Lock the holiday window as a regime, then veto resistance

Two layers around one holiday stretch

The archive workflow isolated a late-year holiday stretch by freezing both the calendar and the stock list. Purchases were locked to the close after Thanksgiving and sales to the final close of the year. The same five names were held at five hundred shares each across three consecutive windows, and individual fundamental and technical study was withheld at entry.

Editorial teaching reads that design as two layers. The first layer tests the window itself as a holiday-regime through a locked-window on a mechanical-basket. The second layer asks whether nearby support and resistance should withhold names whose chart location fights that short hold.

A recurring late-year holiday-regime

On a multi-year holiday overlay of the stock market, many marked holidays showed an advance that began a few sessions earlier and often continued after the closure. The stretch from after Thanksgiving through Christmas was treated as a recurring late-year window. It could still appear in relatively slow or weak years, and it sometimes extended past year-end.

Editorial note: that description is a regime backdrop. It does not say every name participates. The later record in this case showed name-selectivity inside the same holiday-regime.

Lock the window so the calendar is the hypothesis

The case locked purchases to the close after Thanksgiving and sales to the final close of the year so the outcome could be attributed to the window rather than discretionary timing. In editorial terms, that procedure is a locked-window: buys and sells are fixed to named session closes.

The illustration used five names and five hundred shares each, withheld individual fundamental and technical study at entry, and repeated the same list across three consecutive windows. In editorial terms, that design is a mechanical-basket. The test depends on the calendar regime, not on rotating selection.

The same basket, three different market settings

The first of those three identical windows produced a loss on the unfiltered basket even though the broader industrial average rose over that calendar year after an autumn setback. At least one selected name was bought near resistance left by a prior selloff while the market was still absorbing an autumn correction. That chart location was later treated as a poor site for a one-month seasonal hold.

The next window followed a mid-autumn low, so the locked seasonal buy occurred after the market had already turned higher rather than during the decline. Across the three windows, participation was name-selective: some holdings advanced while others lagged or lost inside the same holiday-regime.

The third window included a sharp dip just after Thanksgiving that reversed quickly, and the industrial-average advance continued into the following calendar year.

Add a resistance-veto after the mechanical record

After the mechanical results were recorded, trend and resistance on the individual charts were proposed as a filter for withholding names that sat at unfavorable structure during the short seasonal hold.

Editorial interpretation: a resistance-veto is that withhold step. It does not replace the locked-window. It asks whether price sits at a prior supply area after a decline, a setting already flagged as a poor site for this one-month seasonal hold.