作者: Didier Sornette , Anders Johansen
DOI: 10.1016/S0378-4371(97)00318-X
关键词:
摘要: We propose that large stock market crashes are analogous to critical points studied in statistical physics with log-periodic correction scaling. extend our previous renormalization group model of prices prior and after [D. Sornette et al., J.Phys.I France 6, 167, 1996] by including the first non-linear correction. This predicts existence a log-frequency shift over time oscillations crash. is tested on two largest historical century, october 1929 1987 crashes, fitting index an interval 8 years crashes. The good quality fits, as well consistency parameter values obtained from promote theory have their origin collective ``crowd'' behavior many interacting agents.