Opening-referenced percentile stops for same-session gaps

Same-session gap reversals

A gap-up or gap-down open that reverses during the same session can leave a position entered at the open with an unrealized loss by the close. Gap-analysis reads that gap open against the prior close and the same-session high-low as a repeatable chart condition that can be tested rather than trusted as confirmation.

Editorial: treat the gap open as the origin of risk, not as confirmation. The same print that makes a directional story look finished is the price from which a stop-loss can be measured before a same-session gap reversal rewrites the loss.

Wheat and crude-oil sessions

In a December 1991 wheat session, a long at the 340.00 gap-up open saw a 328.50 low and a 330.50 close, a 9.50-cent unrealized loss versus the open even though that close was 5.75 cents above the prior 324.75 settlement.

In an August 1990 crude-oil session, a short at the 26.05 gap-down open saw a 26.98 high and a 26.84 close, 0.79 above the open, recorded as a 790-dollar unrealized loss per contract.

Why the stop is referenced to the open

Because the reversal forms after an entry at or soon after the open, the stop-loss is placed relative to the opening price rather than the prior settlement. A stop-loss is a precommitted filter that bounds loss or exposure from a stated reference price before entry and while the position is open.

An opening-referenced stop is a protective order whose trigger is offset from the session open rather than from the prior close or a later extreme.

Open-to-extreme spreads and percentile stop distance

An up period is a day or week that settles above its open. A down period is a day or week that settles below its open. An open that is also the period low or high produces a zero open-to-extreme spread.

Quantile-analysis orders historical open-to-extreme observations over a stated lookback so a chosen percentile can be compared with a later session. Stop width is taken from the percentile distribution of the open-to-low spread on up periods and the open-to-high spread on down periods, using a historical sample such as January 1987 through December 1991.

On the wheat up-day sample, the 90th-percentile open-to-low distance was 2.50 cents, so a sell stop 2.50 cents under a 340.00 open sat at 337.50, with a stated one-in-ten chance of trading through that level and still settling above the open.

On the crude-oil down-day sample, the 90th-percentile high-to-open distance was 29 ticks, so a buy stop 29 ticks above a 26.05 open sat at 26.34, with a stated 0.10 probability of trading through that level and still settling below the open.

Closer percentiles and sample consistency

A closer percentile stop reduces the cash loss if hit and increases the chance of being stopped on a session that later settles in the intended direction.

A comparison of those open-to-extreme spread distributions for 1980-87 and 1987-91 was reported as highly consistent.