Subject: Re: 15 year comparison - BRK vs S&P
Apart from being so endpoint dependent, total return graphs and associated numbers are visually deceptive:
Two equities, A and B, that perform the same except that A had an early upblip, easily leads one to conclude from the charts that A consistently outperforms B because the A curve is visually obviously 'better' (i.e. consistently above) the B curve for a long time.

My preference is to graph rolling returns of A and B on the same plot, using time periods of interest for the rolling return. If 1 year, 3 year, 5 year, 10 year returns are of interest, then make four such plots.
Also, separately plotting the *difference* of these rolling returns gives a clear picture of who over/under performs who over what time periods.

It's a bit more work, but it's your money, so you're free to invest however you wish. Some like chicken entrails.