Selling less can reduce tax costs, but a portfolio can also sell less simply by allowing its riskiest asset to become larger. The September 8, 2026 study separated those effects across eleven fixed rebalancing policies.
The full after-tax review covers January 2, 2019–September 4, 2026. Each strategy began with $100,000, no later cash flows and a 60% VT / 30% GLDM / 10% bitcoin allocation.
Ongoing wealth and liquidation tell different stories
“Ongoing” CAGR deducts taxes incurred during the strategy but leaves remaining unrealized gains untaxed. “Liquidated” CAGR also deducts the modeled tax on selling everything at the end.
| Policy | Ongoing CAGR | Liquidated CAGR | Maximum drawdown | Trade days |
|---|---|---|---|---|
| Annual rebalancing | 20.17% | 17.70% | −29.47% | 7 |
| Buy and hold, annual cash investment | 22.39% | 18.97% | −49.46% | 7 |
| Taxable KISS 5.3 | 16.12% | 13.81% | −22.87% | 12 |
| Taxable KISS plus cash recycling | 16.72% | 14.27% | −23.00% | 15 |
| Annual bitcoin cap, cash recycling, long-term-lot guard | 21.25% | 18.52% | −32.84% | 7 |
The bitcoin-cap candidate finished with approximately $368,127 after liquidation, versus $349,151 for annual rebalancing and $269,694 for Taxable KISS. That single start-date advantage was insufficient to promote it.
What the tax model assumed
The simulation used explicit lots, distributions and common next-session closes. Its constant illustrative federal rates were 40.8% on short-term gains and interest, 23.8% on long-term equity gains and qualified dividends, and 31.8% on long-term gold gains. These are scenario inputs, not a reconstruction of changing historical tax law or a statement of any reader’s applicable rates.
The base case did not credit losses against outside income or gains. An optimistic loss-credit sensitivity was tested separately. State taxes, a complete tax-return calculation and all wash-sale complications were outside scope. Bitcoin used spot daily closes, not a historical bitcoin-ETF record synchronized perfectly with ETF trading hours.
How much was tax efficiency, and how much was exposure?
The annual cap candidate used a 14.29% bitcoin threshold. A breach was checked annually, so this was not a continuous 14.29% risk limit. Average bitcoin exposure was 14.50%, versus 11.96% under annual rebalancing; maximum exposure reached 35.80%, versus 28.14%.
The candidate beat annual rebalancing on liquidated CAGR for only two of five overlapping start-year cohorts. It won for 2019 and 2023 starts, and lost for 2020, 2021 and 2022. Those are not five independent trials or a reliable estimate of a future win probability.
Taxable KISS’s lag was largely the opportunity cost of reduced exposure, rather than frequent short-term taxable gains. Adding a long-term holding guard did not change its result in this sample. The all-asset band and valuation-gated candidates also matched the bitcoin-band result because the added conditions did not create independent decisions.
The research retained these candidates for further observation. It did not establish that a more complicated rebalancing policy should replace the existing portfolio. Compare after-liquidation wealth, drawdown and asset exposure together before attributing an advantage to taxes.