w*****t
发帖数: 49 | 1 Credit Risk Parameters (including PD, LGD and UGD).
本版牛牛们的解释通俗易懂，比google的东西好太多了。
期待大牛们的帮助。 | l***a
发帖数: 12410 | 2 prob of defaul
loss given default
don't know UGD...
你去申请sf那个中型银行了？
【在 w*****t 的大作中提到】
: Credit Risk Parameters (including PD, LGD and UGD).
: 本版牛牛们的解释通俗易懂，比google的东西好太多了。
: 期待大牛们的帮助。
| l*******r
发帖数: 33 | 3 basel?
【在 w*****t 的大作中提到】
: Credit Risk Parameters (including PD, LGD and UGD).
: 本版牛牛们的解释通俗易懂，比google的东西好太多了。
: 期待大牛们的帮助。
| w**********y
发帖数: 1691 | 4 PD - Default Probability
In Merton’s framework, If asset price (total net value) of a company is X,
and it will default if X goes below a threshold X*(constant), then PD=P(X
In Statistics, we can assume X is normal/lognormal or whatever;
In Stochastic, this is called First Passage Time
LGD – Loss Given Default
Suppose a collateral of a loan is Y (proportion, ranging in [0,1]).
Then if the counterparty of the loan defaults, you can only get back Y, and
thus the loss is 1-Y (Y>1 is almost impossible). The LGD = Y|X
Basel and Economical capital model use different models for LGD.
LGD could be modeled as a beta distribution..
PD and LGD could be highly correlated, since they are all depending on the
overall economic condition (systematical factor) and the company’s
condition (idio factor).
Actually, I gave an post of “interview question” on quant board, which is
originally motivated by the story here..
【在 w*****t 的大作中提到】
: Credit Risk Parameters (including PD, LGD and UGD).
: 本版牛牛们的解释通俗易懂，比google的东西好太多了。
: 期待大牛们的帮助。
| D******n
发帖数: 2836 | 5 coporate risk is different from consumer risk. Don't know if one can transfe
r from one field to another one.
,
X*)
and
【在 w**********y 的大作中提到】
: PD - Default Probability
: In Merton’s framework, If asset price (total net value) of a company is X,
: and it will default if X goes below a threshold X*(constant), then PD=P(X: In Statistics, we can assume X is normal/lognormal or whatever;
: In Stochastic, this is called First Passage Time
: LGD – Loss Given Default
: Suppose a collateral of a loan is Y (proportion, ranging in [0,1]).
: Then if the counterparty of the loan defaults, you can only get back Y, and
: thus the loss is 1-Y (Y>1 is almost impossible). The LGD = Y|X: Basel and Economical capital model use different models for LGD.
| s*********e
发帖数: 1051 | 6 risk parameters are important for economic models, stress testing, RAROC,
and basel accords.
Here is a little math ==> EL = PD * EAD * LGD
where PD ==> probability of default
EAD ==> exposure at default
LGD ==> loss given default
The measurements of PD and LGD are the same for both termed loan and line of
credit. However, there is big difference in EAD. For termed loan, EAD is as
simple as nominal amount of exposure. However, for line of credit, there
are 3 measures of EAD, each of which describes different situation of EAD
estimation. They are LEQ, CCF, and EADF.
The best reading is "The Basel II Risk Parameters: Estimation, Validation,
and Stress Testing". | H*******s
发帖数: 382 | 7 UR is widely used for EAD
of
as
【在 s*********e 的大作中提到】
: risk parameters are important for economic models, stress testing, RAROC,
: and basel accords.
: Here is a little math ==> EL = PD * EAD * LGD
: where PD ==> probability of default
: EAD ==> exposure at default
: LGD ==> loss given default
: The measurements of PD and LGD are the same for both termed loan and line of
: credit. However, there is big difference in EAD. For termed loan, EAD is as
: simple as nominal amount of exposure. However, for line of credit, there
: are 3 measures of EAD, each of which describes different situation of EAD
| n*********y
发帖数: 474 | 8 现在银行里每个组都想涉及点Basel，呵呵
我也在学习中 |
|
