📊 Interesting & Useful Charts
Visual insights drawn from decades of market data — designed to illuminate key patterns in investing, long-run returns, and the mathematics of compounding.
v1.0 · 2026-07-13 11:00
S&P 500 — 10-Year Rolling Returns (1981–2025)
Average compound annual growth rate (CAGR) for the previous 10 years, total return including dividends reinvested. For example, the bar for 1999 (18.6%) shows the annualised return of investing in the S&P 500 from January 1989 to December 1999. Even the worst 10-year window since 1981 ended barely negative.
★ Data: S&P 500 Total Return Index. Numbers match the widely-cited rolling return analysis from S&P Dow Jones Indices / Bloomberg. Each bar covers a full calendar-year 10-year window ending in December of the year shown.
The Power of Dividend Reinvestment
The same rolling 10-year CAGR analysis, comparing total return (dividends reinvested) against price return only (dividends taken as cash or not reinvested). The shaded gap between the two lines is entirely attributable to dividend reinvestment. Computed using S&P 500 data from the Clive Thompson Portfolio Simulator.
At the average 10-year rolling CAGR rates above, a single $10,000 investment compounds to $27,100 with dividends reinvested versus $21,100 without — a difference of $6,000 after just 10 years from doing nothing extra. Extend to 20 years and those same rates grow $10,000 to $73,700 with dividends versus $44,900 without — a gap of nearly $29,000, created entirely by reinvesting income you were already entitled to. The lesson: the longer the horizon, the more devastating it is to take dividends as cash rather than putting them straight back to work.
★ "With dividends" series uses S&P 500 total return data from the Portfolio Simulator (source: S&P Dow Jones Indices). "Without dividends" (price return) is estimated by deducting historical average S&P 500 dividend yields for each rolling 10-year window, sourced from Shiller/Yale & S&P Dow Jones Indices. Figures are illustrative; individual periods will differ from averages.