A cash-first, signal-driven tactic: buy the cheapest of five ETFs if SMH looks oversold (66.7% of cash to the bottom performer); otherwise tilt to short-term Treasuries (66.7%) if VOO looks overbought and above its long-term trend; otherwise stay in cash. No regular rebalancing.

In plain language: The system starts with cash. If the semiconductor market looks oversold (its RSI indicator reading is very low), it surveys five popular stock ETFs (broad market, Nasdaq, tech, semiconductors, and growth). It picks the one that has recently done the worst over the last 10 trading days and puts about two-thirds of your cash into that single ETF. The remaining cash stays idle. If the semiconductor signal isn’t oversold, the system checks if the market is stretched on the Nasdaq/large-cap side by looking at the broad market’s RSI. If that broad market signal says “overbought” (RSI above 70) and the price is above its long-term trend (its 200-day average), the model shifts to a short-term bond ETF (BIL) with about two-thirds of the cash, otherwise it stays in cash. No automatic weekly rebalancing beyond these rules; there’s a 5% tolerance window, but the plan is not to rebalance unless a rule fires. The overall aim is to try to buy cheap-looking or oversold equity exposures when risk is low and move toward safer assets when the market looks overheated.

Cash-first, signal-driven sleeve that complements SPY by buying oversold semis and shifting to short-term Treasuries when overbought. OOS return ~16%, Calmar ~0.83, Sharpe ~0.76; near-market beta, risk-aware diversification.