BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
 M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20261106T144143Z
DTSTART:20261202T140000Z
DTEND:20261202T150000Z
SUMMARY:Statistics Seminar - Yifan Li (Manchester\, Business School)
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}mhy-mw3ocpg
 m-yhzrmm
DESCRIPTION:Title: Fixed-k Inference for Explosive Drift\n\nAbstract: We 
 propose a new framework for uniform inference on explosive drifts in hig
 h-frequency data. Standard large-bandwidth asymptotics often fail in thi
 s context because the “spot” test statistics computed over short windows
  are far from Gaussian. By treating the window size k as fixed\, we show
  that the spot statistics are coupled with a sequence of dependent Stude
 nt-t variables\, and their maximum converges to a Frechet distribution r
 ather than the conventional Gumbel limit. We establish a novel anti-clus
 tering condition for dependent Student-t processes to justify this limit
  theory under overlapping estimation windows. A local power analysis rev
 eals that explosive drifts induce a multiplicative power transformation 
 of the limiting distribution\, contrasting with the additive location sh
 ift characteristic of Gaussian theory. Empirically\, we show that the pr
 oposed coupling-based test offers superior size control and reveals that
  statistically significant intraday price explosions are far rarer than 
 suggested by Gaussian-based methods. 
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:Frank Adams 2\, Alan Turing Building\, Manchester
END:VEVENT
END:VCALENDAR