h**b
发帖数: 406 | 1 100%default prob. * 100% loss = 1,大牛给解释一下.怎么还会超过1？ | A*****s
发帖数: 13748 | 2 是不是得仔细看一下swap的term？
如果第一笔premium没交的时候就default了，那么implied pd大于1是有可能的吧？
【在 h**b 的大作中提到】
: 100%default prob. * 100% loss = 1,大牛给解释一下.怎么还会超过1？
| r********e
发帖数: 169 | 3 spread quote or price quote? | h**b
发帖数: 406 | 4 bloomberg上面的Greece CDS SR 1y | A*****s
发帖数: 13748 | 5 这样想把：
假设这个swap是annual settlement的
现在距离下一个settlement还有3个月
如果default发生在不到3个月的时候
你给writer按照pro-rate交premium，肯定小于3个月内的expected-loss
而writer要付给你真正的loss，约为LGD
如果writer认为希腊3个月内一定会破产
那么他们即便按照100% expected-loss (100% LGD)来charge你，都是输定了的
原因见上方，因为你交的premium要被pro-rate一下
所以他们就要按照大于100%的expected-loss来charge你
根本的原因就是default近在眼前，都没有人想个你玩几个月以后再settle的游戏了
我记得大概是这样，曾经GM的CDS implied expected loss也是>100%
如果错了请大牛们指正
【在 h**b 的大作中提到】
: bloomberg上面的Greece CDS SR 1y
| c*********n
发帖数: 128 | 6 Nothing wrong about CDS rate>1 at all.
It's the default intensity that relates to the CDS rate, not the default
prob.
The default intensity could be way above 1.
Say if you know Greece will default for sure within 3m, and the recovery
value is 0, then theoretically the CDS rate should be > 400%.
Or think it this way, if the CDS rate is 100%, and default happens at 3m
with prob 1, then seller pays the buyer the Notional (assuming 0 recovery),
and the buyer pays the seller the accrued coupon, which is Notional*100%*(1/
4) = Notional/4. The cashflow is obviously negative for the seller for sure.
【在 h**b 的大作中提到】
: 100%default prob. * 100% loss = 1,大牛给解释一下.怎么还会超过1？
| D********n
发帖数: 978 | 7 All theoretical argument. Indeed, market is assuming 60% hair cut and
trading in terms of points upfront.
,
1/
sure.
【在 c*********n 的大作中提到】
: Nothing wrong about CDS rate>1 at all.
: It's the default intensity that relates to the CDS rate, not the default
: prob.
: The default intensity could be way above 1.
: Say if you know Greece will default for sure within 3m, and the recovery
: value is 0, then theoretically the CDS rate should be > 400%.
: Or think it this way, if the CDS rate is 100%, and default happens at 3m
: with prob 1, then seller pays the buyer the Notional (assuming 0 recovery),
: and the buyer pays the seller the accrued coupon, which is Notional*100%*(1/
: 4) = Notional/4. The cashflow is obviously negative for the seller for sure.
| c****y
发帖数: 3592 | | w******0
发帖数: 4 | 9 借问一下各位大牛。有没有比较好的方法通过CDS time series to calibrate the
intensity process?
【在 D********n 的大作中提到】
: All theoretical argument. Indeed, market is assuming 60% hair cut and
: trading in terms of points upfront.
:
: ,
: 1/
: sure.
| h*y
发帖数: 1289 | 10 There are two quoting conventions: upfront and par spread.
upfront is associated with two standard running coupon: 100bps or 500bps.
But no matter what upfront can never exceed 100%. You will not protect $1 by
using more than $1.
par spread can exceed 100% if market expectation is to default within 3M.
Assuming A/360 and zero recovery, the extreme case is a 36000% par spread, i
.e., entity will default tomorrow for sure. |
|
