TAX OPTIMIZATION · INTERACTIVE MODEL

Start one month later.
Own a different tax problem.

Twelve accounts start one month apart and run independently for 252 sessions. Each adjacent pair is compared on the same rebalance date. Signals, prices, and rules match. Tax lots and wash-sale histories do not.

median overlap in usable loss-lot names between adjacent starts on their latest shared rebalance date. Lower overlap reveals greater tax-state dependence.
1 · Blank accountNo holdings, lots, or wash history.
2 · First buys differA later start sees different signals and prices.
3 · Lots divergeCost bases and holding ages evolve separately.
4 · Tax choices differDifferent names carry usable losses or wash restrictions.
5 · Trades compoundEach tax decision changes the next portfolio state.

Move the actual inception month

The comparison automatically uses the next start and holds the proof date constant; December compares with November.

adjacent blank-account starts
JanFebMarAprMayJunJulAugSepOctNovDec

On the same date, they do not own the same tax opportunity set

Usable means a held loss lot still has negative effective realization cost after the optimizer’s age, wash-sale, and harvest hurdles.

Direct tax-state evidence
Selected start only
Shared names
Adjacent start only
Selected wash-restricted names
Adjacent wash-restricted names
Selected loss names harvested now
Adjacent loss names harvested now

After-tax path by account age

Selected start versus its adjacent start; all twelve cohorts form the gray range.

SelectedAdjacentall-start range

Same-date overlap through shared history

Every dot compares the two accounts on the exact same rebalance date. Lower means their inherited tax states differ.

100% can mean both sets are empty early in life; the locked optimizer suppresses very young loss lots.

One-year outcome by inception month

This is context, not the proof: return dispersion combines different market windows with stateful tax execution.

Selected after-tax return
Adjacent after-tax return
Difference between the two
Selected after-tax max drawdown