Here is report of work in progress (WiP) of Dividend Heritage Project
12 July 2012
More than 3000 US companies paid flat (constant) or growth dividends at least 4 consecutive years. Some distributions are shown below.
Historical comparison: Oldest Dividend Champions vs. Oldest dividend payers:
As anybody can see dividends runs are more stable when a firm sustain more than ~ 25 years of flat or growth dividends.
Diebold Inc. has the longest history (about 60 years) of growth dividends (according David Fish's CCC list) while some companies paid dividends for more than century.
About 1600 firms were qualified for David Fish's CCC list at some moments of time. Now this list contains about 450 companies. IMO it confirm that "buy & forget" approach doesn't work for DGI.
20 July 2012
About 500 companies from 1600 mentioned stopped to pay dividends - about 55% of them were acquired by private or public companies and about 45% suspended dividends. Some of public companies that bought these 500 mentioned companies are current dividend champions or well-known firms like Berkshire Hathaway.
11 November 2012
I recently read 2 interesting papers
"A New Historical Database for the NYSE 1815 to 1925: Performance and Predictability" by William N. Goetzmann, Roger G. Ibbotson, and Liang Peng as well as "Samuelson's Dictum and the Stock Market," by Jung Jeeman and Robert Shiller published in Economic Inquiry, 2005, Vol. 43, No. 5, pp. 221-228.
I'd like to point 3 facts from these papers:
1) During long run dividend change rate (DCR) was mostly positive for US stocks /see fig. 1 from the paper below/
and DCR negatively correlate with yield for yield below about 15%.
2) Only 49 firms have continuous information in Center for Research on Security Prices database during the period from 1926 to 2001. Also there are 125 ﬁrms that have existed during the 1949-2001 period without any missing information on stock prices and dividends. Also Poterba and Summers (1988) found 82 survival ﬁrms during the 1926-85 period in the CRSP tape. The small numbers here apparently reﬂect the continuing disappearance of ﬁrms through time due to liquidation, acquisition, reorganization. Hence "Buy & Hold (and Forget)" approach doesn't work for individual stocks (although in may work for any index fund of stocks in country with quite stable economy /e.g., compare US with Gemany in XX century/).
William N. Goetzmann, Roger G. Ibbotson, and Liang Peng accompanied their paper with database I played with. I tried to find examples of consistent dividend growth policy in firms that were traded in NYSE from 1815 to 1925. From 515 firms listed in the database only 16 firms had 5-6 years of dividend streaks (see fig. below)
Assuming that information is incomplete probably slightly more firms /see 2) above/ had dividend growth streaks but it seems that such events were very rare and streaks were quite short (probably less than 15 years).
The facts 2) and 3) probably show that I should limit the timeline of Dividend Heritage Project to shorter period than I initially wanted to investigate.
26 Feb. 2013
Distribution of quarterly changes in dividends (excluding no change) for all publicly traded US firms paying regular quarterly dividends between 1980 - 2005 is shown in Fig. 1(2/26/13) below. Please note that dividend of ALL firms i.e., totally 113,955 dividend events are considered.
Quite many cuts are 50% reduction of dividends (100% cut = dividend omission). Most of dividend increases are equal to 1,2,3,5 and 10 cents and they are within 5%-20%.