of value for a stock, it is generally a good indication of how one stock compares to another. This is especially true for old blue-chip stocks like the ones you find in the Dow, where dividends are a little more common. To find your undervalued dogs, simply find the ten Dow stocks with the highest dividend yields.

But does this really work? There are always exceptions, but the idea does have some statistical merit......but only a little, recently. In five of the last nine years, the ten highest yielding stocks did indeed do better than the overall Dow average. This didn't mean that the ten dogs didn't lose money in some years, such as 2001 and 2002. But even in those years, you still would have been better off with the dogs. In fact, even though 2005 wasn't one of those years that the theory worked, we still like the idea of relatively low priced stocks. And statistically, the numbers support the theory. Since the early 70's, the average annual return on this theory would have been 17.1%. That's significantly better than the Dow's average annual gain of 10.9%. Take a look at last year's dogs and how they fared against the Dow (and note that we replaced SBC with AT&T, due to the meger). The theory didn't work at all, clearly. Do we abandon the idea then? No, but we do exercise caution. Over the long haul, the theory has superior results, so we're still interested in it. We just don't want to follow it blindly, in case we get another General Motors (GM) or Verizon (VZ) in the bunch.

Dogs of the Dow - 2005

So what about 2006? Take a look below for this year's "Dog" candidates. It's neither logical nor scientific, but it is food for thought. If the theory holds up, these stocks will have outperformed the Dow Average when we check in again at the end of 2006.

Dogs of the Dow - 2006

Price Headley is the founder and chief analyst of BigTrends.com.