Edited: Subdued Dividend Growth

Dividend growth can rescue retirement portfolios. In this investigation, I looked at history to identify failure mechanisms. I looked for conditions that have subdued real dividend growth.

I found that today it is not a matter of business fundamentals. It is a matter of discretion: how corporations treat their shareholders.

Background

Dividends have not always grown in the past (after adjusting for inflation). Careful examination of the modern era, starting from 1950, shows that dividends are much more dependable today than they were earlier. Much of this is because of lower dividend payout ratios.

In Dividend Growth Baselines, I established that a 1% real dividend growth rate is sufficient to maintain a continual withdrawal rate of 3.5% to 3.6%. In Dividend Growth Rates, I found that the S&P500 dividend amount does not always grow by 1% (or more), even in the modern era (starting in 1950), with its lower payout ratios.

Data Analysis: 50% Payout

Actual Reductions

Using a criterion that the single-year payout ratio must be below 50%, there were seven sequences with actual reductions over 5-years: 1973, 1979, 1980, 1981, 1998, 1999 and 2000. There were three years with actual reductions over 10-years: 1973, 1974 and 1975.

Using an alternate criterion (that the average of five years of payout ratios or the ratio of the average of five years of dividends divided by five years of earnings must be below 50%), there were six years with actual reductions over 5-years: 1979, 1980, 1981, 1998, 1999 and 2000. There was only one year with an actual reduction over 10-years: 1975.

Returns Less Than 1%

Using a criterion that the single-year payout ratio must be below 50%, there were twelve sequences with 5-year returns below 1%: 1973, 1974, 1975, 1978, 1979, 1980, 1981, 1982, 1997, 1998, 1999 and 2000. There were nine sequences with 10-year returns below 1%: 1973, 1974, 1975, 1976, 1977, 1978, 1979, 1980 and 1995.

Using an alternate criterion (that the average of five years of payout ratios or the ratio of the average of five years of dividends divided by five years of earnings must be below 50%), there were eleven sequences with 5-year returns below 1%: 1975, 1978, 1979, 1980, 1981, 1982, 1991, 1997, 1998, 1999 and 2000. There were six sequences with 10-year returns below 1%: 1975, 1976, 1977, 1978, 1979 and 1980.

Data Analysis: 40% Threshold

Using a criterion that the single-year payout ratio must be below 40%, there were four sequences with actual reductions over 5-years: 1980, 1998, 1999 and 2000. There were no sequences with actual reductions over 10-years.

Using an alternate criterion (that the average of five years of payout ratios or the ratio of the average of five years of dividends divided by five years of earnings must be below 40%), there were two years with actual reductions over 5-years: 1998 and 2000. There were no sequences with actual reductions over 10-years.

Using a criterion that the single-year payout ratio must be below 40%, there were five sequences with returns less than 1% over 5-years: 1980, 1997, 1998, 1999 and 2000. The 1980 sequence returned less than 1% over 10-years.

Using an alternate criterion (that the average of five years of payout ratios or the ratio of the average of five years of dividends divided by five years of earnings must be below 40%), there were three sequences with returns less than 1% over 5-years: 1997, 1998 and 2000. There were no sequences with less than a 1% return over 10-years.

Starting Today

The last set of payout ratios in my database were in 2004. They were 35.3% (single-year), 34.1% (five year average of payout ratios) and 34.2% (average of five years of dividends divided by the average of five years of earnings).

Any failure to maintain at least a 1% dividend growth rate would be by discretion, not necessity. We have already seen such a choice. The 5-year growth rates after 1998 and 2000 came about because of sharp earnings shortfalls (about 50%) seen in the (January) 2002 and (January) 2003 data. Yet, the single-year payout ratio never exceeded 57% in 2002 and 2003.

Conclusion

A recession similar to what we have seen in the 2002 and 2003 (January) earnings data could stall (real) dividend growth. It has happened before, recently. But the reason would not be economic. It would not be because of a lack of earnings power. It would be the intentional choice of corporate boards.

Dividend investors need to be careful about how companies treat their shareholders.

Have fun.

John Walter Russell
December 28, 2006