Author Topic: Buffett's 10K -> 51M S&P calculation seems wrong  (Read 3851 times)

Graham Osborn

  • Sr. Member
  • ****
  • Posts: 350
Re: Buffett's 10K -> 51M S&P calculation seems wrong
« Reply #10 on: May 15, 2018, 06:23:03 PM »
Wow, thanks Dynamic.  I'm overwhelmed.  You certainly provided ample evidences there that dividends have a huge impact on TR in a range of situations.  And as noted, dividend yields were much higher in earlier times.


  • Hero Member
  • *****
  • Posts: 1549
Re: Buffett's 10K -> 51M S&P calculation seems wrong
« Reply #11 on: May 15, 2018, 09:14:45 PM »
Have fun with this attachment guys

What was very telling when I analyzed the attached data points is that you can have 10 year CAGRs of -1.4% and -1% in the year ending in 2008/2009 in you own the S&P 500 index with dividend reinvested.  The other 10 year CAGR that was negative was the years ending 1938 and 1939 with -2 and -1% 10 year CAGRs.  My key takeaway from investing in the S&P 500 index is that when you go out 15 years, the worst CAGR is -0.2% over 15 years and there are some low single digit 15 year CAGRs like 2% in year ending in 1944, 0% in 1943, 4% in 1974, 5% in 1978, between 4 and 5% for the years ending in 2011 and 2015.  This was a bit surprising given the preaching of long term holds.  I think 10 and 15 years are pretty long and yet the S&P 500 index can produce negative and low single digit returns.

What's my take away from this.  I think the S&P 500 is not some perfect solution for everyone.  In the real world, people have to pay to eat, the IRS, and send their kids to college.  So you can't reinvest every single dividend.  With the personal expenses and inflation, if you generated -2 to 5% CAGR over a 10 or 15 year period, I think you became a lot poorer.   

The other take away is that when the S&P hits a hot decade, it can have 10 year CAGRs of 18-19% like the years ending in the 1997-2000.  That blows my mind a bit. 


  • Hero Member
  • *****
  • Posts: 617
Re: Buffett's 10K -> 51M S&P calculation seems wrong
« Reply #12 on: May 16, 2018, 01:51:57 AM »
I think a hot decade is probably a good warning sign to expect far less in the future and to discount some of your gains in that decade if you're calculating how much of your portfolio you can 'spend' each year, especially if you hold the S&P500 index. Buffett's Fortune article in the early 2000s suggested at best 6-7% above inflation in the future, I recall.

If you're living from your portfolio as some might in retirement, perhaps after a decade of 18-19% returns with modest inflation it would be sensible to assume you are in for a decade of low digit returns on average (with potential market crash somewhere in the middle). Perhaps taking out 3-5 years' income and putting it in cash (even if you miss out on some further gains) would be sensible to ride out the likely crash without needing to sell stock, and potentially even to reinvest much of that cash in the event of a crash.

Certainly, you should not assume you can live on an annual income of 18-19% of your portfolio less inflation rate and not see the value of your portfolio dwindle in real terms. Probably 3-4% of the portfolio is nearer the limit in normal times, and maybe 2-3% after such a bull run.