Page 1 of 1

SP500 + 10y Treasuries since 1900

Posted: Sun Jun 28, 2015 1:33 pm
by Arturo
Hi all,

Does anybody advice me where can i find a study, chart or web analytic tool from were i can get the performance of a SP500 (plus dividends) + 10y Treasuries (plus cupons) portfolio since 1900? lets say 40/60, 50/50 and 60/40.

I would appreciate.

thanks in advance.

Re: SP500 + 10y Treasuries since 1900

Posted: Sun Jun 28, 2015 5:14 pm
by Ad Orientem
That's likely to be a tough one. The S&P 500 wasn't established until after the Second World War.

Re: SP500 + 10y Treasuries since 1900

Posted: Mon Jun 29, 2015 1:58 am
by Arturo
good point :-). ok, lets say something more general: US equities since 1900 :-)

Re: SP500 + 10y Treasuries since 1900

Posted: Mon Jun 29, 2015 8:36 am
by ozzy

Re: SP500 + 10y Treasuries since 1900

Posted: Mon Jun 29, 2015 12:37 pm
by goodasgold
Interesting, but are these long-term returns real or nominal?

Re: SP500 + 10y Treasuries since 1900

Posted: Mon Jun 29, 2015 1:21 pm
by ozzy
Nominal.

Re: SP500 + 10y Treasuries since 1900

Posted: Mon Jun 29, 2015 1:30 pm
by goodasgold
ozzy wrote: Nominal.
So how many percentage points should we subtract to obtain the results in real terms?

Re: SP500 + 10y Treasuries since 1900

Posted: Tue Jun 30, 2015 3:53 am
by Arturo
thank you very much Ozzy. Very interesting.

by the way, i have several doubts. i have found 4 SP500 statistics since 1926:

1. SP500 performance statistic:
https://www.savantcapital.com/uploadedI ... -P-500.png

2. SP500 composite statistic:
http://jlfmi.tumblr.com/image/103220804375

3. SP500 index statistic:
http://stockcharts.com/freecharts/histo ... 00-130.png

4. SP500 Compound Annual Growth Rate: i still couldn't get the chart, but Vanguard link provides the final CAGR from 1926 to 2014 (10,2% for 100% stocks).

So my questions are:

A. which methodology should we use if we want to know the prices? according to first chart, after 1929, the sp500 price recovers in 1935 to the same level than 1928. But on the second chart it recovers in 1965 approximately, while in the third chart in the 50's. But from the point of view of a investor, we should look at CAGR. And if we look at this figures, it recovered in 1934 (similar to performance statistic).

B. the link to Varguard is superB. Is it possible to obtain a year per year CAGR statistic of this stocks/bonds? i mean, from where you can show a person how many years his portfolio would need to recover from a loss year.

thank you for your time!

regards

Re: SP500 + 10y Treasuries since 1900

Posted: Tue Jun 30, 2015 5:10 am
by Arturo
researching on internet today, i have found a very interesting tool. Compound Annual Growth Rate (Annualized Return) calculator for sp500 (source: Robert Shiller data base) since 1871. Obviously, SP500 starts in 1926, but maybe they are using other indexes like Down Jones:

http://www.moneychimp.com/features/market_cagr.htm

From my point of view, and as far as i understand my own portfolio, we should look at the compound rate, not general price level (performance, composite, or wherever), isnt it?

regards

Re: SP500 + 10y Treasuries since 1900

Posted: Thu Jul 02, 2015 3:58 am
by mathjak107
keep in mind those results are only valid for the accumulation side of things.

once you reach the spending down side of things results in retirement will be far different.


the more conservative below 40% equities the higher the failure rates. a failure meaning without taking a pay cut you would have run out of money .

Image

Re: SP500 + 10y Treasuries since 1900

Posted: Thu Jul 02, 2015 3:54 pm
by Arturo
@mathjak107,

thanks for your post. i appreciate it. by the way, is not what i was looking for.

1. i need to know why "performance" and "compound" are the way to analyze data, and not "composite" or "index". i dont understand why using Robert Shiller data you can get that SP500 recovered the same pre crisis level on the 50s or the 60s, while using the compound, it recovered in 1934 aprox.

2. i would need to get the charts (or the excel) from where Vanguard got this: https://personal.vanguard.com/us/insigh ... llocations.

thanks!

Re: SP500 + 10y Treasuries since 1900

Posted: Thu Jul 02, 2015 6:44 pm
by D1984
Arturo wrote: researching on internet today, i have found a very interesting tool. Compound Annual Growth Rate (Annualized Return) calculator for sp500 (source: Robert Shiller data base) since 1871. Obviously, SP500 starts in 1926, but maybe they are using other indexes like Down Jones:

http://www.moneychimp.com/features/market_cagr.htm

From my point of view, and as far as i understand my own portfolio, we should look at the compound rate, not general price level (performance, composite, or wherever), isnt it?

regards
The "modern" Dow Jones Index didn't start until 1918 IIRC. They are using backtested S&P 500 (or more likely S&P 90 or S&P Composite) total return data. The Cowles Commission in the late 1930s did a backtest by hand (it took them over a year) using the known S&P Composite market cap weighting methodolgy in order to get yearly and monthly total returns (including reinvested dividends) from 1871-1938.

Also, as far as when a 100% S&P portfolio recovered from the 1929-32 89% price fall...well, that depends.

1. If you use the Dow Jones index (which is price only; reinvested dividends not included) it took until late 1954. The Dow Jones isn't the bext index to use IMO, though, since it is price only, is not cap weighted or equal weighted but price weighted, and since whoever ran the Dow Jones index kicked IBM out of the index around 1939 or 1940 (no idea why they did this) and IBM went on to become one of the best performing large-cap stocks of the next 25 years so the Dow's performance was lower than what a more broad large cap index (like one that still included IBM) would've been.

2. If you use only the S&P 500 price only, around 1952 IIRC. Stooq has the S&P 500 (interpolated with the Composite for any days before mid-1957 in order to create an uninterrupted series all the way back) going back to the early 1800s if you want an exact date.

3. If you use the S&P 90 or S&P Composite with dividends reinvested, then in early or mid-1944 IIRC.

4. If you just want to know the first time the S&P 500 Composite with dividends reinvested (i.e. total return) hit its mid-1929 value in REAL (inflation or deflation adjusted terms---remember that a dollar in 1936 bought more than a dollar in 1929 because of the significant deflation from 1930-33) then I think it briefly hit it in late 1936 or early 1937 before it crashed again in mid-1937 and didn't hit the same level again in real terms until 1944 or 1945. I think Shiller at Yale has the monthly total return data (in Excel format) in real and nominal terms if you want an exact month.

Re: SP500 + 10y Treasuries since 1900

Posted: Thu Jul 02, 2015 6:50 pm
by mathjak107
D1984 wrote:
Arturo wrote: researching on internet today, i have found a very interesting tool. Compound Annual Growth Rate (Annualized Return) calculator for sp500 (source: Robert Shiller data base) since 1871. Obviously, SP500 starts in 1926, but maybe they are using other indexes like Down Jones:

http://www.moneychimp.com/features/market_cagr.htm

From my point of view, and as far as i understand my own portfolio, we should look at the compound rate, not general price level (performance, composite, or wherever), isnt it?

regards
The "modern" Dow Jones Index didn't start until 1918 IIRC. They are using backtested S&P 500 (or more likely S&P 90 or S&P Composite) total return data. The Cowles Commission in the late 1930s did a backtest by hand (it took them over a year) using the known S&P Composite market cap weighting methodolgy in order to get yearly and monthly total returns (including reinvested dividends) from 1871-1938.

Also, as far as when a 100% S&P portfolio recovered from the 1929-32 89% price fall...well, that depends.

1. If you use the Dow Jones index (which is price only; reinvested dividends not included) it took until late 1954. The Dow Jones isn't the bext index to use IMO, though, since it is price only, is not cap weighted or equal weighted but price weighted, and since whoever ran the Dow Jones index kicked IBM out of the index around 1939 or 1940 (no idea why they did this) and IBM went on to become one of the best performing large-cap stocks of the next 25 years so the Dow's performance was lower than what a more broad large cap index (like one that still included IBM) would've been.

2. If you use only the S&P 500 price only, around 1952 IIRC. Stooq has the S&P 500 (interpolated with the Composite for any days before mid-1957 in order to create an uninterrupted series all the way back) going back to the early 1800s if you want an exact date.

3. If you use the S&P 90 or S&P Composite with dividends reinvested, then in early or mid-1944 IIRC.

4. If you just want to know the first time the S&P 500 Composite with dividends reinvested (i.e. total return) hit its mid-1929 value in REAL (inflation or deflation adjusted terms---remember that a dollar in 1936 bought more than a dollar in 1929 because of the significant deflation from 1930-33) then I think it briefly hit it in late 1936 or early 1937 before it crashed again in mid-1937 and didn't hit the same level again in real terms until 1944 or 1945. I think Shiller at Yale has the monthly total return data (in Excel format) in real and nominal terms if you want an exact month.

the cpi fell 18% during the great depression . that means you needed a recovery 18% less to  have the same purchasing power.

index's did not include dividends . some were 12 to 18% in those times because stocks fell so much .

many great stocks were doing very well but were not admitted in to the index's yet.

when you figure the 2 together you were whole in purchasing power about 5 years later.

Re: SP500 + 10y Treasuries since 1900

Posted: Fri Jul 03, 2015 3:45 am
by Arturo
@mathjak107,

so as far as i understand, if we take into consideration dividends, inflation (or deflation) and "geometric mean" (CAGR), the data from moneychimp is what we have to pay attention, and not the indexes, that doesn't process those variables, right?

once i got this, i just need the CAGR for US bonds (10y minimum) since 1900, including coupons and inflation adjusted. I have found some data:

In here they provide arithmetic and geometric mean (CAGR), not inflation adjusted (they provide at the last column), but with a doubt: they do not provide information about dividends and coupons adjusted.

http://www.jvlassociates.com/files/Avg% ... 202014.pdf

In here, similar issues. They do not provide information about if data is inflation adjusted, and if stocks compound dividends (according to what is written, treasury bonds include coupons, but not bills).

http://pages.stern.nyu.edu/~adamodar/Ne ... retSP.html