Estimating the dividend growth rate

The estimation of future earnings and dividend growth rates is the hardest part in investing. A popular approach is to look at the past dividend growth rate and extrapolate it into the future; however, there is no fundamental reason why the dividend growth rate should continue unchanged. In fact, for any given company, the dividends cannot sustainable grow faster than company earnings and the earnings cannot forever grow faster than the gross national product (GNP). Since the historical average GDB growth rate is around 5-7%, no dividend growth rate above this rate is sustainable in the long term.

For shorter time frames, successful companies can grow earnings and dividends at a fast rate. For example, companies such as JNJ, WMT have raised their dividends by ~10% annually for several decades. Correctly estimating the future growth rates is critical in valuing these dividend growth companies. Here we will cover methods of estimating the dividend growth rates for the near future and long term.

Sustainable growth rate from ROE

The short term sustainable growth rate can be estimated from the company earnings (E) and return on equity (ROE). A portion of the earnings is paid as dividends to the shareholders. The ratio of dividends to earnings (Div/E) is called the payout ratio. The remaining earnings not paid as dividends grow the shareholder equity can be invested to grow the business. Assuming that the ROE remains constant, the future earnings growth rate is therefore

Using Equation (1), we can estimate the sustainable dividend growth rate that can be sustained from earnings growth. As an example, Wal-Mart has ROE=23.18% and payout ratio = 28%. The assuming ROE is not changed, the future earnings growth rate is gE = (1-0.28)*23.18%=16.7%. This is also the sustainable dividend growth rate as the dividends cannot grow faster than earnings without a change in the payout ratio. As a second example, a high payout company (T, ROE = 18.29%, payout ratio = 50%) has a lower growth rate of gE = (1-0.5)*18.29% = 9.1%

The problem with this analysis is that there is no guarantee that ROE will remain constant over time: As companies became larger, it becomes increasingly more difficult for them to find good investment opportunities. For example, Wal-Mart grew rapidly by building stores all over US. Now that most rural towns have a Wal-Mart store, where is the earnings growth going to come from? 

Extrapolating historical growth rate

The easiest and therefore most common method for future dividend growth estimation is to extrapolate the past growth rates. With a short time horizon, this can give fairly accurate predictions. For example, for Wal-Mart has a long history of earnings and dividend increases. Trailing one year dividend growth rate is 11.0% and the earnings growth rate is 12.8%, respectively. The competitive landscape is unlikely to change over the next few years so the dividend growth rate is likely continue unchanged for some time.

Over long periods, say over 5 years, the predictive power of past performance looses its shine. For example, the big box business model has been under attack by online merchants such as Amazon. Wal-Mart has been able hold its own but will this hold in the future?

Unsustainable dividend growth from increasing payout ratio

Transition from dividend growth to slow growth is one of the hardest questions in dividend models. Eventually, the high growth rates will slow down but predicting when this happens is difficult. Getting the estimate wrong even by a few years can significantly change the NPV from 2-stage DDM.

For many “dividend growth companies”, the dividend growth rate gD is larger than the earnings growth rate gE. In this case, the number of years that the dividend growth can outpace the earning growth can be estimated the final payout ratio reaches value that is typical for the industry. The dividend after N years is

The earnings after N years is

The current payout ratio is

payout ratio after N years is

Combining Equations (2) to (5) gives

for the number of years that the dividend growth can be sustained by increasing the payout ratio.

As an example, we will analyze the dividend growth of Medronics (MDT). As the spreadsheet below shows, MDT has been growing dividends at over 19% annually over the past decade. In fact the dividend growth rate has been larger than the earnings growth rate of 15% and larger than the sustainable growth rate from ROE. Extrapolating the earnings and dividend growth rates into the future, MDT will reach 40% payout ratio in 5 years. Larger payout may not be sustainable for a medical company that needs heavy R&D investments to maintain competitiveness.

Assuming that the high growth rate continues for 5 years after which the terminal growth rate is 3% and assuming discount rate of r=10%, the DDM2 model gives target price of $34.42 for MDT. This is close to the current market value ($34.10 as of 11/9/11). The discount rate of 10% is justified by the higher risk of investing in dividend growth company – after all, we could obtain a relatively safe return of r=8% by investing in large telecom such as T. The risks in the investment is that MDT earnings growth slows much sooner than next five years. The upside is that the earnings and dividends will grow for the next decade or more.  

Dividend growth rate estimation v2 ‎(MDT)‎


An editable spreadsheet is located at https://docs.google.com/spreadsheet/pub?hl=en_US&hl=en_US&key=0AlyBMfBlh5aodFFPUV9ZUFVzQmRUb3JWSU5NWFRNTkE&output=html

Long term growth rate

Even for a stellar company, the growth rate cannot forever be higher than the GDP growth rate. The historical GDP growth rate for US is 5.3% from 1790 to 2010 and 6.6% from 1945 to 2010 (see http://www.measuringworth.com/growth/ ). Even these numbers may be optimistic as even solid dividend companies can “go over the hill” (see for example Kodak or GM). A more conservative argument is to use 3% from inflation alone for companies such as Coca-Cola that have solid and historically proven moat and 0% or even negative growth rate for more speculative investments.

Summary

Estimating future dividend growth rates is art, not science. High growth rate predictions are especially problematic as the high growth rates cannot be sustained forever but predicting when the growth stalls is difficult. Best that can be done is to be conservative and never assume that the high growth rate goes on more than ten years.

Comments