Midday baseline: buy the Nasdaq 100 at noon, sell at 3:55
An ML day trader asked how to separate stock-picking skill from simply trading the midday-to-close window. We tested the window itself: buy the Nasdaq 100 at noon New York every day for five years, sell at 3:55, spread included. On its own it was close to nothing, tuning the timing didn't hold up, but only trading on days the index was in an uptrend held up thinly on six unseen months.
From r/Daytrading. The idea, in our words: The control the post asks for: no stock picking and no model, just buy the Nasdaq 100 at 12:00 New York every day and close at 3:55. If this window lost its edge, a model trading only this window will struggle too.
improved version held upAs posted · the rules as described (see what we assumed below)
Optimized: best settings Quant Core found · last months held back as unseen data
Improved: only trade with the trend (price above/below the 60min 200 EMA)
We then tried 8 changes a quant would test first (440 settings in all) and kept the best on past data only. The held-back months played no part in choosing it. It held up on the unseen months.
Ideas tried for this post
Each idea was tuned on past data only, then run on the held-back months (£ won per £1 lost).
| Idea | Tuning months | Unseen months |
| only on days the index is above its session VWAP at noon (a strong morning) | +13% · £1.18 | +2.4% · £1.43 · 68 trades |
| only on days the index sits in the bottom third of its morning range at noon (buy a weak morning) | +10% · £1.19 | -1.1% · £0.76 · 29 trades |
What we had to assume
- Nasdaq 100 CFD instead of individual stocks or SPY (an index baseline, not their stock picks; we have no S&P 500 data)
- Enter at 12:00 New York (the middle of their 11:30-12:30 window), close at 3:55, every trading day
- A wide 3% safety stop the post doesn't have; position sized at 1% risk to that stop
A different reading of the rules could give a different result.
Backtested with the StratOS optimizer on historical data, with spread and costs charged, and the most recent months held back as unseen data. A backtest is not a forecast, and nothing here is investment advice.