Hello, in a synthetic CDO, junior tranches have the following income: (1) sell credit protection to receive the premiums (2) receive income from the high-quality securities However, what are the income sources for the senior tranches? Senior tranches buy credit protection through CDS and pay premiums. Please enlighten me. Thank you.
The sources are the same for both tranches. Senior tranches get paid before junior ones. The Spv sell protection, and from that income pay Senior and junior as the waterfall rules
thank you, but the senior tranches buy protection, are they buying protection from the SPV (which is on behalf of the junior tranches)?
June, whoever buy a tranche whether funded or not is selling protection. In a synthetic CDO it’s just that it works with CDS, it might be funded, there might be a supersenior swap but the logic is still the same
A synthetic CDO is trying to mimic a cash CDO using credit default swaps. Not 100% sure anymore, but I believe the replication goes like this: Risky Bond A = Treasury bond (of same duration) + Premium from selling CDS on A The CDS premium might be spread payments that come at coupon time, or it might be a lump sum payment equal to the PV of expected spread payments. The lump sum would be reinvested as treasuries. The idea is that if you have a T-bond/note + CDS premium, you will be getting the same coupons as you would with the risky bond. If the risky bond defaults, you have to pay out the market value of the bond before default, which is the same as you would lose if you just held the risky bond to begin with. — Now a cash CDO is just a large collection of Bonds from various companies and of various qualities. The tranche structure redistributes the risk of default among various classes through the “waterfall” payment rules. If only a few bonds default, then only the low grade tranches have their income streams affected. If lots of bonds default, then the senior tranches start paying less. How safe the senior tranches are depends on: 1) what % of the total value must be in default before senior cash flows are impacted, 2) overall likelihood of bonds defaulting, and 3) correlation of defaults (are they more or less random, or do they all hit the same time), With a synthetic CDO, you just have a bunch of treasuries and CDS premiums instead of a bunch of bonds. So a synthetic CDO doesn’t really buy protection; it sells it. It is possible that CDO managers may decide that they’ve underestimated the likelihood of default and start to buy protection. If they are right, then they will be adding value by reducing the exposure of their fund to credit events, and protecting all tranches that currently pay. Does that help?
bchadwick Wrote: ------------------------------------------------------- > If the risky bond > defaults, you have to pay out the market value of > the bond before default, which is the same as you > would lose if you just held the risky bond to > begin with. > If the risky bond defaults (or another credit event happens like a restructuring) then the protection buyer delivers a bond which the protection seller buys at par. Most CDS call for delivery.
Thanks Joey, I remembered that there were several ways to do delivery, and that they had to add up to losing the bond. One way is [deliver par, receive risky bond], the other way is [pay cash to make bond whole at par]. I guess there are a bunch of assumptions about recovery rates for defaulted and credit-event bonds too, so if they have good people and lawyers who can recover more than average, it makes sense to go the “take delivery” route rather than the “pay cash” route.
long time lurker, but as my tag might suggest this is something I think (hope?) I can contribute on. bchad- you’re right to the first approximation. In general, delivery of the underlying in a CDO backed by corporates is fairly rare- generally they are settled at the market rate for the defaulted bond (which is theoretically the market’s PV of the recovery). Most CDO managers aren’t paid enough in fees to make it worthwhile to get involved in restructuring/distressed situations, though there are a couple exceptions. Joey- a more correct statement would be ‘Most single-name CDS (and almost all in CDOs) call for delivery’. No one in the market used to like delivery- rating agencies require it bc they think it removes market risk. In the current ABS CDO brouhaha, that’s been a negative for ABS CDO noteholders bc shorts can force money out of the deal via delivery before noteholders get paid.
a more correct statement would be ‘Most single-name CDS (and almost all in CDOs) call for delivery’ I’ll go with that, but getting a little picky here. We were talking about CDS in synthetic CDO’s…
Thank you so much bchadwick, this is very good explanation bchadwick Wrote: ------------------------------------------------------- > A synthetic CDO is trying to mimic a cash CDO > using credit default swaps. > > Not 100% sure anymore, but I believe the > replication goes like this: > > Risky Bond A = Treasury bond (of same duration) + > Premium from selling CDS on A > > The CDS premium might be spread payments that come > at coupon time, or it might be a lump sum payment > equal to the PV of expected spread payments. The > lump sum would be reinvested as treasuries. > > The idea is that if you have a T-bond/note + CDS > premium, you will be getting the same coupons as > you would with the risky bond. If the risky bond > defaults, you have to pay out the market value of > the bond before default, which is the same as you > would lose if you just held the risky bond to > begin with. > > — > > Now a cash CDO is just a large collection of Bonds > from various companies and of various qualities. > The tranche structure redistributes the risk of > default among various classes through the > “waterfall” payment rules. If only a few bonds > default, then only the low grade tranches have > their income streams affected. If lots of bonds > default, then the senior tranches start paying > less. How safe the senior tranches are depends > on: > > 1) what % of the total value must be in default > before senior cash flows are impacted, > 2) overall likelihood of bonds defaulting, and > 3) correlation of defaults (are they more or less > random, or do they all hit the same time), > > With a synthetic CDO, you just have a bunch of > treasuries and CDS premiums instead of a bunch of > bonds. > > > So a synthetic CDO doesn’t really buy protection; > it sells it. It is possible that CDO managers may > decide that they’ve underestimated the likelihood > of default and start to buy protection. If they > are right, then they will be adding value by > reducing the exposure of their fund to credit > events, and protecting all tranches that currently > pay. > > Does that help?