A data-driven critique of dollar cost averaging (DCA) as an investment strategy. Using S&P 500 data going back to 1870, the analysis computes internal rate of return for DCA vs lump sum investments over five-year horizons. Results show DCA provides only ~25 basis points of additional annual return but is actually more likely to be in the red after five years (12.9% vs 11.0%). None of the differences are statistically significant, debunking popular claims that DCA 'minimizes downside risk' or 'drastically reduces market risk'. The post also covers how to compute IRR using binary search, noting that numpy's built-in implementation is slow.
Sort: