Author Topic: Analysis behind the 7% return  (Read 1196 times)

RogerOS

  • 5 O'Clock Shadow
  • *
  • Posts: 6
Analysis behind the 7% return
« on: September 23, 2019, 11:13:12 AM »
Hi all,

Recently I downloaded the S&P 500-index prices from 1950-today (https://finance.yahoo.com/quote/%5EGSPC/history/) to do some analyses. I began with trying to replicate the numbers presented in the MMM blog post 'Dude, where’s my 7% Investment Return?', taken from chapter 14 of 'A random walk down wallstreet'.

I started with all 1-year holding returns: the percentage gained/lost by buying the index at a 'Close' price and selling 1 year later. The most extreme values resemble those presented in the blog post: -44.76% loss (buy 2008-02-01, sell 2009-02-01) and 52.94% gain (buy 1982-06-01, sell 1983-06-01).

However, when I calculate the 5-year holding returns the numbers start to divert: -35.8% total or -7.16% annual loss (buy 2004-02-01, sell 2009-02-01 ), 219.91% total or 43.98% annual gain (buy 1994-12-01, sell 1999-12-01).

Am I misinterpreting the original analysis? Is the difference in me not accounting for inflation or dividends? Has anyone else tried to reproduce these numbers? 'A random walk down wallstreet' speaks of 'common stock' values rather than the S&P500-index, but I can't imagine the values to differ this much.

fattest_foot

  • Pencil Stache
  • ****
  • Posts: 789
Re: Analysis behind the 7% return
« Reply #1 on: September 23, 2019, 03:13:01 PM »
I can't quite figure out what you're doing.

CAGR of the Stock Market: Annualized Returns of the S&P 500

hadabeardonce

  • Bristles
  • ***
  • Posts: 321
  • It's never too early to learn the value of money.
Re: Analysis behind the 7% return
« Reply #2 on: September 23, 2019, 03:37:50 PM »
Last year I had some fun with S&P 500 analysis. The attached spreadsheet can be set up for any time period using "find and replace", but right now it's using data from 03/1990 until 08/2019. I think I wanted to check out returns over multiple overlapping timeframes.... it's been a while since I've looked at it.

Odds of a 6% or greater CAGR have looked good(90%+) 20yr+ period of time.
Odds of a 3.5% or less CAGR have looked good(0%+) 20yr+ period of time.

#StaytheCourse

EliteZags

  • Stubble
  • **
  • Posts: 123
  • Location: Newport Beach, CA
Re: Analysis behind the 7% return
« Reply #3 on: September 23, 2019, 03:38:29 PM »
I can't quite figure out what you're doing.



or more importantly: why

nereo

  • Senior Mustachian
  • ********
  • Posts: 11895
  • Location: Just south of Canada
    • Here's how you can support science today:
Re: Analysis behind the 7% return
« Reply #4 on: September 24, 2019, 05:36:50 AM »

Am I misinterpreting the original analysis? Is the difference in me not accounting for inflation or dividends? Has anyone else tried to reproduce these numbers? 'A random walk down wallstreet' speaks of 'common stock' values rather than the S&P500-index, but I can't imagine the values to differ this much.

Yes - if you do not account for inflation or dividends, the returns you get for any period will be incorrect.  The longer the period, the more off your values will be.  Five years is a considerable amount of time to ignore both inflation and dividends.

You may wish to experiment with calculators that can give you returns of the SP500 and adjust for inflation and dividends.  Here is one that I use:
https://dqydj.com/sp-500-return-calculator/


Also - welcome to the forums.

jeroly

  • Bristles
  • ***
  • Posts: 276
Re: Analysis behind the 7% return
« Reply #5 on: September 24, 2019, 05:45:21 AM »
If you're getting the one year returns exactly right, then you're doing the dividend thing right. Are you perhaps taking an arithmetic average of the returns instead of a geometric average?

For example, up 50% and down 50% is not a (50-50)/2=0% annual return but rather (1-(1.5*.5)^0.5)= -13.4% annual return.

RogerOS

  • 5 O'Clock Shadow
  • *
  • Posts: 6
Re: Analysis behind the 7% return
« Reply #6 on: September 25, 2019, 02:05:12 PM »
Thanks for all the useful replies and the welcome! :)

I will retry in the weekend with the suggestions made and let you know. I thought it would be cool not only to know the min and max annualized return over a certain time period (as presented in the blog post), but also plot the actual distribution of these returns.

nereo

  • Senior Mustachian
  • ********
  • Posts: 11895
  • Location: Just south of Canada
    • Here's how you can support science today:
Re: Analysis behind the 7% return
« Reply #7 on: September 25, 2019, 02:37:39 PM »
Thanks for all the useful replies and the welcome! :)

I will retry in the weekend with the suggestions made and let you know. I thought it would be cool not only to know the min and max annualized return over a certain time period (as presented in the blog post), but also plot the actual distribution of these returns.

www.cfiresim.com might be more to your liking.