AMD 10% dip-buy (60-session hold) on AMD — Published Backtest & Verdict

A 60.9% win rate hid the real score. The rule lost 817 points to holding, took the same maximum drawdown (−65.9% vs −65.4%) while sitting in cash 28.3% of the time, and its 20-session signals were worse than a random day. The slogan survives because the stock multiplied ~25x over the full window — almost any repeatable way of buying AMD made money; this one subtracted value versus doing nothing. We publish losers too.

• Total return: +1369.5% (trailed buy-and-hold at +2186.2%)
• CAGR: +42.1% vs +50.5% buy-and-hold
• Max drawdown: -65.9% vs -65.4% buy-and-hold
• Win rate: 60.9% over 23 trades
• Period: 2019-03-07 to 2026-10-30 (AMD), published 2026-11-04

Cost assumption: slippage and commissions aren't modeled — fills are assumed at the exact closing price with no trading fees, the same assumption stated on How We Backtest. Real execution will differ, usually for the worse.

The claim on trial

Buying AMD after a 10% drop from its trailing 20-session high beats just holding it.

The exact rules

• Drawdown at bar i = close / max close of the trailing 20 sessions (incl. current bar) − 1. Warmup 19 bars.
• Signal: first session per drawdown episode where drawdown ≤ −10% (previous session > −10%). 92 signals, 2019-03-07 → 2026-10-14.
• Claim test: forward 20/60-session returns after every signal vs the all-days baseline over the same history; split by signal-day depth.
• Tradable version: buy at the signal close, hold exactly 60 sessions, exit at that close; signals while invested ignored; after an exit the next signal must be a fresh episode crossing. Entry starts earning the next bar (no look-ahead); exit retains its bar's return.
• Long only, cash earns 0%. No transaction costs, no taxes modeled.

Our honest verdict

A 60.9% win rate hid the real score: the rule lost 817 points to buy-and-hold (+1,369.5% vs +2,186.2%), took the same max drawdown (-65.9% vs -65.4%) at 71.7% time in market, and its 20-session signals were worse than a random day (median -0.37% vs +1.95%). At 60 sessions signals matched any day (+14.6% vs +14.0% mean) — AMD's own rise did the work. Only 1 of 92 first-crossing signals printed at 20%+ depth, so the test cannot grade deep-dip buying.

Full trade log on Substack · Original X thread · StratIQ track record · All published backtests

Charted on TradingView (affiliate link — we may earn a commission at no extra cost to you): tradingview.com

Not financial advice — research and education only; do your own due diligence.