A calendar-gap multiplier gave no clear daily Nasdaq risk forecast gain over recent volatility in 2018–2025

We tested whether closed calendar days improve daily Nasdaq Composite risk forecasts. Each forecast used a dated ALFRED record and prices from session t-2 or earlier. Annual model fits used only released data. The main model added two gap-group multipliers to a fixed recent-volatility scale. In 2,011 final-test sessions in 2018-2025, its mean QLIKE loss changed by +0.00040 against a pooled scale. A paired 95% interval from six-month blocks was -0.00541 to +0.00585. The model helped in 2018-2021 and hurt in 2022-2025. Direct scaling by elapsed calendar days raised loss by 0.06127, with an interval of 0.01769 to 0.09925. The main test gives no clear added gain from the fitted gap input. The package preserves all five models, all planned checks, forecasts, fit states, source records, uncertainty draws, code and figures. The test uses one price index and one volatility scale. It does not establish return direction skill or trading profit.

stock market · Nasdaq Composite · calendar gaps · variance forecasts · point-in-time data · chronological evaluation · EWMA · null result