April 17, 2007 Letters to the Editor

Updated: April 27, 2007.

Inside-the-Box Thinking

I received this letter from Rob Bennett.

The following words appeared in your Note entitled "Inside-the-Box Thinking":

"You can choose to focus on dividends and income streams. Doing so, you can reap most of the market's long term Investment Return--even today."

I understand the point being made and to a large extent I am in agreement with it. Investing in high-dividend-payers makes sense to me.

What I have a hard time getting a handle on is the extent of the lost-opportunity cost paid by investing heavily in dividend-payers at a time of high valuations. If we knew that prices were going to come down quickly, the smart thing to do would be to hold off. If we knew that prices were going to remain high for a long time, the smart thing to do would be to invest in dividend-payers now. The reality is that we don't know whether prices are going to come down quickly or not.

I can accept that one can do "okay" by investing in dividend-payers now. But even dividend-payers will offer better value propositions when prices are lower. Are there effective ways of analyzing whether it is better to buy dividend-payers now or to wait?

HERE IS MY RESPONSE

The Total Return of an investment equals the Investment Return plus the Speculative Return. The Speculative Return is the effect of changing multiples, such as the price to earnings ratio. Mathematically, assuming only that the multiples are bounded, the Speculative Return eventually becomes very close to 0% per year (annualized). Mean reversion prevents multiples from expanding without limit. Over long periods of time, the Speculative Return becomes very small.

However, the Speculative Return is not identically zero. Over the course of the twentieth century, it was 0.7%.

The long term return of the stock market has been 6.5% to 6.8% after adjusting for inflation. Removing 0.7% for the century long Speculative Return, the remaining Investment Return is very close to 6.0%.

Under typical conditions, the Gordon Equation approximates the long term Investment Return very well. It does not do nearly so well at extreme conditions.

The Investment Return, according to the Gordon Equation, equals the initial dividend yield plus the dividend growth rate.

The dividend growth rate does not depend on the price. It is 1.1% to 1.5% per year (annualized) after adjusting for inflation. Subtracting this from the long term Investment Return of 6.0% leaves a typical dividend yield of 4.5% to 4.9%.

Today’s prices are double what is typical. Today’s long term Investment Return is 2.2% to 2.5%. High prices have removed 2.2% to 2.5% from the potential return of today’s stock market.

Adding the dividend growth rate back into the equation, today’s long term Investment Return is between 3.3% and 4.0% after adjusting for inflation.

Today’s dividend yield of the S&P500 index is between 1.5% and 2.0%. This is because of a lower than normal payout ratio. The lower dividends should show up in faster earnings growth, but it is far too early to tell whether or not it actually does. Early reports are contradictory, initially indicating a lack of earnings growth, now indicating accelerated earnings growth.

Notice that prices influence standard calculations twice, both in the dividend yield and in the Speculative Return adjustment. Because of today’s high prices, we can expect the Speculative Return to be negative (reducing the Total Return) over the next decade or so. Dividend and Income approaches will not suffer this penalty.

It works the other way during times of favorable valuations. The Speculative Return gives an extra boost as prices rise. Dividend and Income approaches fall behind.

Let us look more closely at the Speculative Return.

A price doubling increases returns by 7% per year (annualized) over a decade. A price cut to one half subtracts 7% per year (annualized) over a decade. In today’s market, with a 3.3% to 4.0% Investment Return, the traditional Total Return would be -3.7% to -3.0% per year (annualized) if multiples returned to normal over a decade.

The Stock Returns Predictor indicates a more favorable result. The detailed assumptions behind the original calculations are overly pessimistic. More dividends are purchased at lower prices.

When prices return to highly favorable valuations, at one quarter of today’s multiples (P/E10=7 at that time), the dividend yield will rise from the typical 4.5% to 4.9% to 9.0% to 9.8%. Adding 1.1% to 1.5% per year (real) dividend growth, the Investment Return would rise to 10.1% to 11.3% per year (annualized) after adjusting for inflation.

Total return would receive a boost of 7% per year (annualized) over the following decade as multiples expand. This ranges from 17.1% to 18.3%.

The Stock Returns Predictor shows the Most Likely return as 17.2%, ranging from 11.2% to 23.2%.

SUMMARY

The Gordon Model does a good job of explaining what to expect over the next decade or so (from 5 to 15 years). You should exclude the Speculative Return component with Dividend and Income strategies. You should include the Speculative Return for Total Return strategies.

Prices show up in the dividend yield of the Investment Return. Prices show up in the Speculative Return as well.

Standard Total Return approaches receive a boost during times of low valuations as multiples expand. At today’s valuations, we can expect the Total Return to be dragged down as multiples contract.

Dividend and Income strategies currently deliver MORE than today’s Investment Return when adjusted for prices. They deliver almost as much as the typical long term Investment Return.

Major incorrect assumption in use of TIPS

I received this letter from Jim.

Just want to mention a major incorrect assumption you are using with regard to TIPS.

Because of substitution and hedonic adjustments factored into the officially reported CPI, the CPI does not truly track the rate of increase in living costs. The CPI understates true inflation or cost of living by several %'s. Therefore, if TIPS are paying a coupon rate of 2%, it does not mean that TIPS are returning 2% over the true rate of inflation/cost of living. Shadowstats.com is a good website for explaining the substitution and hedonic adjustments placed into the CPI which dramatically understate inflation.

I loved your website and had been studying it in depth, but when I learned more about hedonic adjustments and substitutions in the CPI, making the TIPS coupon rate meaningless and in reality a negative real rate of return, it stopped me dead in my tracks. I'm thinking there might be a lot of other people seeing this same thing. Otherwise, your website is the most scientific I have found yet on the web.

Thanks.

HERE IS MY RESPONSE

Thank you for your kind words. I appreciate them.

This is the adjustment that you are seeking: Reduce my withdrawal rates by 0.3% to 0.4% for portfolios that include stocks and 0.6% for TIPS-only baselines.

I do not expect you to change your opinion. I will simply point out that there is another side of this argument. Many economists contend that the CPI overstates inflation. Congress split the difference between the original CPI rates and the 2% reduction recommended by Federal Reserve Board of Governors Chairman (at that time) Alan Greenspan. The fact is that we expect quality improvements even in the absence of inflation. Today’s poverty level corresponds to the upper middle class fifty years ago.

Mathematically, the adjustment that you seek is 1% per year (annualized) over 30 years. This reduces buying power by 1.01^30. That is, divide by 1.348 or multiply by 0.742.

I pressed the Year 30 SWR button. I looked at the Likely Success rates with 50% stocks and 80% stocks with Year 30 balances of 70%, 80% and 100%.

With 80% stocks and P/E10=28, the Likely Success rate that maintains 70% of the original balance is 3.22%. It is 3.09% at 80%. It is 2.83% at 100%. The differences between 100% and the lower balances are 0.39% and 0.26%. The adjustment to reach 74% would be 0.3%.

With 80% stocks and P/E10=14, the Likely Success rate that maintains 70% of the original balance is 5.77%. It is 5.65% at 80%. It is 5.41% at 100%. The differences between 100% and the lower balances are 0.36% and 0.24%. The adjustment to reach 74% would be 0.3%.

With 50% stocks and P/E10=28, the Likely Success rate that maintains 70% of the original balance is 3.26%. It is 3.11% at 80%. It is 2.80% at 100%. The differences between 100% and the lower balances are 0.46% and 0.31%. The adjustment to reach 74% would be 0.4%.

With 50% stocks and P/E10=14, the Likely Success rate that maintains 70% of the original balance is 4.74%. It is 4.59% at 80%. It is 4.28% at 100%. The differences between 100% and the lower balances are 0.46% and 0.31%. The adjustment to reach 74% would be 0.4%.

The 30 Year withdrawal rate of 2% TIPS is 4.46%. The 30 Year withdrawal rate of 1% TIPS is 3.87%. The difference is 0.6% (actually, 0.59%).

Early retirees would do well to consider these adjustments in their planning. Most would like to maintain their relative standard of living, not simply maintain their existing standard of living.

The best way to handle a refinement of this sort is at the edges. That is why we have five levels of probability built into the Retirement Risk Evaluators (30 Year SWR and 15 Year SWR buttons). You might insist upon the full level of safety (Safe Withdrawal Rate) with the standard CPI, but permit a lesser level of safety (Reasonably Safe) when it comes to the 0.3% to 0.4% adjustments.

30 Year Retirement Risk Evaluator (P/E10)

I received this letter from Larry.

In your 30 Year Retirement Risk Evaluator what is the meaning of the "P/E10"? Does that mean the same thing as the P/E of the market?

Thank you.

HERE IS MY RESPONSE

Thank you. You have asked a good question.

P/E10 is the current price of the S&P500 index divided by the average of the previous ten years of earnings. All amounts are adjusted for inflation.

I keep an updated conversion between the current S&P500 index level (price) and P/E10 on the Stock Returns Predictor. The relationship varies slowly over time since smoothed earnings grow.

Benjamin Graham came up with the idea of smoothing earnings over as much as a decade when evaluating individual companies. Professor Robert Shiller, author of “Irrational Exuberance,” applied this idea to the S&P500. Professor Shiller includes P/E10 in his S&P500 database, which is a widely accepted standard.

Single year earnings fluctuate dramatically. Smoothed earnings do not. P/E is an unreliable indicator. Even a healthy company will have an occasional earnings shortfall. P/E10 is a reliable indicator. The P/E10 of the S&P500 is the best indicator of the overall market that I have found. I have identified two competitive alternatives: smoothed dividend yield and Tobin’s q.

Letters to the Editor in 2007

Letters to the Editor in 2006

Letters to the Editor in 2005