A proximity based macro stress testing framework

Boris Waelchli. "A proximity based macro stress testing framework." Dependence Modeling 4.1 (2016): 251-276, electronic only. <http://eudml.org/doc/287095>.

@article{BorisWaelchli2016,
abstract = {In this a paper a non-linear macro stress testing methodology with focus on early warning is developed. The methodology builds on a variant of Random Forests and its proximity measures. It is embedded in a framework, in which naturally defined contagion and feedback effects transfer the impact of stressing a relatively small part of the observations on the whole dataset, allowing to estimate a stressed future state. It will be shown that contagion can be directly derived from the proximities while iterating the proximity based contagion leads to naturally defined feedback effects. Since the methodology is Random Forests based the framework can be estimated on large numbers of risk indicators up to big data dimensions, fostering the stability of the results while reducing inaccuracies in estimated stress scenarios by only stressing a small part of the observations. This procedure allows accurate forecasting of events under stress and the emergence of a potential macro crisis. The framework also estimates a set of the most influential economic indicators leading to the potential crisis, which can then be used as indications of remediation or prevention.},
author = {Boris Waelchli},
journal = {Dependence Modeling},
keywords = {Random Forests; Machine Learning; Stress Testing; Early Warning Indicators; Big Data; random forests; machine learning; stress testing; early warning indicators; big data},
language = {eng},
number = {1},
pages = {251-276, electronic only},
title = {A proximity based macro stress testing framework},
url = {http://eudml.org/doc/287095},
volume = {4},
year = {2016},
}

TY - JOUR
AU - Boris Waelchli
TI - A proximity based macro stress testing framework
JO - Dependence Modeling
PY - 2016
VL - 4
IS - 1
SP - 251
EP - 276, electronic only
AB - In this a paper a non-linear macro stress testing methodology with focus on early warning is developed. The methodology builds on a variant of Random Forests and its proximity measures. It is embedded in a framework, in which naturally defined contagion and feedback effects transfer the impact of stressing a relatively small part of the observations on the whole dataset, allowing to estimate a stressed future state. It will be shown that contagion can be directly derived from the proximities while iterating the proximity based contagion leads to naturally defined feedback effects. Since the methodology is Random Forests based the framework can be estimated on large numbers of risk indicators up to big data dimensions, fostering the stability of the results while reducing inaccuracies in estimated stress scenarios by only stressing a small part of the observations. This procedure allows accurate forecasting of events under stress and the emergence of a potential macro crisis. The framework also estimates a set of the most influential economic indicators leading to the potential crisis, which can then be used as indications of remediation or prevention.
LA - eng
KW - Random Forests; Machine Learning; Stress Testing; Early Warning Indicators; Big Data; random forests; machine learning; stress testing; early warning indicators; big data
UR - http://eudml.org/doc/287095
ER -