@quantpad: Where to put your stop when you hold S&P 500 futures overnight: a stop-loss backtest on every ES night since 2011 YouTube asks whether you should hold futures overnight at all. If you do, you need a stop, so how far away does it go? I bought one ES contract at the 4 pm close and sold it at the 9:30 open on all 3,885 valid nights from 2011 to Nov 2026, with a sell stop at a quarter, a half, three quarters, 1, 1.5, 2, 3 and 4 times the previous night's range, and with no stop. Costs $16.50 a round turn, plus one tick of slippage on every stop fill. No stop made +$138,335 after costs. The tightest stop (a quarter of last night's range) was hit on 65.8% of nights and lost -$52,674. Net rose with every step wider, and on the S&P none of them beat no stop. A stop at 1x last night's range cut the worst drawdown from $51,954 to $46,553 (about 10%) and cost $38,046 in profit. Walked forward (expanding by year, 2014 to 2026, each year the stop with the best net per drawdown on past years): ES +$95,134 vs +$136,446 with no stop; Nasdaq NQ +$234,791 vs +$263,144. On NQ a 1x stop did edge out no stop in-sample (+$273,714 vs +$266,953), but picking it from the past still lost. The worst ES night, Aug 5 2024, lost 217 points ($10,854 on one contract); the 1x stop halved it. On Apr 7 2025 the same stop lost 228 points where holding lost 167. Overnight, the gap is your real stop. Takeaway: if you hold a night, size for the worst night rather than leaning on a stop. Not modelled: overnight margin changes and partial fills. I built this in QuantPad. It has every trade on the CME since 2010. Link in bio. #QuantPad #stoploss #futurestrading #backtesting #riskmanagement

QuantPad

Open In TikTok:

Region: US

Thursday 12 November 2026 17:58:31 GMT