Act on a pattern change, not on formula signals

Mechanical flow belongs on the daily bar

Listed options, futures and stock-to-futures basket arbitrage changed how technicians read markets by adding mechanical, nonfundamental decisions that show up in volatility and swings.

What looks like noise after those instruments arrived was treated as a real market force to include in the reading, not a residue to strip from the daily bar.

The public share of transactions was recalled as about two thirds in an earlier exchange-floor period and later estimated at less than one third.

Who is forced to act

Institutional response to a toppy chart or a rating cut sped from a monthly meeting, to a weekly meeting, to waiting for a manager to return from lunch, to executing immediately without details.

House rules moved toward a fully invested or almost fully invested mandate, so selling one holding required buying another. Consultant reallocations enforced that constraint and were cited as one reason the tape rotated on a two-to-three-day horizon.

Same-session broadcast and desktop charts compressed reaction: put buying within 30 seconds of a drop and a correction-headline frame within two minutes.

Editorial reading: this is a regime of who is forced to act. Specialists, institutions, consultants and derivative-linked flows set the constraints a single trade has to sit inside.

A few hand-tracked internals

Hand-posted daily bars were treated as the least abstract, time-linked summary of the tape. Tick charts were too short-term, weeklies and point-and-figure more abstract, and computer oscillators such as a stochastic crossover above 80 were called seductive but too inconsistent to rely on.

A 10-day moving average of the advance-decline differential was kept as a simple overbought/oversold oscillator after 9-, 11- and 13-day windows were compared, and a 10-day versus 30-day cross was the only mechanical breadth rule retained.

Daily new-high and new-low counts were favored as a raw internal. Fewer new lows against a falling average were read as fading downside momentum, including around the June 1982 low and the 1990 low after the Kuwait invasion.

Editorial reading: those counts turn advance-decline, new-high and new-low participation, and price structure into a checkable hypothesis about whether a move is still intact. Standing aside stays inside the same procedure until that hypothesis changes.

A lagged forced-sale measure

Specialist short sales as a share of total shorts were read as a late-rally forced-sale measure, reported with a two-week lag. Contemporary bands were about 48-52% on the high side and 29-32% on the low side.

A recent 41 reading was not treated as late-cycle until a rise toward 44 or 45 or a rising trend against a still-rising market appeared.

Editorial close: entry, exit and abstention wait on a change in that participant pattern. A specialist share still at 41 is the same standing-aside case as an oscillator print that has not yet been joined by a change in who is forced to act.