Draft — These backtests have not yet been independently verified. Do not trade based on this data.
Home
Study — Premarket All-Time Highs

When SPY hits a new ATH
before the open, it fades—
but less than you think.

Across 243 premarket all-time highs spanning 2013–2026, the morning drawdown is actually 48% smaller than average. But the risk/reward skews 35% in favor of shorts — because the upside is even more suppressed. The edge lives in the filters.

A premarket ATH day is any day where SPY's high between 4:00 AM and 9:29 AM ET exceeds the highest price ever recorded in regular trading hours. This captures the moment when overnight futures or early-morning buying pushes SPY into uncharted territory before the opening bell.

243
Premarket ATH days
out of 6,582 trading days
3.7%
Frequency
~12 days per year in bull markets
2013–2026
Date range
premarket data availability

Premarket ATH days cluster in bull markets: 44 in 2021, 37 in 2017, 32 in 2024. There are none in bear market years. The sample starts in 2013 reflecting when reliable premarket data becomes available.

The thesis says premarket ATH days should produce large morning drawdowns, making them good short entries. The data says the opposite: ATH days have smaller drawdowns than average because they are low-volatility, bullish-momentum days. The entire price range is compressed.

-0.25%
Mean max morning drawdown
Premarket ATH days
-0.48%
Mean max morning drawdown
Non-ATH days (baseline)
48%
Smaller drawdowns
on PM ATH days

Running worst drawdown from the open

Cumulative minimum price vs open, by minutes elapsed (mean across all days)
PM ATH days (n=243) Non-ATH days (n=6,157) PM ATH max run-up

Why smaller? All-time highs happen during strong uptrends with low volatility. The daily ATR is narrow, the trend is firmly bullish, and there's no panic. The range compresses in both directions — less downside, but also less upside from the open.

Here's the non-obvious finding. Although PM ATH days have smaller absolute drawdowns, the risk/reward is skewed in favor of shorts. On non-ATH days, the upside and downside from the open are roughly symmetric (R:R ~1.0). On PM ATH days, the downside consistently exceeds the upside by ~35%.

1.35
Median reward/risk
PM ATH days (all windows)
1.01–1.06
Median reward/risk
Non-ATH days (symmetric)
WindowPM DownPM UpR:RNon DownNon UpR:R

Translation: On PM ATH days, if you short at the open, the average maximum profit exceeds the average maximum loss by 35%. On normal days, they're equal. The edge isn't in the size of the move — it's in the directional skew.

How often does price drop at least X% from the open during the morning? PM ATH days underperform non-ATH days at every threshold. If you need a 25bp drop for your short to work, it happens 39% of mornings on ATH days vs 60% baseline.

Left bar = PM ATH days, right bar = non-ATH days. Morning session (open to 12:00 ET).

The average PM ATH day is a weak short. But specific conditions sharpen the edge. Large ATH extensions, gap-up opens, and Phase Oscillator readings outside distribution/extended_up all produce meaningfully larger drawdowns.

By ATH extension size

ExtensionNDD MorningDD Day

By gap size (prev close → open)

GapNDD Morning≥25bp≥50bp

By Phase Oscillator zone at open

PO ZoneNDD Morning≥25bp≥50bp

By open position in ATR framework

ATR PositionNDD Morning≥25bp≥50bp

Kill condition: When the 10m Phase Oscillator is in distribution or extended_up at the open, don't short. These days have a morning drawdown of just -0.06% to -0.11%. Momentum is carrying through — the ATH is real, not a trap.

The majority of PM ATH days have a max morning drawdown between -0.10% and -0.50%. The distribution is tightly clustered — you rarely get the kind of large morning selloffs that can happen on non-ATH days.

YearNDD Morning≥25bp