A Dividend Based Design Thread

Research on Safe Withdrawal Rates

Moderator: hocus2004

JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

A Dividend Based Design Thread

Post by JWR1945 »

A Dividend Based Design Thread

I recommend readers visit the Early Retirement Forum to read a thread started by hankjoy. It is critically important that you read the entire thread and not just the first post.
http://early-retirement.org/cgi-bin/yab ... 1084752712

Hankjoy suggests looking at retirement needs in terms of specific streams of income that must grow to match inflation. He invests for dividend income alone, never selling any shares of stock. He focuses on how much you need to spend on purchasing stock. When he mentions price, he is talking about the purchase price, not the portfolio's current price. When he mentions dividend yield, he is talking about the initial dividend yield, not the portfolio's current yield.

He pays very great attention to the price at purchase, the long-term stability of the dividend and the growth of the dividend amount so as to, at least, match inflation. His approach requires considerable price discipline. You must be willing to let an attractive investment get away if its price does not drop enough (at least, momentarily) to meet your target.

His approach requires a very careful selection of individual securities. Many of its advantages would be lost with indexing. Suitable securities fall outside of what many people would even consider. They include selected purchases of REITS. But they also include such things are royalty trusts, which are likely to provide you with an expensive education if you are not careful. [With royalty trusts, the danger is depletion. The key is to find trusts with appreciating assets that are highly unlikely to disappear, obviously a difficult task.]

Notice that this approach requires equity investment. There must be growth so that the dividend amount will grow. Because hankjoy never sells any shares of stock, he will never experience the sting of volatility: having to sell too many shares when prices are low.

This approach requires work and attention. Some companies have looked attractive from a distance while their balance sheets were falling apart. Benjamin Graham has mentioned Penn Central as an example. It was a star performer with a growing (purchase price) premium until just before it filed for bankruptcy.

I think that there is a speculative opportunity surrounding dividend based approaches as the Baby Boom generation retires. A disproportionate number of people should be looking for retirement income. This provides the kind of opportunity that Wall Street loves to fill (for a fee). My guess is that we are talking about an excellent opportunity for both retirees and speculators that will last for 5 to 10 years before it disappears through arbitrage. I do not expect the opportunity for retirees to disappear completely.

One thing that I will probably do (to quote unclemick) is "invest 45 bucks or so in Mergents Handbook of Dividend Achievers."� I am also interested in finding out about hankjoy's favorite book, "The Single Best Investment" by Lowell Miller.


Have fun.

John R.
Mike
*** Veteran
Posts: 278
Joined: Sun Jul 06, 2003 4:00 am

Post by Mike »

My guess is that we are talking about an excellent opportunity...
Does this mean that you think the high dividend method is an excellent opportunity to increase withdrawal rates?
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Mike wrote: Does this mean that you think the high dividend method is an excellent opportunity to increase withdrawal rates?
Yes. Just remember that it takes a lot of work to do it right. There are some traps hidden in the high dividend territory. One of the posts on the thread provided examples related to royalty trusts.

You have to be disciplined as well. It helps if you are driven by an overwhelming desire to observe the market intently and then to do nothing.

I think that the advantage will decline with time but not disappear entirely. Retirees need income streams. Other investors have other needs. It is wrong to state that there has to be a winner and a loser on every trade. There can be two people who both obtain their objectives. It is just that their objectives differ.

Have fun.

John R.
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

After ten years of putzing with building a small dividend income stream, I'm a little leery of my 'SWR' in the following sense: if I lump my dividend stocks with my balanced index the takeout number is 'higher' than splitting out and running the income along side pension and SS income streams. ?Retirement calculators are keyed to indexes - right? Sooo - a lump of div. types(utilities, oils, banks, drugs, etc) added in is shifting the numbers in an unknown manner? My gut tells me persuing div income requires treating it as income and struggling to quantify a div. growth - different than TSM or S&P 500.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

uncklemick wrote: Sooo - a lump of div. types(utilities, oils, banks, drugs, etc) added in is shifting the numbers in an unknown manner? My gut tells me pursuing div income requires treating it as income and struggling to quantify a div. growth - different than TSM or S&P 500.
Yes. I believe so. This is largely a limitation of our existing tools (i.e., Safe Withdrawal Rate calculators).

One key difference is that we usually run or TSM or S&P500 holdings down to zero when we use our calculators. I doubt that we would want to treat dividend income streams in the same way.

Have fun.

John R.
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson »

Mike,

The answer to this is, no.

People assume that royalty trusts with their 10% yields do it for you, but they do not as I've already pointed out to John. The dividend on oil trusts is dependent on the price of the barrel of oil and how that generates free cash flow. It doesn't necessarily follow that oil will steadily rise with inflation and that your dividend stream with too. While crude oil has beaten inflation over the very very long term, there are cases where a decade or more goes by and things don't look pretty. Therefore, one would want to be prudent and reinvest part of the income stream to cover inflation, just like one would wish for bonds coupons. You pay tax on the gross dividend but have to reinvest to cover inflation out of what is left. When you run the numbers using this reality, the return is nothing like 10% real after tax. It is also worth noting that you have a choice between US trusts which are a depleting resource until the wells are dry or a Canadian trust where they can buy more undrilled wells to replace the ones that are spent but a US citizen pays a 15% Canadian withholding tax (25% for UK buyers :-0 ). The Canadians are also putting in a new law that says that Canadians oil trusts must have 50%+ Canadian investors and they can force foreign buyers to sell their shares. A number of the best trusts have high foreign ownership and will come up against this problem in around 2006 I think.

REITs can deliver a consistent 4% dividend even in years when the capital value of the shares fall substantially. The bonds are connected to the property and their yields, not necessarily the value of the REIT security and so things remain steady. There are times when REITs pay out closer to 7% but that is not a consistent average. REIT return has been between 11%-13% depending on which 20 year snapshot you take. Inflation figures have been 5% over the past 30 years, which leaves 6%-8% real. Part of that tends to be in the increased cash flow/FFO and part from dividends. Therefore, even with historically high REIT returns (I don't use return estimates that high), you are unlikely to see a 7% avg yield. If we said REITs would deliver 10% long-term, and inflation were 4-5% long-term, then returns might reach 5-6% tops. You then balance that against other investments which pay less, many people do this with bonds which brings your overall returns down. You should also compare this math to the way swr works with stocks. Stocks delivered 10% long term with 5% inflation but only deliver 4% w/d rate because sometimes you are selling in at low valuations and that eats more of your capital away, thus lowering the s/w rate from 10%-5% = 5% down to 4%.

The other equities are common stocks which have fallen in price and are commanding a higher dividend purely because of the lower buy-in price. Such low prices are usually only indicative of a poor business model and bad finances & future prospects. A bad place to place your precious capital.

Bottom line, if you are willing to have a mix of equities and REITs, and avoid bonds and all other lower returning investments, you may be able to add another 1% to your w/d rate. I would suggest that one would do better aiming to buy superior growth businesses that deliver a better than average 10% return and that will raise your w/d $$$ by means of having more capital from better growth rates. Attention paid here is far more likely to add extra benefit that attention to withdrawal rateas. Also focusing on a large stock portfolio runs the risk that you will suffer greatly if the global markets take a tumble and don't recover for a decade or more. Most people will take a lower return in order to have a more balanced portfolio that enables them to sleep at night.

Petey

Mike wrote::
My guess is that we are talking about an excellent opportunity...


Does this mean that you think the high dividend method is an excellent opportunity to increase withdrawal rates?
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

Trusts are out of my league. But even with utilities, oils, banks, REITS and a few others - sleep at night is optional - I try to watch the div. end of the portfolio and use individual stock dips as chances to buy. Utilities, banks, and REITS especially are affected by the interest environment. When my favorite dog (con ed) dropped from 45 to 25, I bought more. Watching my overall portfolio value would upset me - if I took it serious - so I ballpark total only once in a while. This is a tough row to hoe - you've gotta 'enjoy' this and accept the gyrations.
Mike
*** Veteran
Posts: 278
Joined: Sun Jul 06, 2003 4:00 am

Post by Mike »

Also focusing on a large stock portfolio runs the risk that you will suffer greatly if the global markets take a tumble and don't recover for a decade or more. Most people will take a lower return in order to have a more balanced portfolio that enables them to sleep at night.
Since my current withdrawal rate is less than 1%, risk taking for the possibility of a larger withdrawal rate is not high on my agenda. I am more interested in mitigating risk. I recognize that building up the portfolio during good years is the best way to improve the odds of survival, but current valuations are unprecedented. I have opted for 63% equity as a play on the economic recovery, using strategies that have been associated with risk reduction in the past. However, I am constantly interested in evaluating alternative strategies. Flexibility is important as situations change over time.
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

Several points/confessions? After getting a little on the RE board for my hobby of 'laddering dividend stocks' - JWR applied a little math power which bucked me up some. Anywise after ten years and counting - ala Bogle I will 'press on regardless'. I call it a hobby and keep it 'capped' at 20% (80% is balanced index).

So my ulterior motive is to see if this forum can apply SWR analysis techniques to 'quantify' what I'm doing anyway and see how the method 'sacks up' against balanced index. My take of reading previous post leads to believe there is some interplay between SWR, historical div.s, and with a stretch even P/E 10 valuations.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

unclemick wrote: My take of reading previous post leads to believe there is some interplay between SWR, historical div.s, and with a stretch even P/E 10 valuations.
No stretch at all!

When Professor Shiller developed P/E10, he also had the P/D (price to dividend ratio or 1/[the dividend yield] ) in mind. In their paper to the Federal Reserve Board of Governors, Shiller and Campbell showed that both had predictive power for the medium term (10-year) stock market behavior but that P/E10 was a little bit better.

Dividends provide an income stream. That income stream provides the theoretical foundation for determining whether a stock price is reasonable. There are many other theoretically determined prices, but the starting point is always with this income stream (and showing that something else is equivalent).

P/E10 has worked a little bit better because dividends always come out of earnings and the dividend payout levels can be adjusted by management (subject to shareholder approval). P/E10 tends to get around the problem of unpleasant surprises known as dividend cuts. [Looking forward, many believe that dividend yields will be the better indicator because of accounting distortions buried in the reported earnings.]

Have fun.

John R.
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

The ducks are quacking, the ducks are quacking! Feed them - Blackrock Dividend Achievers Trust got another plug for retired investors in my Dick Young's Intelligence Report(I subscribe to it and Moneypaper). They claim to use the Mergent's 10 yrs of rising div.'s idea. $0.90/share payout and 0.85% expenses. Don't plan to buy any - unless the market future presents a big discont to NAV that makes it compelling. JWR's guess that the market may arbitrage the div. thing may be starting to play out. I expect to see more of the 'D' word in Wall Street's future offerings.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

unclemick wrote:They claim to use the Mergent's 10 yrs of rising div.'s idea. $0.90/share payout and 0.85% expenses.
I received my copy of Mergent's about two weeks ago. They mentioned that they had just licensed their selections to Blackrock to form the Blackrock Dividend Achievers Trust.

Mergent features Canadian companies in this copy (Spring 2004).

Have fun.

John R.
MacDuff
* Rookie
Posts: 31
Joined: Wed Jun 18, 2003 8:20 pm
Location: Centralia, WA

Post by MacDuff »

JWR1945 wrote:P/E10 has worked a little bit better because dividends always come out of earnings and the dividend payout levels can be adjusted by management (subject to shareholder approval).
I think there are some errors of wording here. Dividends usually come from earnings. However loss making companies can and often do pay dividends, at least for a while. In fact, loss making companies with negative book value can pay dividends. For tax purposes, these dividends may be classified as a return of capital, but they are still dividends from a corporate finance point of view.

Also, the second point, that management can adjust dividend payout levels with the approval of shareholders is also incorrect. Management can adjust, cancel, or otherwise mess around with dividends without any input from shareholders. Only the board must go along. Sometimes bond convenants may require that bondholders be solicited. And it may be that powerful large shareholders need to be consulted for political reasons before making a big dividend change. But no shareholder vote or other manifestation of shareholder democracy is needed.

Mac
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Thank you, MacDuff. You are right on both counts.

A better choice of words from me might have been Dividend Amounts.

Have fun.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

It turns out that hankjoy overstated what Lowell Miller recommends. Lowell Miller doesn't go anywhere near Royalty Trusts, Limited Partnerships and the like. He sticks with dividend paying stocks and REITS.

His formula is High Quality + High Yield + High Dividend Growth = the Single Best Investment.

I am at a loss as to why people assume that a high current dividend automatically means a low total return. I understand that some high dividend stocks no longer have dividend growth. But it seems reasonable that some excellent performers have high dividends and high dividend growth as well.

I am also at a loss as to why people would not wait for favorable prices. There are lots of companies to choose from. Since stock prices swing wildly each year, [with rare exception] you have to be able to pick a good entry point. It seems to me that you can always find something worthwhile at a favorable price.

Have fun.

John R.
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

Chickenheartedness. Perhaps not a real investment word but it 'minimizes regret'. Bernstein has recommended reading Edelson for value averaging and my trusty Moneypaper provides a % investec number for those who follow their value ave. method. I try to pick a suitable entry point for the DRIP - loosely DCA some, and throw in a few extra $ on perceived dips over 7-10 yrs.

In the land of little guts, little glory - just sent off to Moneypaper for ONE share of AT&T for a little DCA/value averaging in the future. Ala Ben Graham and Fama and French - .8 price to book - the div ain't bad either - NOT in Mergent's though.
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

JWR

Now that you have your Mergent's at hand - have you given any further thought as to how to approach the design of a dividend stream portfolio and what metrics/benchmarks as suitable bogies to measure against. off the top of my head - CPI change over time, ???DOW dividends instead of S&P, DOW utilities??. Once you assemble individual stocks - are you stuck with Bernstein's - The 15 stock diversification myth - problem of very widely scattered outcomes - OR is there data out there that shows a 'tighter' grouping for dividend stocks.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

I have taken a little time away from my reading in order to make my latest calculator modifications (Deluxe Calculator V1.1A02) and to collect some data.

I work in terms of real returns, where I use the CPI-U for making inflation adjustments, except as noted. I recommend your doing that. The long-term real growth rate of dividend amounts has been around 1.1% per year. The long-term real growth rate of earnings has been around 1.4% per year. These numbers are from memory and they may be off slightly. In addition, long-term means the really long-term where 50 years is the minimum. If your income stream consistently grows by 2% per year above inflation, you are doing exceedingly well.

[WARNING: There is an apparent contradiction between this observation and the observation that stocks have returned 6.5% to 7.0% in real dollars over the long-term. Looking at the Gordon Model, we would expect the total return to equal the initial dividend yield plus its growth. The dividend growth rate has been 1.1% per year in real dollars. The total return includes the initial yield.]

Using Mergent's, I have reexamined Lowell Miller's examples. At the time of his writing, they had been spectacularly successful. In terms of their performance since then, their total returns have been around 5.5% (nominal) or worse. The high priced winners are not nearly so high priced as before. Even when income streams are outstanding, there is a time to take money off of the table.

In contrast, I have noticed that there are a few Dividend Achievers such as Merck (which I have purchased recently) that have lost heavily in the recent past. I anticipate future price growth among such stocks. Surely, there has been bad news, but can it really be that bad when a company is signaling a bright future through its dividend policy?

I will be getting back to my reading shortly. An earlier discussion about the Gordon Model has clarified my thinking. I may be able to improve the discussion about measuring in terms of earning streams over what is typically presented. I seems to me that ignoring prices entirely is a bad idea. It is good not to overreact to price swings, but there are times when valuations are ridiculously high (see page 108 of the fourth revised version of Benjamin Graham's The Intelligent Investor) so that you should sell out to Mr. Market.

Have fun.

John R.
unclemick
*** Veteran
Posts: 231
Joined: Sat Jun 12, 2004 4:00 am
Location: LA till Katrina, now MO

Post by unclemick »

JWR

If you go back to my old 'dividend stock ladders' where you graciously helped anaylize my numbers - you'll notice "Mr Market' has a wicked habit of cash out mergers, spin - offs and combo's there of. So I took more cap gains than intended - which ?luckily? where close to my reinvested dividends over time. Also - the list of 1994 vs 2003 has changed considerably - but many stalwarts are still there (Merck).

So close to your idea - Try to find a 'good/decent price' entry point, reinvest divs., maybe DCA a little, take money off the table(not always voluntarily) along the way and track your (dividend stream?) progress over time against CPI-U.

I'm thinking drop my - if they don't double in 7-10 yrs, take them out and shoot them(er sell). Think long term but tweak if some get overvalued along the way. Oh I also don't rebalance or specifically diversify because of larger part of toal portfolio is balanced index. Interestingly - when you use div growth plus current div as a metric - different stock groups at different times seem to fall to good value - at least during last ten years. I'm also looking at things like Merck (have Glaxo, Lilly bought in 94). Also - do not automatically sell when they fall from the list - loosely watch taxes and progress - sometimes too long.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Watch out, unclemick. We might begin to refer to you an investment genius in the not too distant future.

I am continuing to read Lowell Miller's book about The Single Best Investment. I have another book of his as well. There is a lot of there there so to speak. I am highly impressed. Most of it fits in nicely with what you have already been doing on your own.

I am cautious as well because I know how easily subtle errors can creep into any investment strategy and corrupt the whole. I am more strongly convinced than ever that hankjoy gravely misrepresented Lowell Miller's approach when he mentioned exotic investments such as Royalty Trusts, Timber holdings, Limited Partnerships and so on. That is why I am so cautious. IMHO, a sound dividend strategy has already been corrupted by such influences.

There are distractions as well such as the approach of The Dividend Growth Investment Strategy. The concept was generally sound. But it ignored valuations entirely. IMHO, that is a fatal flaw.

There are lots of loose ends. For example, Lowell Miller prefers to use a select group of higher yield stocks as a substitute for bonds in a balanced portfolio. I need to look into that very carefully. His arguments were compelling when written. I don't know if they still hold up. I know how they could break down: a meaningful loss of capital that would destroy the rebalancing of one's portfolio.

You have mentioned rebalancing. More specifically, not rebalancing. I think that you have something there. The benefit in the Historical Surviving Withdrawal Rates of the past was small: of the order of 0.1%. Academic studies and our calculators all rebalance at zero cost. Actual costs are real and, as Bogle would point out, they compound. Does rebalancing pay for itself? In terms of risk and reward, most likely no.

There is the matter of separating what is important during accumulation versus what is important during retirement. There is even the matter of distinguishing those elements of uncertainty or risk that are knowable (at least, in a statistical sense) and that we can control from those that we cannot control but must accept.

Have fun.

John R.
Post Reply