Does the Opening-Range Breakout Survive an Honest Backtest? Evidence from the S&P 500

This paper tests a five-minute opening-range breakout (ORB) strategy on S&P 500 stocks from 2024 to 2025. Four “judges” decide which stocks may trade: the index must be above its 200-day average, the stock's RSI must show a persistent uptrend, the stock must outperform the index over 60 sessions, and its sector ETF must be at a 20-day high. Every trade uses a stop of 1.5 times the average true range, risks at most 1% of equity, and closes at the end of the day. During development the strategy's reported Sharpe ratio moved from 10.56 to 2.19 without any change to its rules, as implementation errors were found and corrected. The most damaging was a mismatch in price adjustment between intraday and daily data vendors, which created artificial profits on every short trade and single-day windfalls around stock splits. With all pricing errors corrected, the strategy takes 191 trades and loses 7.7%, with a Sharpe ratio of −1.06, over a period in which the S&P 500 rose more than 40%. Short trades, which bet against stocks the judges have just rated as strong, lose $10,556 (p < 0.001); long trades earn a statistically insignificant $2,895. A bootstrap of the trades gives a 3.6% probability of any profit. The paper catalogues thirteen implementation biases and argues that pipeline errors, not overfitting, are the first threat to an intraday backtest.