A switching rule realises its gains every time it switches, and with a few switches a year almost every gain is short-term, taxed as income. Buy-and-hold never realises anything until you sell, so it compounds untaxed. At top US federal rates, the original study found tax cost the switching rule about 10.7 points a year and buy-and-hold only 0.7.
The obvious fix, trading less often, made things worse. Slowing the rule down or checking it monthly lost more in missed crash protection than it saved in tax.
What helped was keeping the signal and changing which exits you take. If the signal says leave TQQQ while the position is sitting on a short-term gain, hold until the gain turns long-term, unless the drawdown gate has shut, which always overrides. Losses are never deferred. In the original study that raised after-tax returns from 25.9% to 32.5% a year at top rates, though some of that came from simply holding TQQQ more often. Matched for risk, the gain was about 2.2 points a year, and it only paid at the 32% bracket and above.
If you can, hold this in a tax-sheltered account and follow the main signal. Members in a taxable account at a high bracket can follow the taxable-account track on the dashboard. This isn't tax advice; your situation will differ.