Accumulation Stage: Edited

How should you invest BEFORE retirement? Here are some insights for the accumulation stage.

I have edited these posts heavily. They are from the old SWR Research Group discussion board at the NoFeeBoards.com web site.

POST A. Accumulation is Different

Rob Bennett had made the following remark:

“What makes this confusing is that SWR analysis has generally been designed to be applicable only to retirees. It is not appropriate for non-retirees to draw conclusions from SWR studies as to how to invest during the asset accumulation stage. Everyone understands this is so as a technical matter. The reality, however, is that many have improperly drawn inferences from SWR analyses as to how to invest during the accumulation stage. There are many posts in the archives showing this to be the case.”

I proceeded to collect accumulation data.

I started with an initial balance of $1000. I deposited $1000 per year in subsequent years.

These results compare switching portfolio allocations to sticking with a 100% stock portfolio. I investigated using 2.0% TIPS, 2.0% ibonds and commercial paper as the component other than stocks.

I used the optimal switching program from previous studies for the distribution phase. The P/E10 thresholds were 9-12-21-24 and the corresponding stock allocations were 100-50-30-20-0%. I set the interest rate for TIPS and ibonds at 2.0%.

30-Year Results

With rare exceptions, an investor was much better off sticking with 100% stocks. Even with the exceptions, any advantage in favor of switching was small. The worst case without switching was 1952. The all-stock allocation had a final balance of $45231. Switching with 2% TIPS, 2% ibonds and commercial paper produced balances of $53121, $53044 and $51289, respectively.

Balances with 100% stocks frequently exceeded $100000. The best cases were $215873 in 1932 and $202734 in 1970. With switching, there were only a handful of final balances above $100000.

Ibonds produced the best results with switching.

Final balances were almost entirely independent of the percentage earnings yield 100E10/P. R-squared values were less than 0.01.

Switching improved the worst case results only a small amount while taking away the upside.

15-Year Results

The behavior at 15 years was different.

The 100% stock portfolio’s final balance y increased with the percentage earnings yield x (i.e., 100E10/P). The equation was y = 2794.7x+9956.9 and R-squared was 0.3279. My visual estimates of the confidence limits are plus $20000 and minus $15000 at lower levels of earnings yield and minus $25000 at high levels of earnings yield.

With 2% TIPS and switching, the equation was y = 460.18x+21200 and R-squared was 0.1132. My visual estimates of the confidence limits are plus $8000 and minus $5000.

Switching is superior when the final balances are low. If the final balance without switching exceeds $30000 or so, switching severely limits returns.

In essence, switching produces a non-trivial improvement when overall stock returns are low, but with a serious penalty on the upside. The entire decade of the 1960s produced better results with switching than without. The decade from 1925-1934 favored switching as well.

The percentage earnings yields [at the very start of a historical sequence] during which switching produced better results was typically between 4% and 6% (i.e., P/E10s of 17 and above).

The 15-Year results favor switching at today’s valuations, but not at valuations typically found in the historical record.

POST B. Accumulation is Different (Follow-On)

Since switching helps in times of high valuations, I extended my investigation.

First, I invested $1000 per year into 100% stocks for 15 years and then I left the account untouched [i.e., I reinvested all dividends] for another 15 years.

The worst case in the modern era (1923-1972) with this approach occurred in the 1967 historical sequence. The final balance was $12594.

Next, I invested $1000 per year into switching with 2.0% TIPS [and stock allocations of 100-50-30-20-0% and P/E10 thresholds of 9-12-21-24] for 15 years. Then I invested the entire balance into stocks for another 15 years, leaving it untouched and reinvesting all dividends.

The worst case in the modern era (1923-1972) with this approach also occurred in the 1967 historical sequence. The final balance was $17361.

This particular comparison favors switching. Looking at the rest of the data in the modern era, switching during the first 15 years provides a small improvement (less than $20000) in the final balance about half of the time. It costs a large reduction (about $40000 to $160000) in a fourth of the conditions.

The years 1959-1971 favored switching. The years 1936-1958 favored sticking with 100% stocks. The years 1940-1954 strongly favored switching.

Finally, out of desperation and also to be thorough, I reversed the sequence of events. I made the $1000 per year deposits into an all-stock account for 15 years. Then I changed to switching with 2.0% TIPS (and with 9-12-21-24 for the P/E10 thresholds and 100-50-30-20-0% stock allocations) for the final 15 years, but without making any additional deposits.

This time I hit pay dirt. The worst case balance after 30 years was $36284 in 1963. Compare that with the previous final balances of $12594 and $17361 (both in 1967). The upside was limited once again, this time to $82912 in 1942 for the 1923-1972 era. The improved balances under worst case conditions justify sacrificing some of the upside.

POST C. Dollar Cost Averaging Historical Returns

There were several instances in which dollar cost averaging would have lost money at 15 years. This has caused me to reexamine my recommendations for investing today.

Normally, I would have recommended dollar cost averaging into stocks. But today, we are in the early stages of an extended bear market. We need to revise our short- and intermediate-term recommendations. These are general updates, not just a few special cases.

Data Collection

I collected data on the balances at 5, 10, 15 and 20 years.

Summary

There are several sequences that did worse than inflation-matched cash equivalents (i.e., TIPS and/or ibonds at an interest rate of 0.00%). Those in recent times started in the 1960s. Poorly performing portfolios included the year 1975 as part of the dollar cost averaging period.

There were several additional sequences in earlier times during which dollar cost averaging did worse than inflation-matched cash equivalents.

The period of poor performance easily extends to ten and fifteen years.

Dollar cost averaging was superior most of the time. It was not superior all of the time. There are times when it is better to preserve capital.

There is a weak relationship between the dollar cost average bonus and the percentage earnings yield (at the start of a sequence). R-squared ranges from 0.18 to 0.26. It takes an earnings yield above 6 to guarantee a positive advantage over 10 or 15 years. This corresponds to a maximum P/E10 level of 17. It takes an earnings yield above 8 to assure an advantage at 5 years. This corresponds to a maximum P/E10 level of 12 to 13.

Extrapolating (visually): at today’s P/E10 levels, the odds of getting an advantage from dollar cost averaging (versus inflation-matched cash equivalents) is only slightly better than 50-50%. [This is with periods up to 20 years.]

At today’s valuations and with today’s stock market outlook, it is a good idea to consider alternatives to dollar cost averaging into the market as a whole. One attractive approach is to purchase income streams directly by focusing on dividends (i.e., using one of our dividends-based strategies). Buying TIPS and ibonds (while being careful about taxes) is an attractive alternative, especially for those who may need to convert (some of) their holdings directly into cash quickly.

POST D. Switching during Accumulation

We already have two important findings related to switching stock allocations during the accumulation stage. First, dollar cost averaging into a 100% stock portfolio is always best when the time frame is long. Second, switching allocations between stocks and TIPS at a 2% interest rate makes sense when you have already accumulated a large amount of money.

We have examined what happens when we start from scratch. If we break a 30-year investment period into two 15-year periods, we should start out by dollar cost averaging into a 100% stock portfolio. After 15 years, varying stock allocations according to P/E10 provides enough protection on the downside to compensate for the reduction of the upside potential.

In contrast: varying allocations during the first 15 years reduces performance without providing protection. Varying allocations over the entire 30-year period reduces performance without protecting the downside.

When switching is used during accumulation, it is with money that has been invested in the first 15 years. A portion of the balance at year 15 is set aside for switching. That portion protects the downside. The remainder is included in the 100% stock portfolio with dollar cost averaging for new deposits.

During accumulation, there are no withdrawals for the allocation switching account. Previous optimizations of the switching algorithm (i.e., thresholds and allocations) have been during distribution when there are withdrawals. Does this make a difference? The answer turns out to be NO.

Initial Survey

In my first survey, I varied the allocation percentage when P/E10 fell between 12 and 21.

The tradeoff region appears to be for percentage allocations between 30% and 70%.

The minimum growth multiplier generally decreases as the stock allocation increases.

The maximum growth multiplier generally increases as the stock allocation increases.

These are the effects of holding larger stock allocations: the spread of the data increases.

Follow-up Surveys

I examined three middle level stock allocations: 30% and 50% and 70%. In each case, I varied the higher P/E10 threshold between 12 and 26 in increments of one.

At 30% and 50% middle range stock allocations, we select a P/E10 threshold of 21.

When we look at the 70% middle range stock allocation, P/E10 thresholds of 18 through 21 are best.

Summary

I stopped my surveys at this point. The old allocations are still valid. The data behave in a manner similar to what we have seen before. Allocations are not excessive sensitive to P/E10 thresholds. [The older investigations were during distribution when there were withdrawals. In this investigation, there were none.]

It is very important to learn the right lessons from these investigations. We do not depend upon any particular number. We search for ways to reduce sensitivities. Along those lines, we have found that having many thresholds for making small shifts in allocations is best. It reduces the sensitivity to the exact threshold levels and stock allocations. It supports the more general common sense notion of taking a little money off the table (in this case, out of stocks) when valuations are high.

POST E. Numbers for Transition Planning

Preserving capital becomes more and more important as one approaches retirement.

I have collected 5-year and 10-year balances for portfolios with 0%, 20%, 50%, 80% and 100% stocks (as represented by the S&P500) with the remainder invested in TIPS with a 2% interest rate.

This information should help those who are close to retirement to assess the risk in their portfolios.

I collected the appropriate regression equations relating final balances to the percentage earnings yield 100E10/P. Today’s earnings yield is around 3.5%.

5-Year Balances

Putting today’s earnings yield into these equations, a $100000 portfolio is likely to grow (or decline) to the following balances.
1) With 0% stocks and 100% TIPS, the balance at year 5 will be $110408.
2) With 20% stocks and 80% TIPS, the balance at year 5 will be between $117000 and $97000. The most likely balance at year 5 will be $107478.
3) With 50% stocks and 50% TIPS, the balance at year 5 will be between $141000 and $61000. The most likely balance at year 5 will be $101360.
4) With 80% stocks and 20% TIPS, the balance at year 5 will be between $183000 and $33000. The most likely balance at year 5 will be $93169.
5) With 100% stocks and 0% TIPS, the balance at year 5 will be between $187000 and $7000. The most likely balance at year 5 will be $86547.

Comparing these numbers with the tables, the worst case balances for 1923-1980 were:
1) With 0% stocks and 100% TIPS, the balance was $110408.
2) With 20% stocks and 80% TIPS, the worst case balance at year 5 was $100796.
3) With 50% stocks and 50% TIPS, the worst case balance at year 5 was $86760.
4) With 80% stocks and 20% TIPS, the worst case balance at year 5 was $73321.
5) With 100% stocks and 0% TIPS, the worst case balance at year 5 was $64764.

10-Year Balances

Putting today’s earnings yield into these equations, a $100000 portfolio is likely to grow (or decline) to the following balances.
1) With 0% stocks and 100% TIPS, the balance at year 10 will be $121899.
2) With 20% stocks and 80% TIPS, the balance at year 10 will be between $142000 and $102000. The most likely balance at year 10 will be $122331.
3) With 50% stocks and 50% TIPS, the balance at year 10 will be between $191000 and $71000. The most likely balance at year 10 will be $120738.
4) With 80% stocks and 20% TIPS, the balance at year 10 will be between $247000 and $37000. The most likely balance at year 10 will be $116583.
5) With 100% stocks and 0% TIPS, the balance at year 10 will be between $313000 and $13000. The most likely balance at year 10 will be $112568.

Comparing these numbers with the tables, the worst case balances for 1923-1980 were:
1) With 0% stocks and 100% TIPS, the balance was $110408.
2) With 20% stocks and 80% TIPS, the worst case balance at year 10 was $111743.
3) With 50% stocks and 50% TIPS, the worst case balance at year 10 was $96368.
4) With 80% stocks and 20% TIPS, the worst case balance at year 10 was $81149.
5) With 100% stocks and 0% TIPS, the worst case balance at year 10 was $71269.

Conclusions

When valuations are very high, as they are today, stocks have a substantial downside risk. Their most likely returns are about the same as having a 100% TIPS portfolio.

Looking very closely at the data, one can argue that [my eyeball estimate of] the downside risk is somewhat exaggerated. [The amount of scatter increases with earnings yield. I have reported confidence limits that do not vary with earnings yield.] This is only a matter of degree. The worst case historical balances for high stock portfolios have been fortunate. Bad luck caused by randomness and poor returns caused by overvaluation have not occurred simultaneously.

Looking at the charts, the critical threshold seems to be an earnings yield just above 6%, which corresponds to a maximum P/E10 = 16 or 17. The median or typical historical value of P/E10 has been around 14.

Translating this into plain language: you should pay attention to preserving capital as you approach retirement unless stocks are cheap. They are not cheap. They are expensive.

Have fun.

John Walter Russell
August 2, 2005