Title: Stress test techniques using drawdown metrics: a Brazilian case study

Authors: Arthur Geronazzo; João Luiz Chela

Addresses: Getulio Vargas Foundation (FGV), Rua Itapeva, 432, São Paulo, Brazil ' Getulio Vargas Foundation (FGV), Rua Itapeva, 432, São Paulo, Brazil

Abstract: The main objective of investors is to obtain the highest possible return, by running the lowest risk. This paper attempts to present a series of risk metrics based on maximum drawdown historical distributions. Maximum drawdown provides the information of the largest drop in the asset value that an investor can have in a given time interval. The metrics use historical simulation with different time intervals, holding period and confidence intervals. Backtest of these metrics is done to verify their adherence, so it shows that the maximum drawdown at risk using GEV metric is the metric that presents the highest approval rates in the different scenarios. The main contribution of this paper is the presentation of diferent risk metrics based on maximum drawdown, analyse of the best metric for each situation and applications of the metrics to risk management and to stress scenarios.

Keywords: maximum drawdown; risk metrics; stress test; extreme value theory; maximum drawdown at risk; MDaR; conditional expected drawdown.

DOI: 10.1504/IJFMD.2020.111886

International Journal of Financial Markets and Derivatives, 2020 Vol.7 No.4, pp.315 - 336

Received: 26 Sep 2019
Accepted: 06 Apr 2020
Published online: 18 Dec 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article