Bubble detection in Greek Stock Market: A DS-LPPLS model approach

Abstract

Upfront bubble detection is one of the holy grails in Financial Markets. In the present paper, in order to archive this goal, we consider two different methods based on the Log Periodic Power Law. We implement this early detection algorithms in the Greek Stock Market, which is a relatively “shallow” and underdeveloped market. We have examined a period from 1997 until the end of 2019, an epoch before the rise of COVID-19 virus. Using this methodology, we managed to detect with a relatively good accuracy the formation and the critical time for both positive and negative financial bubbles that occurred during the examination period. © 2021 Elsevier B.V.