Dated wave and ratio cases need a later-sample test

The later-sample question

A reader who had already fitted a weekly-close stock model asked whether adding more names from 2000 forward and applying the same rules would count as a test on unseen data.

The editorial reply treated a development-sample test as rerunning a system on the same historical set used to build it. It treated a later-sample test as running those unchanged rules on data they had not previously seen in the same market and time frame.

A gold analog that matched only in point distance

One gold-market letter said a 2011 advance matched a late-1970s peak only in point distance and differed in the other market characteristics the writer cited. That letter also argued that dominant bullion-market institutions could generate almost any desired technical picture.

A reply on gold chart work said reading charts and locating event patterns is not a complete answer for every allocation decision. The same reply described the presented gold technical discipline as aiming to capture a large share of a large move rather than nearly the whole path of every major swing.

A long-cycle-wave still needs a case-checkpoint

A reader returned to a late-2008 long-cycle-wave discussion that had suggested share accumulation around 2010. The reader asked whether a major 2012 pullback should be avoided or whether a fully invested stance through 2016 was the intended path.

The wave-piece author replied that a correction was expected, was not described as severe, that the projected advance ran through 2016, and that a presidential cycle was said to line up with that long wave.

Editorial reading: file the count, then wait

This section is editorial, not archive attribution. TradersWeek treats an elliott-wave as a labeled swing sequence taken from price structure so a later bar can confirm or break the count. It treats a fibonacci-retracement as a ratio grid drawn on a completed swing to mark pullback zones that later price can hold or reject. Neither is a finished forecast in this reading.

Editorial view: a case-checkpoint is a calendar date attached to that wave or ratio call so a later reader can see what would have counted as a miss. Lock the structure and the invalidation before unused bars print. Then let a later-sample of the same market and time frame decide whether the case still stands.

Editorial view: the gold letter's claim that dominant institutions could generate almost any desired technical picture is the manufactured-technical-picture risk. Chart reading and event patterns, in the archive reply, were not a complete answer for every allocation decision, and the stated aim was a large share of a large move rather than nearly the whole path of every major swing. The 2012 pullback question is the nearer case-checkpoint on the long-cycle-wave: whether a correction is routine or a thesis break.

Editorial note: the archive does not report whether those later stretches confirmed or broke the cases. It records how the questions were framed.