The Statistical Properties of the Maximum Drawdown in Financial Time Series

10 Pages Posted: 2 May 2012 Last revised: 15 May 2013

Alessandro Casati

Antares Technologies

Serge Tabachnik

Antares Technologies

Date Written: August 20, 2012

Abstract

The purpose of this work is to study the statistical properties of the MDD for stochastic processes characterized by the stylized facts of real financial time series. The numerical results obtained using a Monte Carlo code are firstly validated against the analytical predictions available within the academic framework of the Brownian motion. The statistics of the maximum drawdown is then analyzed in term of both its expectation value and distribution function for processes whose increments are not independent and present non vanishing excess kurtosis and skewness. The expectations for the maximum drawdown are finally compared, along with their confidence intervals, to the events observed in the historical financial time series of different asset classes.

Keywords: quantitative risk management, maximum drawdown, Monte Carlo, auto-regressive process

JEL Classification: C50, C53, C22, C15, C40

Suggested Citation

Casati, Alessandro and Tabachnik, Serge, The Statistical Properties of the Maximum Drawdown in Financial Time Series (August 20, 2012). Available at SSRN: https://ssrn.com/abstract=2049584 or http://dx.doi.org/10.2139/ssrn.2049584

Alessandro Casati (Contact Author)

Antares Technologies ( email )

19, Boulevard Malesherbes
Paris, 75008
France

Serge Tabachnik

Antares Technologies

19, Boulevard Malesherbes
Paris, 75008
France

Paper statistics

Downloads
767
Rank
24,876
Abstract Views
2,502