The universe was chosen after the race

Any study whose asset list contains known ex-post winners carries hindsight selection and survivorship bias by construction. Read counterfactual DCA results as expectation-setting against a broad benchmark, never as evidence for a selection rule.

Fireworks with a footnote

An income-scaled DCA counterfactual produces spectacular headlines: NVDA at a 70.98% XIRR on the fair window, TSLA topping the own-inception view. The footnote carries the real content: the 26-asset universe was configured today, with full knowledge of how the last quarter-century went. Every stock on the list survived; several are there precisely because they won. The delisted, diluted, and forgotten never got a row in the table.

The honest row

That is why the SPY row matters most. The same growing contribution schedule — 317 monthly buys over 26 years — compounds to $4.63M at a 12.25% money-weighted return, with a -52.9% account drawdown along the way. That is what the rule delivers without asset-picking skill. Most of the universe’s “steady” names cluster far below the superstars, and long-duration Treasuries (TLT) manage a negative XIRR on the fair window.

The discipline

Use hindsight universes to calibrate ranges and drawdown experience, never to validate selection: the right reading is “had one held NVDA throughout, this is what the cash-flow rule would have felt like” — not “rules like this find NVDAs”. If a counterfactual’s conclusion survives replacing its winners with the broad benchmark, it is a finding; if not, it is a story about the winners.