·
Stress
Testing - alerts
bank management to adverse unexpected outcomes related to a variety of risks
and provides an indication of how much capital might be needed to absorb losses
should large shocks occur.
o
Bottom-up approaches tend to use the results of
sensitivity analysis to identify sensitive dependence on risk factors as
starting points. As a consequence, those scenarios are chosen which involve
risk factors having the largest impact. For example, for a bank focusing on real
estate, GDP, employment rate, inflation rate, spending capacity in the
countries, it is acting in, will be of more relevance than the oil price,
exchange rates, etc. Thus, it will look for scenarios involving the relevant
risk factors.
o
Top-down approaches start with a chosen scenario, e.g.,
the terror attack in New York on September 11, 2001, and require the analysis
of the impact of this scenario on the portfolio. The task in this situation is
to identify those tests which cause the most dramatic and relevant changes.
Ø Historical
Scenarios
Historical scenarios are easy to
implement, as one only has to transfer the values of risk factors corresponding
to a historic event to the current situation. In most cases, it does not make
sense to copy the value of the risk factors, but to determine the change of
value (either in absolute or in relative form) which is accompanied by the
insertion of the event and assume it also applies to the actual evaluation. The
use of scenarios like: Economic or industry downturn, market-risk events,
liquidity shortage.
Method of Stress Testing-
·
If
macroeconomic parameters are not part of the input for determining the risk
parameters which are stressed, there are three steps required for macro stress
tests.
o
Firstly,
it is necessary to model the dependence of the risk parameters on the risk
factors.
o
Secondly,
it is necessary to choose values for the risk factors which are representative
for stress events.
o
Thirdly
to reproduce correlations and causal interrelations between risk factors and stress
events, intricate (macro-economic), methods of estimation and validation are
needed.
Once the candidate variables for inclusion were selected, we
examined the importance of each variable separately so as to decide whether it
should be included into the model or not. In order to perform this analysis,
two different approaches have been used:
o
Pearson’s chi–squared statistic
o
Univariate logistic regression (likelihood ratio)
Multifactor sensitivity analysis-
Stressed PDs are modeled as a
function of stressed macroeconomic variables such as GDP, unemployment rate,
CPI, industrial production, housing price index, etc.
Once a history of portfolio PDs and
macroeconomic factors is collected, a relation between the portfolio PD and
macroeconomic factors could be estimated.
•As an example, a linear regression
model could be used for estimation:
Historical PD = β1 x GDP + β2 x unemployment rate + e
After this by adding stressed values
of macroeconomic variables we can calculate stress PD.
Stressed PD = 1/ (1+exp (- (β1 x GDP
+ β2 x unemployment rate + e))
•In the context of this model, a
distribution of future portfolio PDs could be derived. This allows an
alternative definition of a stress portfolio PD: It could be defined as a
quantile of this PD distribution, e.g. a 95% quantile.
Exposure at Default – PV*(Recoveries) + PV*(Costs) / Exposure at Default
PV is calculated by Weighed Average cost of capital-
The weighted average cost of
capital, WACC, is the minimum rate of return allowable and still meet financing
obligations.
We used balance sheet data with a
time gap of 12 to 24 months prior to default to characterize insolvent firms.
Second, due to missing bankruptcy data in the data sample to
No comments:
Post a Comment