Unused bars and dropped names decide a vertical-runup signal

A hygiene drill, not a finished call

A published figure showed a six-bar vertical-runup that a reader thought met the four-bar-minimum. The reader asked why that advance was omitted. A second question asked whether a file limited to still-listed names would change the candidate count. Together those questions turn the figure into a research-hygiene drill.

Runup-consolidation-runup is a three-stage price structure: a sharp advance, a pause, then another advance. In this archive it is a testable signal hypothesis, not a forecast. The first leg is a vertical-runup, a steep multi-bar advance. That leg is usable only after the chartist can explain every unused bar and whether delisted names sat in the sample.

Why the six-bar rise was unused

A reader asked why a six-bar vertical advance on the published figure was omitted, arguing it appeared to meet a four-bar-minimum before a sizable overlapping-bar. An overlapping-bar is a price bar whose range substantially covers a prior bar and interrupts a clean vertical advance. The four-bar-minimum is the visual threshold in which an advance is counted only after at least four bars of rise before a material overlap.

The same reader asked whether two smaller bars, similar to an annotated pause segment, or the later overlap caused that advance to fail the author's visual test.

The author said the unused October 2015 advance was shown only to display two different post-run outcomes and sat outside the study window, which closed in November 2013.

Editorial reading: placing a later rise outside the window explains why it was not a study case. It does not, by itself, say whether the two smaller bars or the overlapping-bar failed the visual test.

Survivorship-bias in the candidate file

A reader argued that a file limited to still-listed names could change both the number of vertical-runup candidates and any bull-market results drawn from names that lasted an 18-year window. That reader reported 1,846 delistings from May 2008 to August 2011, a little more than 1,000 names surviving a 20-plus-year span beginning in 1995, and about 7,000 or more stocks on a typical date.

The author said the research used both a daily-follow list and an archive-database of merged, unlisted, or dropped names, and acknowledged the survivorship concern. Survivorship-bias is sample distortion that appears when a study keeps only names that remain listed at the end of the window. An archive-database is a research file that retains merged, unlisted, or dropped names so pattern counts are not limited to current listings.

Editorial reading: using an archive-database answers the listing-universe question only if dropped names were scanned with the same four-bar-minimum as names that remained listed.