Notes starting from September 12, 2009

Updated: October 30, 2009.

Fixed Allocation Investors

Those choosing to stick with a fixed stock allocation will do well to take advantage of my Retirement Risk Evaluators, the Year 30 SWR and Year 15 SWR buttons on the left. It tells you the full range of likely outcomes based on current valuations.

Traditional Safe Withdrawal Rate studies fail miserably by not taking valuations into account. They are highly misleading.

My calculators use the historical sequence method. This is the best approach when it can be used.

I also have Monte Carlo simulations that take valuations into account. They can extend the range of investigation. They are the Scenario Surfer and the Investment Strategy Tester. They include two forms of mean reversion as well.

Although I recommend keeping risk steady as opposed to stock allocations, I allow you to investigate all options fully.

Today’s P/E10 is just above 18.

Sometimes Yes, Sometimes No

Does a higher total return translate into a higher Safe Withdrawal Rate? Sometimes, it does.

Look at the Year 30 SWR [button on left] Retirement Risk Evaluator. As P/E10 rises, Safe Withdrawal Rates fall. Yet, looking at the Stock Returns Predictor [button on left], rising P/E10 means falling real, annualized, total returns.

At today’s valuations of P/E10=18, the most likely Year 10 total return is 3.9%. This has the greatest influence on Safe Withdrawal Rates. The Year 20 and 30 total returns are higher. The Year 30 Safe Withdrawal Rate is 4.3% (plus inflation) with a high stock allocation.

Now visit the TIPS Table [button on left]. With a 3.5% interest rate, which is also its total return, the 30-year payout is 5.44%. This is substantially higher than the 4.3% available from a portfolio with a high stock allocation.

The difference is volatility. At the same volatility, a higher total return delivers a higher Safe Withdrawal Rate. At a lower volatility, a lower total return can deliver a higher Safe Withdrawal Rate.

Outstanding Site

Rob Bennett is hitting home runs. Visit his PassionSaving.com site and his A Rich Life blog. You will enjoy his offerings. They are outstanding.

PassionSaving.com
A Rich Life blog

Avoid Rebalancing

Here are two classic articles. Rebalancing is a horrible idea whose time has passed. Control risk by shifting allocations in accordance with valuations.

Shun Rebalancing
Orders of Magnitude

Historically Typical P/E10

Norbertc, in a post at the Morningstar Portfolio Design/Management discussion board, points out that “the historical ‘P/E10’ average value from 1881 to the present is 16.34.” He is right.

I am referring to the “modern” era starting from 1921 when I identify P/E10=14 as historically typical.

This should not cause a big change in your investment decisions. The effects of P/E10 are gradual.

Expensive Market

Stocks are expensive once again with P/E10=19.

I went to ETF connect. I found that today’s distribution yield for DVY is 3.77%. It is 7.14% for PFF. I put these numbers into my Simplified Automatic Allocator. I used a DVY dividend growth rate of 5.5% nominal, same as for the S&P500, and 0% for PFF. I assumed 3% per year inflation. I started with 50% DVY and 50% PFF. I used 2% TIPS for my cash management account.

The result: you can safely withdraw 5.1% (plus 3% inflation) indefinitely if you avoid selling shares for income. You must allow for occasional dividend shortfalls, possibly reaching as high as 25%, followed by dividend growth exceeding the rate of inflation. Your income level would always remain above 3.8% of your original balance (plus 3% inflation). That would be the worst case.

How Often to Buy and Sell?

The Scenario Surfer tells us a lot about the market. Although it normally tells us to trade infrequently, there are exceptions. In some markets, you end up buying and selling quite often. Typically, you would shift allocations about once or twice in a decade. Sometimes, valuations shift more dramatically. Today’s market might be such a market. Only time will tell.

Is $200000 Enough?

How much do you need? For a traditional retiree, $200000 can be enough. A younger retiree needs more. Partial retirement may be a better choice.

Here is an oldie but goodie from August 2007.

Is $200000 Enough?

Hand Calculations for a Dividend Blend

You can still determine what works with a dividend blend even if you are unable to build a spreadsheet.

The formula for the real income of an investment at year N is:

Inflation adjusted dividend income = (initial dividend amount)*{[1+(nominal dividend growth rate)]^N}/{[1+(inflation rate)]^N}

Typically, you would use a nominal dividend growth rate of 5.5% per year in the absence of other information and 3% per year inflation. If so, the formula becomes:

Inflation adjusted dividend income = (initial dividend amount)*(1.055^N)/(1.03^N)

With preferred stock and/or bond income, use a nominal dividend growth rate of 0%.

Make you allocations and list your initial dividend amounts. Calculate each investment’s inflation adjusted income at Years 0, 5, 10, 15, 20, 25 and 30. For each of these years, add the incomes together. You will see how the income starts high, becomes lower and then grows once again.

You make a dividend blend work by taking some of the excess funds from the early years and use them to fill in the shortfall in the intermediate years.

This is how you do it by hand.

When P/E10=19

Stock Returns Predictor:
Most Likely Year 10 return is 3.5% (real, annualized, total return).

Year 30 Safe Withdrawal Rate with fixed allocations:
4.2% (plus inflation) with 50% stocks and 4.1% (plus inflation) with 80% stocks.

Year 30 Safe Withdrawal Rate with variable allocations:
4.8% (plus inflation) with Switching A and 4.9% (plus inflation) with Switching B.

TIPS-only portfolio with 2% TIPS lasting for 30 years of withdrawals:
4.46% (plus inflation).

Straight dividend screen at Morningstar (average growth and profitability, outstanding financial condition): 23 securities listed with an above 4% dividend yield (continues indefinitely).

Recent Dividend Blend: 5.1% (plus 3% inflation) with DVY and PFF (continues indefinitely).

Capitalization Weighting

The S&P500 index uses capitalization weighting. It has a lot going for it. But it is not perfect.

Capitalization weighting has the advantage that it avoids selling to maintain its structure. This cuts expenses. This advantage is important. It can be of primary importance.

Capitalization weighting has a disadvantage as well. It ignores dividends. Dividends are collected and reinvested according to prices within an index. They are NOT applied to the companies that generated them. This overemphasizes price appreciation as opposed to the total return of individual companies.

Those who invest in individual stocks have greater flexibility. They can reward dividend payers if they so choose.

What Is Capitalization Weighting?

Not everyone is familiar with investment jargon. Capitalization weighting is a term that applies to the S&P500 index. It means that you add up each company’s contribution to the total value of the index. You calculate each company’s contribution by multiplying its current price by its total number of shares.

A Fine Point

You must match the time frame of a prediction with the prediction itself. You cannot extrapolate short term stock market data into the medium and long term. This was the most important lesson in my study of the Forsey-Sortino model in Current Research E: Managing Downside Risk in Financial Markets. Monthly data or even annual data is not enough for predicting what happens in ten or twenty years. You must use ten year data to estimate what happens in ten years. Similarly, you must use twenty year data to estimate what happens in twenty years.

Current Research E
Monthly Returns

Free Lunches for Everyone

There are free lunches out there. Lots of them. From October 2007.

Free Lunches for Everyone

While Working on a Prototype

I am working on a prototype of a new calculator.

I was amazed at how large the stock market spread is at Year 10. Then I realized that applying valuations cuts this in half. At Year 10, the total stock market spread after applying valuations is roughly plus and minus 6% (real, annualized, total return). If you fail to account for valuations, the spread is huge.

Mean reversion helps, but it does not do the entire job. Making things more difficult, stock returns are highly correlated from one year to the next, reducing the effective number of degrees of freedom.

The prototype will tell us about the sequence of returns. In this version, I separate dividends from the total return so that the effect of purchasing more shares becomes apparent. It will be interesting to see how it turns out.

Still Safe at 5%

I brought up the Investment Strategy Tester [button on left]. Withdrawing 5% of the original balance (plus adjustments to match inflation) is still safe for 30 years using Valuation Informed Indexing. I looked at stock (S&P500 index) allocations of 100%-80%-50%-20% and corresponding P/E10 thresholds of 8-10-18. I assumed that P/E10=20 in a Bear Market and that TIPS delivered 2% (real) interest.

I also looked at a fixed allocation of 50%-50% stock and TIPS. It eventually failed about 25% of the time. But it performed well over the first decade.

Retirees just starting out who are not yet ready to embrace Valuation Informed Indexing would do well to begin with a 50%-50% allocation. It too will allow 5% (plus inflation) withdrawals, but only during the first 10 to 15 years. Stock prices should drop dramatically somewhere during that interval, at which time they should shift to a high stock allocation.

Refusing to See the Obvious

This is a classic article that I need to mention from time to time. The statistics behind P/E10 projections are overwhelming. It takes effort to miss the obvious.

Refusing to See the Obvious

Confidence Limits

If you look at the Stock Returns Predictor Year 10 results, you will see a total spread (outer confidence limits) of plus and minus 6%. This is huge, but only half of what you would expect without including valuations and mean reversion.

If forecasters were to include confidence intervals, we would see much less nonsense. Who would be interested in a precise one-year forecast based on the Gordon Equation if he knew that its outer confidence limits were plus and minus 40%?

Dividend Modeling

Dividends increase steadily in NOMINAL dollars over time. Year 10 total return vary with valuations in (inflation adjusted) REAL dollars. I have recently been modeling the two. The one thing that really stands out is this: dividend yields vanish if you assume a high rate of (total) return throughout a decade.

How low can dividend yields go? In January 2000, they were only 1.16%. [Use the S&P500 Dividends button on the left.]

A Time for Skill

This last decade has been a time for skill. Those paying attention to valuations have done well. Others have done OK. Those sticking to high stock allocations have not done as well.

When P/E10=26, as it did for so long in the first decade of the 2000s, the 30-year Withdrawal Rates ranged from 3.1% to 6.1% with an 80% fixed stock allocation and 3.6% to 5.6% with a 50% fixed stock allocation. With Switching A and Switching B, the ranges are 4.2 to 6.0% and 4.4% to 6.1%, respectively.

This demonstrates the advantage of skill. Varying allocations can lift the Safe Withdrawal Rate by more than 1%.

Paying attention to dividends helps as well. It adds another dimension to your choices. Dividend strategies would have allowed you to retire with a continuing withdrawal rate of 5%.

All of these are percentages of the original balance. They are adjusted to match inflation. Dividend strategies are subject to occasional setbacks due to dividend cuts and/or delayed growth. They are followed by full recovery and growth.

Predictability and Dividends

If you remove dividends and look only at prices, you find that there is a good amount of predictability in terms of valuations. When you reinvest dividends and look at total return, there is more. Dividends themselves show the highest degree of predictability.

Dividends versus Capital Appreciation

Real Growth of Dividends

Dividend amounts rise steadily in terms of NOMINAL (without adjustments for inflation) dollars. They rise erratically in terms of REAL (with adjustments for inflation) dollars. Still, if you are interested in seeing numbers that have the effect of inflation built in, read this article from when I was first trying to characterize dividends as a function of valuation.

Dividend Growth Rates

John Walter Russell, RIP

This is Rob Bennett, owner of the site at www.PassionSaving.com.

I have sad news to impart to regular readers of this site and to all members of the Retire Early and Indexing discussion-board communities, which John Walter Russell (my good friend and the author of all of the material that has appeared at this site thus far) has served so well for over seven years now. John died on Tuesday, October 27, 2009, of emphysema. He was 63 years old. He had been ill for some time but the death came as a shock. He found much joy in the last seven years of his life generating the material that appears at this site and helping thousands of middle-class investors discover a more effective and reasonable way to invest for retirement.

John expressed a desire that this site be transferred to me upon his death. The site is now in my hands.

My intent for the time being is to leave the site as is, so that those who learn of it may tap into all the wonderful materials that John provided so generously to all of us. In my discussion with John's wife Karen, she noted that John viewed his work here as akin to missionary work (John was a Christian). I think that goes a long way to explaining why the research done here is so special.

John was not employed by The Stock-Selling Industry. He was free to tell us the straight story re how stock investing works in the real world. If John's work had been better known, we would not be suffering an economic crisis today. Getting John's work better known is the quickest way out of the crisis that threatens to devour both our economic and political systems if we do not soon get people of influence to take serious, positive steps. My primary goal for the site is to spread the word about the wonderful insights that John mined here. My secondary goal is to develop those insights even further than John was able to develop them during his short journey through the Valley of Tears.

I would like to be able to find someone capable of updating the site occasionally, co-developing new calculators and responding to reader questions. John's shoes are extremely hard shoes to fill. So I doubt that I am going to be able to pull that one off. There are two individuals who come to mind as long-shot possibilities. But the odds are strong that we are not going to be able to find someone qualified to advance John's work into new areas until the Valuation-Informed Indexing strategy and the Rational Investing model for understanding how stock investing works are far better known and appreciated than they are today.

In the event that I am not able to find someone to take on research duties for our communities, I may elect to move some or all of the materials at this site to my site (www.PassionSaving.com). John intended for his missionary work to help all middle-class investors come to a better understanding of how stock investing works than has been provided by The Stock-Selling Industry. Such investors number in the millions. My job is to inform them all that there is a better way. It will be easier for me to build one site than two.

So my expectation is that, unless I am able to find someone to take on research duties, I will ultimately move some or all of the material that now resides here to my other site. I don't expect to do that quickly, as it would take a good bit of time to do the job properly and I have much else on my plate today.

A third possibility is that I may leave the site up in its present form indefinitely. It would of course remain a wonderful resource for all working in the investment advice field to link to for many years to come. One thing that I am sure of is that I want those researchers who see the need to explore the path that John pioneered to have access to the materials they need to do so successfully.

If anyone reading these words knows of someone who could help us with research work, I would be grateful if you would contact me (you can use the "Contact Us" button at the left side of this page to do so).

My favorite section of this site has always been the "Letters to the Editor" section. Readers who have questions or observations should continue to send them in. If the questions are technical in nature, it is unlikely that I will be able to respond effectively (my skills are in journalism, not statistics). But I will of course be happy to share what I know (I was intensely involved in the development of all of the calculators so I might be able to provide background despite not possessing an understanding of the technical issues) or what I think. I also can invite other community members who possess a better understanding of the technical issues to provide input.

Please say a prayer for John's soul. And please say a prayer asking for strength for Karen to bear her loss. If you have any prayers left in you after that, please say one for my efforts to spread word of John's work far and wide and one for a softening of the hearts of those who have somehow come to believe that it is in their best interest to spend their life energies blocking that effort.

John was a Numbers Guy. I'm the Emotions Guy. But I am proud to be able to say that I am not so dumb about numbers that I couldn't figure out that John's numbers were calculated through the power of love. That's what makes them so exciting and valuable and special and -- why not just say it? -- accurate.

Rob

Note: Links to tributes to John posted by his many friends in the Retire Early and Indexing discussion-board communities (along with a few stinkbombs tossed in by the Goon Element!) are set forth in the "Letters to the Editor" section of this site.

Notes Index Starting from June 25, 2009

Notes Index starting from November 23, 2007