Treat an opening gap as a completed session event

An opening-gap is the difference between the regular-session open and the prior regular-session close, expressed as a percentage so large and small price regimes stay comparable. In the archive workflow that overnight print is already complete. Session-analysis then separates it from the regular-session path so the leftover day can be filtered rather than traded as unfinished business.

Why the sample used relative price change

In a two-and-a-half-year daily sample of more than 650 observations beginning 2 January 2000, the average opening-gap in either direction was about 1 percent, with the largest recorded down gap 9 percent and the largest up gap 6 percent.

Prices were converted to relative-price-change percentages because the sample spanned a high of 118 and a low of 21.80. The large observation count was described as leaving an approximate plus-or-minus 3 percent sampling error around the reported frequencies.

A histogram of overnight gaps

A price-change-histogram is a binned frequency map of percentage price moves used as an explicit quantitative baseline for later comparison. When overnight gaps were placed in 0.5 percent bins, most were small: about 30 percent of down gaps and 30 percent of up gaps were under 0.5 percent, and the next 0.5-to-1 percent bin held about 27 percent of down gaps and 26 percent of up gaps.

Little link from gap direction to the close

Gap direction was nearly even in the sample, at about 51 percent up and 49 percent down. It showed no strong link to the same-day close relative to the open: after a gap up the close exceeded the open about 48 percent of the time, and after a gap down the close finished below the open about 51 percent of the time.

A leftover-session gap-fade filter

The same gap histogram implied a much stronger session-path regularity. After a gap-up open there was about a 75 percent chance the regular session would attempt to close or fade that gap in the following hours. A gap-fade-filter asks whether the open-to-close remainder is likely to fill part of the overnight gap, not whether the close will beat the open.

Quantile bands from the open to the session extreme

Quantile-analysis reads those bins as probability thresholds that mark when a move from the open is already unusual versus still typical. An open-to-high histogram placed a 2.5 percent rise from the open near a 70 percent quantile: once that band was reached there was only about a 30 percent chance of a still-higher session high. An open-to-low histogram placed a 2.5 percent decline from the open near a 60 percent quantile, so further downside beyond that band occurred in about 40 percent of sessions. Each open-to-extreme-band is a quantile band on the distance from the open to that session high or low, used as a day-scale support or resistance reference.

Bands as session order filters

The same session-analysis frame treated the vehicle as highly executable: average daily volume of about 60 million shares, a peak of 219 million, tight quoted spreads, and stock-like margin terms. The histogram bands were meant as order filters rather than as a forecast of the close.