Maintenance Stage Survey

The maintenance stage occurs after you have accumulated enough funds but before you make withdrawals.

I conducted a brief survey using three approaches: switching allocations in accordance with P/E10, maintaining a fixed 50%-50% allocation of stocks and TIPS and keeping everything in stocks. I made an excursion by introducing Latch and Hold.

Careful analysis shows that varying allocations is best. Latching onto bargains and holding them for several years is better than mindlessly cutting back as soon as prices rise.

Overview

I used a very simple switching algorithm. I used my Version 1.61 Deluxe Calculator V1.1A08b. It is available in my Yahoo Briefcase.

I set the P/E10 thresholds to 10-15-20-25 and the allocations to 100%-75%-50%-25%-0%. I used 2% TIPS. I set expenses equal to 0.00%. I set the initial balance equal to $10000.

I determined the real (inflation adjusted) dollar balances at Years 5, 10, 15, 20, 25 and 30. I generated a series of regression equations relative to the initial 100E10/P (that is, 100/[P/E10]) level. I included all completed sequences of Years 1901 through 2000 using Excel’s LINEST function. I looked at the Years 1921 through 2000 separately. I set the confidence limits at plus and minus 1.64 times the standard deviation. The confidence level is of the order of 90% (by formula). I calculated expected values and upper and lower confidence limits for 100E10/P = 10, 6.7, 5, 4 and 3.3. This corresponds to P/E10 = 10, 15, 20, 25 and 30.

I collected a set of data with a fixed 50%-50% allocation of stocks and 2% TIPS. I collected another set of data for a 100% stock portfolio.

I collect another complete set of data using my Latch-Hold Threshold Calculator LH02. I refer to this as LH Switching. I left the Latching Threshold and Holding Thresholds unchanged from their originally selected values. The thresholds were 24.1 (4 years) and 8 (7 years), preference lower.

I also took a complete set of data for a fixed 50%-50% allocation and a 100% stock allocation on LH02 calculator to prevent hidden errors.

I constructed comparison tables. The most useful tables showed the difference between the 50%-50% portfolio and the other dollar amounts (switching and 100% stocks) and between the 100% stock portfolio and the other dollar amounts (switching and 50%-50%). These tables included the Year, P/E10, 100E10/P and real dollar amount information. I construct them so that the selected portfolio (50%-50% or 100% stocks) would show up in black and the differences would show up in red when the alternatives were superior.

I have placed these tables into my Yahoo Briefcase in the “Maintenance Stage” Folder.

Yahoo Briefcase

Data Analysis

In all cases, the fixed 50%-50% portfolio had the narrowest confidence limits. If you focus on minimum balances and higher valuations and use the standard formulas “as is,” you would choose this portfolio. [For greater detail, look at Sheet 4 of the spreadsheets in the “Maintenance Stage” folder.]

Doing so would be a serious mistake.

On sheet 5 of my spreadsheets, I show side by side comparisons of the actual dollar amounts. I include the Year, P/E10, 100E10/P as references.

To make these data easy to interpret, I constructed similar tables below. One of them shows comparison information with the 50%-50% portfolio. It shows all of the balances of the 50%-50% portfolio, as in the original list. But under the Switching (or LH Switching) column is the difference: I subtracted the Switching amount from the 50%-50% amount. The color is black when the 50%-50% portfolio is better. The color is red (indicating a negative number) when the Switching portfolio is better.

I did the same at the bottom using the 100% stock portfolio as the reference.

By Year 10, the story is clear: switching allocations is vastly superior to a fixed 50%-50% portfolio. We can see this in the Year 5 data as well, just not so dramatically. If we include Latch and Hold, the story is obvious even by Year 5.

The comparison of Switching with 100% stocks requires a little bit more effort, but not much. Looking at the bottom table in Sheet 5, it looks like a toss up initially. Looking more carefully, however, we notice that the 100% stock portfolio is best only when the final balances are high. The Switching (or LH Switching) portfolio is better whenever the final balance is low. [This advantage diminishes at years 25 and 30. All final balances are high.]

As might be expected, the final balances are low when valuations are high.

Interpretation

Varying allocations according to P/E10 is best during portfolio maintenance when there are no deposits or withdrawals. Latch and Hold is a good idea as well.

There was no optimization of the Switching and LH Switching algorithms.

Based on previous experience, it is highly likely that individuals can learn to do even better. Just practice on one of my Retirement Trainers. The latest, which is available as a free download from my Yahoo Briefcase, is the “Simplified Retirement Trainer with Dividends A” or “Simp Ret Tr w Div A2” for short. Just keep the “Total Withdrawn” always equal to zero. Any amount withdrawn from stocks is automatically added to TIPS. Any amount deposited into stocks (a negative withdrawal) is automatically taken from TIPS.

Have fun.

John Walter Russell
December 7, 2006