Claim
A monthly rotation between S&P 500 and gold, driven by the S&P/gold price ratio vs its 20-month SMA, achieved (1985–Jul 2020) CAGR 9.9% with max drawdown 22%, vs S&P 500 buy-and-hold max DD 55% — and beat both buy-and-holds on the equity chart (≈2.35M vs 1.4M SPX vs 0.55M gold on 100k from Jul 1986).
- Source: quantified-strategies backtest, rules fully disclosed, presented in
gold-spx-rs-ratio-rotation. No transaction costs stated; dividends explicitly
excluded (
^GSPCprice index). - Why it might be true: relative-strength rotation between low-correlation assets is a well-documented family; the 2001–09 gold regime alone can flip the DD comparison.
- Why to be skeptical: dividend exclusion inflates the strategy’s relative CAGR by ~2%/yr of SPX exposure time; single tested parameter (20-month SMA); edge appears only post-2008; backtest window cherry-ended Jul 2020.
- Test: gold-spx-ratio-rotation-replication — direct replication (rules are fully specified), price-only and total-return variants, cost sensitivity, SMA-length grid.