From Earnings Yield

Research on Safe Withdrawal Rates

Moderator: hocus2004

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

Post by JWR1945 »

Here are the annualized real returns Return0 for 80% stocks and 20% commercial paper (and 0.20% expenses) after 10 years. This covers the years 1871-1920 (50 years).

11.84
9.95
9.74
9.84
8.84
10.27
12.47
10.28
8.61
8.35
5.05
7.84
7.10
6.53
7.05
4.88
4.49
6.17
7.37
4.95
7.48
6.63
6.55
5.68
7.00
8.36
7.46
4.49
4.89
6.40
5.05
4.24
4.40
4.76
2.26
2.54
2.32
0.93
(1.70)
(2.03)
(3.24)
(1.80)
0.37
1.03
3.41
3.03
4.66
10.75
14.84
14.23

(more follows)

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

Post by Mike »

Thank you John. When you copy data from the web, you can make Excel drop the HTML coding by right clicking and choosing special paste. It asks what type of coding you want to use when inserting the data into cells.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Here are the annualized real returns Return0 for 80% stocks and 20% commercial paper (and 0.20% expenses) after 10 years. This covers the years 1921-1980 (60 years).

13.93
8.94
7.09
10.48
7.46
9.45
10.46
4.73
2.62
3.59
3.81
4.69
4.95
2.95
5.09
4.10
(0.62)
1.35
0.74
1.88
4.31
7.25
7.64
6.33
8.42
8.02
11.15
11.12
13.30
12.26
10.91
11.22
9.78
11.18
9.01
7.66
6.56
8.33
6.10
4.38
4.13
3.51
4.91
1.42
(2.36)
(1.10)
0.05
(2.00)
(2.03)
(0.86)
(0.02)
(1.79)
(1.08 )
2.13
5.27
4.77
6.55
7.36
7.93
9.18

Have fun.

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Here are the annualized real returns Return0 for 100% stocks (and 0.20% expenses) after 10 years. This covers the years 1871-1920 (50 years).

12.47
10.34
10.04
10.10
8.73
10.57
13.33
11.17
9.25
8.37
4.47
7.60
6.84
5.98
6.86
4.30
3.80
5.86
7.52
5.05
8.13
7.28
7.29
6.44
8.20
9.87
8.85
4.85
5.36
7.08
5.29
4.22
4.35
5.01
1.91
2.34
2.31
1.11
(1.83)
(2.15)
(3.65)
(2.12)
0.38
1.25
4.17
3.71
5.40
12.59
17.19
16.04

(more follows)

John R.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

Here are the annualized real returns Return0 for 100% stocks (and 0.20% expenses) after 10 years. This covers the years 1921-1980 (60 years).

15.60
9.04
6.60
10.73
7.14
9.53
10.94
3.87
1.47
2.76
3.37
5.40
6.19
3.82
6.48
5.31
(0.20)
2.62
1.90
3.28
6.44
9.94
10.22
8.50
10.98
10.39
13.95
13.70
16.35
15.05
13.20
13.48
11.65
13.36
10.69
9.00
7.57
9.72
6.95
4.77
4.48
3.75
5.45
1.05
(3.52)
(1.93)
(0.43)
(2.96)
(2.95)
(1.34)
(0.42)
(2.76)
(1.93)
2.01
5.82
5.12
7.19
8.16
8.79
10.22

Have fun.

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

Post by Mike »

I am not sure that individual columns will transfer directly to Excel (or Word). If nothing else, they should be easier to cut and paste.
They did. I right clicked on an Excel cell, chose paste in unicode for formatting, and was able to make charts with the resultant data with no further modifications. Thanks again.
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

This is the impact of our adjustments to the number of degrees of freedom.

Instead of confidence limits corresponding to plus and minus 1.64 standard deviations, the 90% confidence limits are widened by 10% to 40% (actually, by the square root of 58/48 to the square root of 2). The existing confidence limits are actually at plus and minus 1.64/1.1 = 1.49 to 1.64/1.4 = 1.17 standard deviations. Instead of changing the existing confidence limits, we will change the confidence levels associated with them. The new confidence levels are 86% to 75% instead of 90%.

It is important to recognize that the larger adjustment (to a 75% confidence level) is an extreme upper bound. It assumes that Historical Database Rates vary much more than they do. It is worth having this bound since is assures us that the existing confidence limits will still be meaningful even if future calculations widen the 90% confidence limits further.

By focusing on earnings yield and the effects of valuations, we have separated out the effects of year-to-year price fluctuations. We have been able to capture a large fraction of the degrees of freedom buried within the historical record. This is in stark contrast to previous historical-sequence methodologies, which used only a small fraction of the information available.

Have fun.

John R.
Post Reply