Strict backtest

Signals. 12- and 6-month momentum: price change over 12 or 6 months, skipping the latest month. 3-month momentum: the last 3 months. 1-month reversal: last month's biggest losers.

Rules. At each month-end close, rank stocks on the signal, buy the top 15 at equal weight, hold one month, repeat. Decisions use only prices up to that close. Turnover in the strict run: about 408% of the portfolio a year.

Hyped uses today's S&P 500 list for every past month (so it only ever picks companies that later survived and grew into the index) and pays nothing to trade. Strict uses the index members on each date and charges the cost you choose on every buy and sell.

Survivorship caveat. Some bias remains even in strict mode: former members that were acquired or delisted have no price data in this free dataset, so they can't be picked. Real results would likely be somewhat lower. No taxes, slippage beyond the cost setting, or dividends beyond SPY's total return.

Data: Yahoo Finance daily prices through 2026-10-16, S&P 500 history from public membership records. 129 months tested. Same harness as SURVYR.'s published strict model results.