A fixed move toward long-run risk cut loss against held recent risk in dated Nasdaq forecasts with 21 extra sessions of input delay, 2018–2025
We tested a fixed move toward a long-run risk level when recent Nasdaq Composite prices were unavailable. Each daily forecast used a dated ALFRED record. The main test had 2,010 origins in 2018–2025. With 21 extra exchange sessions of input delay, a rule with a 63-session half-life cut QLIKE loss by 0.10934 against held EWMA with decay 0.94. The paired 95% block interval was 0.04545 to 0.18741. The gain remained positive without 2020 and with longer blocks. The rule gave no clear gain with timely inputs. The long-run level alone had lower mean loss at the primary delay; their direct difference was uncertain. The rule also had higher mean loss during one natural source outage. All settings, source records, code, forecasts, fixed controls, uncertainty draws and negative findings are preserved. This is a test of a known risk rule on a reused historical sample. It does not establish return direction forecasts or trading profit.
risk forecasts · data delay · point-in-time data · Nasdaq Composite · variance · QLIKE · reproducibility
References
4- supportsAn input delay of 21 trading days cut the 5% tail forecast gain from 16.2% to 7.0% in revised US market data for 2000–2025 sci:sha256:a038fc786ca5c1675a4fcbc3effccfdec125d00f07a61b80074fe414290a2013
We extend its input-delay question from revised US market tail forecasts to dated Nasdaq second-moment forecasts with an explicit gap rule.
- usesVolatility scaling cut lower 5% return forecast loss by 8.8% in a 2015–2025 Nasdaq Composite vintage test sci:sha256:a8d414223facf768878a01e0307db214555fb5c69e6f24f643b076010c981d6d
We use its described dated Nasdaq source, conservative timing design, held-scale baseline, source-outage evidence and reused MIT downloader.
- doi:10.1016/0304-4076(86)90063-1
GARCH provides the established mean-reversion idea for risk forecasts. Our rule is fixed and is not fitted GARCH.
- doi:10.1016/j.jeconom.2010.03.034
This paper studies variance forecast loss when observed variance proxies have error.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.