One-day fade-the-gap rules remain unfinished across market climates

The restated one-day rule

A correspondence item restated a one-day fade-the-gap procedure as rank one day and hold one day. Gap-trading in that restatement is a one-day reverse of the prior rank day: after an up rank day the next day is short, and after a down rank day the next day is long.

Days that fail the gap test are left untraded. That abstention keeps time in the market far below a continuous holding rule. Fade-the-gap, in this sketch, means entering against the opening dislocation rather than continuing it, then exiting after a short, fixed hold.

The same rule in more than one climate

The same rule was then tabulated over successive calendar windows. Those windows covered an early span, the original demonstration span, a later declining span, and the full decade, so the same one-day entry-exit rule could be inspected in more than one market climate.

The letter writer argued that using a strongly rising interval as the demonstration window can distort any judgment of whether the fade-the-gap entry and exit rules are robust. Walk-forward-analysis, in this exchange, is the act of re-applying the same entry, exit and abstention procedure on those successive windows so the rising demonstration interval is not the only climate inspected.

The original author agreed the demonstration interval was strongly rising and said a non-rising period should be examined before the procedure is treated as complete.

A sample too thin to settle the method

The letter writer treated a short run of one-day trades on a single name, set against a much longer calendar span, as too thin a sample to decide whether the gap-fade method has value.

An illustration, not a closed recipe

The original author framed the write-up as a directional illustration of an idea rather than a finished, closed-form recipe, and added that running the exact same rule set long enough will eventually lose.

Instead of a fixed gap, the original author proposed setting the day's gap threshold from recent close-to-open-volatility, a measure of overnight dislocation used to widen or tighten the gap that must print before a signal is taken.

The original author described a fixed share count as a risk-pattern choice because dollar exposure then moves with the price level. That is fixed-share-exposure rather than compounding a fixed equity unit. Volatility-position-sizing, in this sense, makes the gap threshold or the share count depend on recent overnight movement and on how dollar exposure changes with price, so the loss shape is chosen before the fade is placed.

The original author also said a sparse single-name signal becomes more active only across a large set of names, and that transaction costs were still unmeasured in the letter writer's table.

Editorial reading

Editorial interpretation: the archive leaves a classroom sketch plus a request to inspect that sketch in more than one climate. A finished entry and exit recipe would still need walk-forward-analysis, a thicker sample, and volatility-position-sizing before the one-day fade can be treated as a closed procedure.