Alpha Often Lives in Time-of-Day, Not Pattern

Markets have schedules: session opens, closes, fixes, rollovers, and auctions.
Liquidity, spread, and participation change with these schedules, so behavior changes too.

If you don’t tag time-of-day, you misattribute cause. You think you found a pattern, but you actually found “London open” or “illiquid lunch”.
When you trade the same pattern at a different time, it fails and you call it randomness.

Time is not a detail. It is a regime variable.

Model risk: small parameter changes that flip your results are a warning, not a feature. Robustness is a survival requirement.

• Look for wide plateaus, not sharp peaks.
• Measure drawdown shape, not just final equity.
• Assume the future will be different in the exact way that hurts.

Backtest sees pattern → trade it at wrong time window → spreads/liquidity differ → fills worsen → stopouts cluster → strategy declared ‘broken’.

How it kills accounts:

1. Model works on clean history.
2. Regime changes and execution friction increases.
3. Performance decays slowly, so you rationalize.
4. You optimize parameters instead of reducing risk.
5. Drawdown becomes the teacher.

Tag every trade by session and liquidity conditions.
Treat time-of-day as a filter: some setups are only valid in certain windows.
Assume thin-hour results don’t generalise.

Rule that survives:

• Stress test tails and execution, not averages.
• Prefer robust plateaus over optimized peaks.
• When performance decays, reduce risk before “fixing” the model.

You build a breakout rule that ‘works’ in your backtest. Live, you try it during a thin session and it fails repeatedly. The rule wasn’t universal. It was time-bound.

Tell: if small parameter tweaks flip your results, your model is fragile.

What to do instead

Start with the obvious: The open is price discovery, not a pattern factory. Opens and closes behave differently because participation changes.

Then apply Illiquid hours create fake confidence: the market can look “clean” when it’s just thin.

Related

Glossary: liquidity, regime.

Field checklist

• Stress-test the tails. The worst days define survival.
• Use variable spreads and slippage in testing.
• Prefer robust plateaus over optimized peaks.
• Look at drawdown shape, not only profit.
• If a tiny parameter change breaks the model, the model is fragile.