S&P500+Govt Long Bonds: Bad news

Research on Safe Withdrawal Rates

Moderator: hocus2004

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

S&P500+Govt Long Bonds: Bad news

Post by JWR1945 »

I copied the 1928-2000 Government Long Bond data from the Gummy's database.
http://www.gummy-stuff.org/returns.htm

Data in Gummy's database are more realistic than the existing data in the historical sequence calculators. The existing data use single-year returns only from interest. Gummy's data include single-year capital gains and losses as well as interest payments. Neither set of data includes transaction costs.

I made a copy of an existing calculator. (I used my latest, my Gummy 02 version of the Deluxe Calculator 1.1A08.) I pasted the Government Long Bond data on top of the commercial paper entries for 1928-2000 (in cells BG177 through EA177).

I pasted on top of the commercial paper data because I was interested in varying stock and bond allocations according to P/E10. My calculators allow switching with commercial paper, TIPS and I-bonds, but not with 5-year Treasury Notes and not with 30-year Treasury Bonds. The Retire Early Safe Withdrawal Calculator, Version 1.61 dated November 7, 2002, from which I made my original modifications, allows switching only with commercial paper.

Conditions

I set the starting balance at $100000. I set expenses to 0.20%. I varied the withdrawal rate. I used the CPI for inflation. I examined 30-year sequences starting in 1928-1980. There are 53 sequences. Stock allocations consisted of the S&P500 index.

I took a brief survey. I varied the stock allocation depending upon P/E10. When P/E10 was below the lower threshold (which varied), the stock allocation was 100%. When P/E10 was between the two thresholds, I used an intermediate allocation of 30% or 50% or 70% as indicated. When P/E10 exceeded the upper threshold, which I set at 21, the stock allocation was 0%.

The best intermediate stock allocation (when there was only one intermediate allocation) from a previous survey using commercial paper was 30%. The best P/E10 thresholds were 12 and 21.

I collected baseline data with fixed stock allocations of 0%, 30%, 50%, 70% and 100%.

Procedure

I increased the withdrawal rate in increments of 0.1%. I recorded the highest rate at which all portfolios from 30-year sequences beginning in 1928-1980 survived. I have listed those rates as HSWR.

I continued increasing withdrawal rates in increments of 0.1%. I recorded the lowest withdrawal rate at which 1 or more, 5 or more and 10 or more portfolios failed.

This method allows me to survey a large number of conditions rapidly. By including data with 5 and 10 failures, I am able to spot difficulties associated with probability distributions.

Results

This was a brief survey. This was not a full optimization. This did not include a full sensitivity study.

The Survey of Thresholds and Allocations

Calculator data: 1928-2000.
30-year sequences from 1928-1980.
$100000, 0.20% expenses.
[Calculator settings:
P/E10 thresholds: 4-varies-21-80.
Allocations: 100-100-varies-0-0.]

P/E10 threshold = 9 and [Intermediate stock] Allocation = 30%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 3.9
Ten failures: 4.0

P/E10 threshold = 9 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.0
Ten failures: 4.5

P/E10 threshold = 9 and Allocation = 70%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.1 in years 1962-1966
Ten failures: 4.7

P/E10 threshold = 12 and Allocation =30%
30-year Failures in 1928-1980:
HSWR: 3.7
First failure: 3.8
Five failures: 4.0
Ten failures: 4.1

P/E10 threshold = 12 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.7
First failure: 3.8
Five failures: 4.1
Ten failures: 4.6

P/E10 threshold = 12 and Allocation = 70%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.1 in years 1962-1966
Ten failures: 4.7

P/E10 threshold = 15 and Allocation = 30%
30-year Failures in 1928-1980:
HSWR: 3.2
First failure: 3.3
Five failures: 3.5 in years
Ten failures: 4.3 in years 1959-1968

P/E10 threshold = 15 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.3
First failure: 3.4
Five failures: 3.7
Ten failures: 4.3

P/E10 threshold = 15 and Allocation = 70%
30-year Failures in 1928-1980:
HSWR: 3.4
First failure: 3.5
Five failures: 3.9
Ten failures: 4.4

Summary of results with the best intermediate stock allocation, which was 50%.

P/E10 threshold = 9 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.0
Ten failures: 4.5

P/E10 threshold = 12 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.7
First failure: 3.8
Five failures: 4.1
Ten failures: 4.6

P/E10 threshold = 15 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.3
First failure: 3.4
Five failures: 3.7
Ten failures: 4.3

Additional conditions with an intermediate stock allocation of 50%.

P/E10 threshold = 10 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.0
Ten failures: 4.5

P/E10 threshold = 11 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.0
Ten failures: 4.5

P/E10 threshold = 12 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.7
First failure: 3.8
Five failures: 4.1
Ten failures: 4.6
This confirms earlier findings.

P/E10 threshold = 13 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.7
First failure: 3.8
Five failures: 4.1
Ten failures: 4.6

P/E10 threshold = 14 and Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.3
First failure: 3.4
Five failures: 3.7
Ten failures: 4.5

Baselines

Calculator data: 1928-2000.
30-year sequences from 1928-1980.
$100000, 0.20% expenses.
[Calculator settings:
Fixed allocations. No switching, but with annual rebalancing.
Stock Allocations: 0%, 30%, 50%, 70%, 100%.


Stock Allocation = 0%
30-year Failures in 1928-1980:
HSWR: 2.3
First failure: 2.4
Five failures: 2.7
Ten failures: 2.8

Stock Allocation = 30%
30-year Failures in 1928-1980:
HSWR: 3.3
First failure: 3.4
Five failures: 3.6
Ten failures: 3.9

Stock Allocation = 50%
30-year Failures in 1928-1980:
HSWR: 3.5
First failure: 3.6
Five failures: 4.0
Ten failures: 4.4

Stock Allocation = 70%
30-year Failures in 1928-1980:
HSWR: 3.6
First failure: 3.7
Five failures: 4.2 in years 1962, 1964-1966, 1968
Ten failures: 4.7

Stock Allocation = 100%
30-year Failures in 1928-1980:
HSWR: 3.7
First failure: 3.8
Five failures: 4.2 in 1929, 1965-1966, 1968-1969
Ten failures: 4.7

Comparisons

Two conditions with switching did better than all of the fixed allocations (of 0%, 30%, 50% and 70%) that included Government Long Bonds. They had lower P/E10 thresholds of 12 and 13. The upper P/E10 threshold was 21. The intermediate stock allocation was 50% (making the stock allocations 100-50-0% when P/E10 fell below, between and above the thresholds).

Failures clustered together. For conditions with 5 failures that I examined, they always occurred in 1962-1966. For the condition with 10 failures, they occurred in 1959-1968.

A 100% stock portfolio did better than all of the conditions that included Government Long Bonds.

Failures with 100% stocks were a little bit different. The 5 failures at a withdrawal rate of 4.2 occurred in 1929, 1965-1966, 1968-1969. That is, the Great Depression year of 1929 produced a failure in addition to the stagflation associated with retirements starting in the 1960s.

Summary

These results remind us of how badly owners of Government Long Bonds were hurt by the inflation and rising interest rates of the late 1960s and the whole decade of the 1970s.

Although switching stock allocations helped a little bit, the best decision would have been to abandon the long-term Treasuries entirely. Based on previous studies, commercial paper would have been a good alternative. But investors were better off abandoning Government Long Bonds even if they went to 100% stocks.

Have fun.

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

Post by Mike »

But investors were better off abandoning Government Long Bonds...
They did, and switched to equity and real estate. However, this helped drive the S&P dividend yield below 2%, making future returns potentially less promising than before.
User avatar
Alec
Admin Board Member
Posts: 31
Joined: Wed Sep 10, 2003 4:00 am
Location: Crofton, MD

Post by Alec »

John,

How about intermediate term bonds or 5 year treasuries? A lot of people think that the choice for bonds is an "either-or" scenario when it can actually be a combination of ST, IT, and LT bonds.

Jaye Jarrett did a study on "The Fixed Income Portion of a Retirement Withdrawal Strategy" in 2001. Unfortunately, his website seems to be down, but i can email you a copy of the pdf if I cannot find another copy on the web.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Alec wrote:John,

How about intermediate term bonds or 5 year treasuries? A lot of people think that the choice for bonds is an "either-or" scenario when it can actually be a combination of ST, IT, and LT bonds....
The good news is that I can now make such combinations as you have mentioned. It requires what I call the Gummy 04 Calculator. It uses Gummy's data as replacements for both the stock data and the commercial paper data for the years of 1928-2000.

Gummy's S&P500 data are slightly more generous than Professor Shiller's. Using comparable conditions, Gummy's data show first failure at 4.1% (and safety at 4.0%) while Professor Shiller's data shows the first failure at 4.0% (and safety at 3.9%).

The replacement data for both stocks and commercial paper during 1928-2000 can consist of any combination of Gummy's investment types.

It is time to be building wish lists for new studies. We now have calculators that can handle a lot of new investment types and their combinations.

IIRC, Jarrett's studies have the same limitations as the earlier historical sequence studies. There were no capital gains and losses from year to year. Everything was a single-year trading vehicle. That is, only the interest rates changed and nothing could ever be locked in. Gummy's data, although still changing from one year to the next, include capital gains and losses.

Have fun.

John R.

P.S. I have made my Gummy 02, Gummy 03 and Gummy 04 calculators available for downloading from our special SWR Research section. The link to the special section is on the left side of the NFB home page.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

These data with the Government Long Bond have the same timing problem that I had with Large Cap Value. I needed to shift Gummy's data one year to the right when I put this together.

I will have to redo this study. Expect the actual results to be better.

Have fun.

John R.
User avatar
Alec
Admin Board Member
Posts: 31
Joined: Wed Sep 10, 2003 4:00 am
Location: Crofton, MD

Post by Alec »

IIRC, Jarrett's studies have the same limitations as the earlier historical sequence studies. There were no capital gains and losses from year to year. Everything was a single-year trading vehicle. That is, only the interest rates changed and nothing could ever be locked in. Gummy's data, although still changing from one year to the next, include capital gains and losses.
Nope. Jarrett's study used the total return data of (1) S&P 500, (2) CRSP 9-10 Index, (3) Salmon Bros Long Term High Grade Corp Bond Index (4) 20 Year Treasuries (5) 5 Year Treasuries (6) 30 Day Tbills (7) CPI-U. His annualized return, standard deviation, and sharpe ratios are almos exacly the same as I get using Gummy's bond return data. Jarrett's data is from 1926-1999, from the Ibbotson's Associates 2000 Yearbook, and Gummy's is from 1928-2000.

- Alec
Post Reply