The concept of buying dogs, is of course nothing new. It is a type of “reversion to the mean”. It is predicated on an assumption that all stocks are broadly similar, with similar growth patterns, and that nothing is genuinely new. Therefore, any stocks that have outperformed have more propensity to react negatively, and those that have underperformed have a propensity to outperform.
The first problem with this theory is that there are as we know genuine success stories in the market, those that have revolutionized an industry or properly executed their strategy. One example might be Apple. However the proponents of the theory may accept that exceptions exist to the general rule, or they may, as with all bell curve statistics, simply suggest that by definition some stocks will take longer to revert. Dell was a success story for a long period and then reverted negatively. Hewlett Packard was a laggard, and then reacted positively. Another factor to note is that there are likely to be a higher proportion of growth stories with smaller companies, because fewer companies have technology or process that is so innovative as to make a difference at the multi billion dollar level. Google and Apple again being different.
One should not confuse buying dogs or reversion to the mean with high yielding stocks. Yes, there are similarities, because many stocks that have seen their share prices slip will be higher yield, but as noted by others , that may show confidence by the company in their prospects which is a separate indicator. Equally, the share price may have slipped precisely because the dividend has been slashed, for instance British Petroleum. Finally, there are sectors with very disparate cash flow characteristics which dictate their dividend potential. Utilities tend to be higher yielding than technology, precisely because there is limited growth. Unless fundamentals support the declared dividend there is no sane reason why higher yielding stocks per se should outperform for fundamental reasons. However lets not start a war over that one. Value stocks also cross into the same field, and may be juxtaposed, possibly only on a technical level, with momentum buying.
Finally a word on the technical constituency of the Dow. Good stocks enter the larger Indices, and bad stocks drop out. Therefore statistics gleaned from the historic bases are in effect biased. Picking the poor stocks in one period you will tend to be picking those stocks that have moved upward rather than downward, which may have fallen so far as to have fallen out of the reckoning.