@quantpad: How to test if your trading entry works: five day-trading entries against no signal on 15 years of E-mini S&P 500 futures. A viewer asked us to show what actually works. This is the test an entry has to pass first. For every trade a signal took, we took the same trade, same direction, same minute and same hold time, on every other day of that year with no signal at all. Whatever the entry earns above that is its edge. If it cannot beat that, the market's drift did the work. Five entries we have tested before: the 15-minute opening range breakout, the initial balance breakout, the Williams breakout, the RSI 85-90 fade and Linda Raschke's Holy Grail, each held 15 minutes, 1 hour and to the close. ES minute bars 2011 to 2025. The opening range breakout held 15 minutes did worse than no signal: -0.11 points a trade on 3,843 trades, ahead in only 4 of 15 years. Of all 15 entry and hold tests, only one cleared two standard errors: the Holy Grail held 1 hour, +1.56 points over no signal, on just 185 trades. Then we walked it forward: expanding train, annual refit, 1-session purge, keeping only cells with t above 2. In 9 of 13 years nothing passed. In sample the picks beat no signal by 0.82 points; out of sample by 0.02. 716 trades, -$10,153 after $16.50 a round trip, -$21,967 at $33. The Williams breakout held to the close made +$54,065 after costs, but its edge over no signal was +0.73 with t of 1.7, and buy and hold made +$281,408. Fifteen years of ES minute bars are included with QuantPad, no data vendor account needed. #QuantPad #backtesting #algotrading #futurestrading #tradingstrategy

439

2026-10-03 17:02:56

To see more videos from user @quantpad, please go to the Tikwm homepage.