If history is any indication, economically disruptive rescission, like the 1991 variety, tend to produce two things in their wake. There is a predictable increase in companies reporting losses or profits based mostly on cost-cutting initially and, correspondingly, companies begin to be creative in their financial reporting to meeting earnings expectations. However, studies continue to show that the best way to mitigate against these dangers is investing in dividend stocks.

First, let’s state the obvious. In order to be a long-term dividend payer, a company has to be able to sustain cash flow necessary to pay dividends quarter over quarter. The implication being that dividend paying stocks do not necessarily have to be profitable all the time but rather they cannot be losing money over the long-term lest they jeopardize the dividend payment and investor confidence.

However, we appear to be investing in a world characterized by increasing loss reporting. A University of Chicago study on the relationship between dividends and earnings found that the fraction of losses reported by non-dividend payers in the U.S. grew from 28% in 1974-1979 to 52% by 2000-2005. Dividend payers followed the same general trend of increasing fraction of losses reported but in a far lower percentage than non-dividend payers with losses compromising of 3.5% of dividend payers in 1974-1979 to 11% in 2000-2005 (but this percentage of losses has been more or less consistent since 1985-1989). In other words, dividend paying stocks are less likely to report losses.

The losses reported by dividend payers also tend to more palatable than from non-dividend paying stocks. The authors found that over half of losses reported by dividend payers were due to one-time items versus approximately 25% of all losses derived from special items by non-dividend payers.  The authors conclude from this finding that losses tend to be better (for lack of a better term) than non-dividend payers since they are not attributed generally to operational inefficiencies.

The issue with this conclusion is that there is nothing barring companies from repeatedly reporting “one time” items quarter after quarter to hide operational inefficiencies. However, the large differentiation between loss reporting between dividend and non-dividend payers does tend to indicate that if this is occurring it is not in sufficient scale.

The flip side of the analysis is the author’s also found that dividend paying stocks have greater “earnings persistence” than non dividend earnings; the implication being that a dividend paying company showing a profit one quarter are more likely to show profits subsequently.  Yet, caution should be exercised for companies who rely on share buybacks. Their earnings persistence tends to be less than its counterparts who rely on steady dividend payments and are more likely to report losses (although still in fewer instances than non-dividend paying firms).

The study is interesting but should be read with two caveats. The sample size (1974-2006 studying NYSE, AMEX and NASDAQ) is relatively small and it excludes utilities and financial firms (which, given both are traditionally dividend payers, may distort the data somewhat).

If you are interested in finding the best dividend stocks, I would reiterate what other bloggers have written. The most reliable dividend data is found in the company’s financials themselves and understand what makes a good dividend stock. Start there and move outwards towards an industry comparison. While the study is interesting, please remember that Enron also paid dividends so don’t judge a book purely by its dividend paying cover.