BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:Finance Research Workshop: "Financial Market Fragility in the 
 Era of AI Planning"
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260719T031717Z
UID:tag:localist.com\,2008:EventInstance_52239373266864
DTSTART:20260306T153000Z
DTEND:20260306T170000Z
DESCRIPTION:This Friday\, we will host our final seminar speaker for the sp
 ring semester\, Professor Winston Dou\, Wharton\, University of Pennsylvan
 ia.\n\nProfessor Dou will present their preliminary draft of\, "Financial 
 Market Fragility in the Era of AI Planning"\, co-authored with Itay Goldst
 ein and Yan Ji  at 10:30 AM in the Finance Department conference room\, 12
 00 RCB. \n\nAbstract:\n\n"This paper examines how AI planning\, the core t
 echnology behind agentic AI systems that pursue long-horizon objectives by
  anticipating how current actions shape future payoff-relevant states\, af
 fects financial market stability. We develop a dynamic trading framework w
 ith positive-feedback investors\, constrained arbitrageurs\, and oligopoli
 stic informed speculators who may coordinate intertemporally: trading aggr
 essively in tandem to generate (negative) bubbles and subsequently unwindi
 ng their positions in a coordinated manner to extract profits. Such intert
 emporal coordination differs from traditional collusion because it faces t
 wo unique\, fundamental obstacles: time inconsistency\, as coordinated pla
 ns become incentive-incompatible once a large (negative) bubble has formed
 \, and weak punishment\, as deviations are difficult to penalize when no l
 arge (negative) bubble materializes. We characterize equilibria featuring 
 coordinated creation of manipulative\, exploitative (negative) bubbles. In
  simulation experiments\, AI speculators trained via reinforcement-learnin
 g algorithms with explicit planning modules autonomously discover and impl
 ement intertemporal collusive trading strategies based on compounded price
 -trigger rules\, coordinating without communication or shared intent. When
  feedback trading is strong\, these AI-planning speculators dynamically co
 nverge on destabilizing strategies that create and exploit (negative) bubb
 les\, manipulate feedback traders\, and significantly amplify market fragi
 lity."
GEO:33.755064;-84.389989
LOCATION:Robinson College of Business\, Finance Department Conference Room\
 , 1200 RCB
SUMMARY:Finance Research Workshop: "Financial Market Fragility in the Era o
 f AI Planning"
URL;VALUE=URI:https://calendar.gsu.edu/event/finance-research-workshop-fina
 ncial-market-fragility-in-the-era-of-ai-planning-1755
CATEGORIES:Lecture
END:VEVENT
END:VCALENDAR