opVaR-package: Computing operational risk - opVaR package

In barryrowlingson/opVaR: Computes operational risk

Description

opVaR package computes operational risk by Monte Carlo simulation performed by mc function. It provides reading losses, merges them by chosen periods, fits frequency and severity distributions. Finally, VaR is computed basing on simulated yearly losses.

Details

Package:	opVaR
Type:	Package
Version:	1.0
Date:	2010-05-30
License:	GPL-3
LazyLoad:	yes

Author(s)

Anna Patrycja Zalewska, Faculty of Mathematics, Informatics and Mechanics at the University of Warsaw.

Maintainer: Anna Patrycja Zalewska <[e-mail]>

Examples

data(loss.data.object)

D<- loss.matrix(loss.data.object);D
	# matrix of loss summaries

loss.matrix.image(data=loss.data.object)
	# table of loss summaries

loss.density(a=1,b=2,loss.data.object)
	# loss denisties for every business line
	# and second risk category
	# "Clients, Products and Business Practices"

x<- read.loss(3,2,loss.data.object)
	# read losses from third business line
	# and second risk category

hist.period(x,"days")
	# frequency histogram for days
root.period(x,"days","nbinomial")
	# fitted frequency - nbinomial

z<- x[,2] # these are losses amounts
par(mfrow = c(1,2))
fit.plot(z,dnorm, param = list(mean = mean(z),sd = sd(z)))
fit.plot(z,dnorm, param = list(mean = mean(z),sd = sd(z)),draw.diff=T)
	# empirical and fitted severity distributions

loss.fit.dist("lognormal",x)
	# fitting lognormal distribution 
	
mc(x,rfun="log-normal",nmb=200)
	# 200 yearly losses simulated for x
	# with given lognormal distribution as severity distribution
	# and chosen binomial distribution as frequency distribution

barryrowlingson/opVaR documentation built on May 11, 2019, 7:24 p.m.