Dividends – a better retirement model

I’ve talked a lot about how much it takes to retire and safe withdrawal rates and ways to increase safe withdrawal rates. See the series of posts here. All of these discussions and studies I reference are based on a certain retirement model. Lets call it the portfolio value retirement model. One critical assumption in that model is that it assumes re-investment of all stock dividends and income from bonds and relies on selling off assets yearly to fund retirement spending needs. This assumption makes the longevity of a retirement portfolio critically dependent on big down years in the value of the portfolio. I think there is a better and simpler retirement model – the dividend retirement model.

A dividend retirement model is one where an investor lives off the dividend income received from their retirement portfolio. No asset sales are required to fund retirement. So, if you have a portfolio that yields 4%, you live off that 4%. Simple. In such a model, the market prices of the stocks, the value of the portfolio, are much less important than in the other model. However, other factors are much more critical. In the dividend retirement model, the stability/consistency of the dividend stream and its growth relative to inflation are the most critical parameters. Lets look at the history of both these factors to see how feasible such a retirement model is. We’ll use the S&P500 index, the same index used in the traditional retirement model. Using the dividend history of the index going back to 1900, the inflation adjusted (real) growth in dividends has been 1.2%. Thus, dividend growth from the index has been enough keep up with inflation. The second test is more difficult – the stability of the dividends. The table below shows every year since 1900 where there was a reduction in the S&P500 dividends and the extent of the reduction.

The first column shows the dividend reduction in a given year and the second column shows the cumulative reduction in the cases there were several years in a row of cuts. The first column highlights those years with sequential reductions and the second column highlights cumulative reductions above 20%. Not very stable and some very large reduction in dividends that would make living off the S&P500 dividend stream near impossible. And this is the reason I think most academics, financial planners, don’t even discuss the dividend retirement model. Based on the S&P500, it is not a feasible model. Since most academics, financial planners, are efficient market proponents anyway, the conclusion is that an investor cannot do better than this. Well, BS.

In order to implement the dividend retirement model, an investor cannot rely on indexes due to the instability of the dividend stream. The investor needs to invest in a portfolio of dividend paying stocks with consistent and rising dividends. This is not as difficult as it sounds. There are many stocks that have long histories of steady and rising dividends. A good reference on these dividend stocks lists can be found here. For example, the S&P dividend aristocrats are a group of 42 companies in the S&P500 that have raised dividends for at least 25 years in a row. And if you remove the large cap requirement, the US dividend champions list, also with 25 consecutive years of payments and growth, has over 200 members. In addition, there even are larger lists of 10 year payers, up and coming payers etc… And many of these lists do not even include stocks such as REITs, MLPs, and international stocks which can enhance yields further. I’ll post more on these lists in the future and some more thoughts on portfolio construction but the point is there is a large pool of consistent quality dividend payers to fund a stable retirement.

Outside of choosing consistent dividend payers with long track records, at least 10 years I would say, a dividend retiree can add an additional layer of safety by having a dividend income buffer. For example, I personally use an income buffer of 20%. If my retirement income needs are $40K then I target a dividend income from my portfolio of $50K. This way I can target a worst case dividend cut of 20% across my portfolio, a very dire assumption if one is choosing the right dividend stocks. Such a strategy allows me to sleep much easier at night knowing that one unforeseen dividend cut will not impact my income requirement for the year.

Another upshot of the dividend retirement model is the lesser importance of the value of the portfolio as determined by the price of the stocks in the portfolio or the price of the market indexes for that matter. As long as the dividend is stable and growing more than inflation, the value of the portfolio, prices of stocks, is of much less importance to retiree. We know that in the long run dividends and dividend growth account for most of stock’s returns but there can be long periods of mispricings and under performance. This is yet another layer of comfort with such a portfolio. I like to use JNJ as an example here. Over the last 10 years JNJ’s stock price has basically been flat but during that same time it’s dividend growth rate has been 13% per year. If I’m using JNJ’s dividend as a piece of my retirement income stream I feel pretty darn good despite the flat stock price.

In summary, I think its quite realistic and feasible to build a safe portfolio of dividend paying stocks that funds an investor’s retirement needs adjusted for inflation. Most planners seem to dismiss such a model due to the poor history of the S&P500 with respect to the stability of dividends and the fact that most advisors/planners don’t believe in active investing anyway. This is very unfortunate because such a model can provide a safe, comfortable, and peaceful retirement. I’ll continue to post on the portfolio value retirement model because that is what the majority of retirees will still use and there are ways to improve that model but the dividend retirement model is my preferred method of funding my retirement.