A high-risk, rule-based tactical strategy using leverage and hedges to ride short-term moves, with a semiconductors tilt (SOXL/SOXS) replacing older QQQ bets. It mixes volatility, inverse/short ETFs, bonds, and a TSLA long bias, driven by tight short-window signals and multi-criteria screening to keep total exposure at 100%.

In plain terms, the system starts with cash and then, based on very short-term price signals, chooses whether to place bets that make money if markets fall or rise. It looks at a set of instruments (volatility bets like UVXY, bearish bets on tech like SOXS or QID, bullish/ Bear tilts on semis like SOXL, and occasional stock bets on TSLA) and picks the best one or two using criteria such as how much they’ve moved recently, how strong they’ve been versus other assets, and whether their short-term price trends are up or down. Then it assigns weights so the total exposure adds up to 100% of the capital, favoring hedges when risk is high and favoring long bets (especially TSLA in this design) when conditions look favorable. The strategy explicitly substitutes SOXL/SOXS in place of TQQQ/QID/SQQQ, shifting emphasis to semiconductors rather than broad Nasdaq leverage. The rules are nested: different market states trigger different combos of assets (sometimes hedges, sometimes longs), with occasional long-only or short-only legs, all chosen by ranking criteria (top/bottom by cumulative return or RSI, etc.) and constrained to keep total exposure at 100%. The numbers you see (windows like 7, 11, or 13 days; weights like 48/100 or 88/100) are the specific knobs the model uses to decide which leg to favor in each state. Because it relies on short-horizon signals and leveraged ETFs, it is highly sensitive to regime changes and can produce substantial upside but also serious drawdowns.

Out-of-sample upside: ~60% annualized return vs ~23% for the S&P, aided by a semiconductor tilt and hedges. Calmar ~1.46 indicates solid risk-adjusted upside; max drawdown ~41%, higher than SPY.