Predicting critical crashes? A new restriction for the free variables

TL;DR

This paper introduces a new restriction based on hazard-rate positivity to improve stock market crash prediction, significantly increasing the accuracy of identifying imminent crashes compared to previous methods.

Contribution

It proposes a novel inequality constraint on free variables that enhances crash prediction accuracy when combined with existing approaches.

Findings

25% of data satisfying the new inequality precedes crashes within a year

The new method predicts crashes with 54% success rate when combined with other approaches

Previous methods predicted only 9% of crashes, showing substantial improvement

Abstract

Several authors have noticed the signature of log-periodic oscillations prior to large stock market crashes [cond-mat/9509033, cond-mat/9510036, Vandewalle et al 1998]. Unfortunately good fits of the corresponding equation to stock market prices are also observed in quiet times. To refine the method several approaches have been suggested: 1) Logarithmic Divergence: Regard the limit where the critical exponent \beta converges to 0. 2) Universality: Define typical ranges for the free parameters, by observing the best fit for historic crashes. We suggest a new approach. From the observation that the hazard-rate in cond-mat/9510036 has to be a positive number, we get an inequality among the free variables of the equation for stock-market prices. Checking 88 years of Dow-Jones-Data for best fits, we find that 25% of those that satisfy our inequality, occur less than one year before a crash. We compare this with other methods of crash prediction, i.p. the universality method of Johansen et al., which are followed by a crash in only 9% of the cases. Combining the two approaches we obtain a method whose predictions are followed by crashes in 54% of the cases.