A one-ETF-at-a-time, RSI-driven regime switch between a pool of leveraged bulls and leveraged bears, selecting the weakest performer within the chosen pool on a 4-day lookback, with 90% exposure and 10% cash, and no automatic daily rebalancing.

In plain terms: the model looks at two markets—bond-like assets (SCHZ) and a very aggressive stock market proxy (URTY). It measures recent price momentum (RSI) for each over about 10 days. If bonds look stronger (SCHZ RSI is higher), the system picks the single worst-performing bull ETF in a small set (e.g., TECL, TARK, URTY, TMF, YINN, EDC, SOXX, LABU, HIBL) over the last few days and buys it with 90% of capital, keeping 10% in cash. If stocks look stronger (URTY RSI higher or SCHZ not higher), it instead picks the single worst-performing bear ETF in a different set (PSQ, SARK, TMV, DRV, JDST, LABD, SH) and executes the same 90/10 allocation. The ranking uses a 4-day moving-average return, computed over a 15-day window, and picks the bottom candidate. There’s no ongoing daily rebalancing; the position only changes when the rule fires. The bull and bear ETF pools are composed of highly leveraged funds, so the strategy aims at short-term bets with amplified exposure, not long-term buy-and-hold diversification.

Out-of-sample, this RSI-driven levered-ETF strategy targets higher upside (26.5% vs 24.2% for S&P) with a disciplined 90/10 long-cash stance, using contrarian bets in bull/bear pools. Higher upside, but greater drawdown risk.