Rule-based investing, in plain sight

Tactical asset allocation usually lives behind closed doors at hedge funds and institutional quant desks — built on proprietary data, refit quarterly, and rarely shown end-to-end to the people whose money is being moved. Most retail investors are left with two options: buy & hold and ride out 50% drawdowns, or pay an advisor whose process they can't see.

Taunusstar publishes one specific alternative: a transparent, deterministic rule engine that allocates daily between SPY (US equity), LQD (investment-grade corporate bonds), and gold. Six independent stress signals decide whether the portfolio sits in its normal 50/30/20 allocation or flips to a defensive 10/60/30. A drift-only rebalance rule keeps trading costs low. The whole engine is open and reproducible from public FRED and yfinance data.

We commit to transparency about what works and what didn't. The allocations were picked via a 14,400-variant grid search and validated across seven walk-forward folds (out-of-sample, 2007–2026). We show the full crisis decomposition — including the 2022 bear, where the strategy still lost money — and we don't cherry-pick the in-sample optimum. There's no model to retrain, no calibrator to maintain, no proprietary feature set. Just rules you can audit yourself.