E10/P stuff

Research on Safe Withdrawal Rates

Moderator: hocus2004

Post Reply
User avatar
gummy
*** Veteran
Posts: 107
Joined: Wed May 14, 2003 4:00 am
Location: Hogsville, ON
Contact:

E10/P stuff

Post by gummy »

It occurs to me that, although Schiller's E10/P has a high correlation with the HSWR withdrawal rates, the Schiller 10-year average is an ordinary, garden variety average over 120 months.

That gives just as much weight to the monthly earnings 10 years ago as it does to the current month.

If one uses an exponential average (which gives more weight to recent months than to earnings many months in the past.)
... and one uses a "multipler" = 0.998, one gets a better correlation.

Here's a pretty picture:
JWR1945
***** Legend
Posts: 1697
Joined: Tue Nov 26, 2002 3:59 am
Location: Crestview, Florida

Post by JWR1945 »

I like what I just read.

If you can come up with replacement data, I can enter them on the P/E10 line in our calculators. Then we would be able to use them with switching.

Our calculators use the January P/E10 values for each year. This includes dummy values for 1871-1880. IIRC, there are no values assigned for 2003-2010. If so, Excel converts them to zeroes.

This is an exciting period. We are gaining capabilities fast. It will take a while for us to catch up with what we now have available.

Have fun.

John R.
User avatar
gummy
*** Veteran
Posts: 107
Joined: Wed May 14, 2003 4:00 am
Location: Hogsville, ON
Contact:

Post by gummy »

To get the exponential moving average numbers, it's probably easiest to download the spreadsheet, as described here:
http://www.gummy-stuff.org/E10-P.htm

There are lots of comments to explain most columns and how they're generated.

The E10/P values (using the Exponential Moving Average of monthly earnings) are in column BF (1871 to 2000).

I used Schiller's monthy earnings data (from 1871) to calculate the EMA each month, then used the January EMA value in the spreadsheet.

The spreadsheet weighs in at about 153K.

P.S.
The withdrawal rate (increasing with inflation) that would have reduced your portfolio to $0 (after 30 years) is also given, using the magic formula given here:
http://www.gummy-stuff.org/magic_sum.htm#WITHDRAWALS
MacDuff
* Rookie
Posts: 31
Joined: Wed Jun 18, 2003 8:20 pm
Location: Centralia, WA

Post by MacDuff »

gummy wrote:I used Schiller's monthy earnings data (from 1871) to calculate the EMA each month, then used the January EMA value in the spreadsheet.
Then, if I understand correctly, you are using real or indexed earnings?

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

Post by JWR1945 »

MacDuff wrote:
gummy wrote: I used Schiller's monthly earnings data (from 1871) to calculate the EMA each month, then used the January EMA value in the spreadsheet.
Then, if I understand correctly, you are using real or indexed earnings?

Mac
Professor Shiller lists both nominal and real earnings.

Because he averages ten years of earnings, he uses real dollars only when calculating P/E10. That is, Dr. Shiller uses the real price (or index level) and the average of ten years of real earnings.

Most likely, gummy uses the real price and real earnings in his calculations as well.

Have fun.

John R.
User avatar
gummy
*** Veteran
Posts: 107
Joined: Wed May 14, 2003 4:00 am
Location: Hogsville, ON
Contact:

Post by gummy »

Yes, I used the CPI-adjusted values (Schiller's price/CPI and earnings/CPI ) in calculating the EMA.

Since there are a jillion "weighted averages" (at least) that one may use, one-of-these-days I might try 'em all to see which has the highest correlation with historical "Safe" Withdrawal Rates :D
Post Reply