Hi Guys, I am having trouble understanding why US swap/GOV spread could be negative for 5 year onward. How can the credit quality of banks be more than US government? I know this have been going on for a while now but I haven't found any satisfactory response online. Comments? Thanks!Fish

It is obviously deeply disturbing, but if you can find a way to arb it, please let us know!

Swap spread is a value of insurance while bond yield signifies value of US Treasury USA and world demand. These are close notions but not the same

If you search the fora here, you will find a few old discussions that cover this subject. You will have to filter out list's gibberish, though. If you need more help, you can then ask more specific questions.Again, list/list1, I have to ask you to please stop posting on this subject. I am certain that my requests will fall on deaf ears, but I have to try.

Martinghoul are you still think that irs swap rate is a discount factor and one can invest $s and get $1 at a future moment T? If one opens then can read a handbook at irs fixed rate can protect himself against variability of the floating leg. This feature is a specific of insurance.If we consider in theory the case when floating leg tends to a linear function with known coefficients then in limit we get that fixed rate and floating leg payments are equal. Hence the spread between floating US Treasury rate and corresponding IRS is subject to Treasury rate variability over specified period. As we mentioned above the use IRS fixed is an insurance against variability. It is a first order approximation as we ignored external world effect.Of course, it takes a couple lines to present alternative point.

QuoteOriginally posted by: list1Martinghoul are you still think that irs swap rate is a discount factor and one can invest $s and get $1 at a future moment T? If one opens then can read a handbook at irs fixed rate can protect himself against variability of the floating leg. This feature is a specific of insurance.If we consider in theory the case when floating leg tends to a linear function with known coefficients then in limit we get that fixed rate and floating leg payments are equal. Hence the spread between floating US Treasury rate and corresponding IRS is subject to Treasury rate variability over specified period. As we mentioned above the use IRS fixed is an insurance against variability. It is a first order approximation as we ignored external world effect.Of course, it takes a couple lines to present alternative point.Yet again, you spout utter gibberish, which not only demonstrates a total lack of understanding of the market instruments being discussed, but, also more generally, makes no sense whatsoever. This was the case in the older thread dedicated to the same topic and you have clearly not gotten any saner and/or wiser. Therefore, I refuse to engage in any discussions with you here and I ask you, please, to stop, just like I did previously. As an aside, where is the "Ignore" feature when you need one?

QuoteOriginally posted by: MartinghoulQuoteOriginally posted by: list1Martinghoul are you still think that irs swap rate is a discount factor and one can invest $s and get $1 at a future moment T? If one opens then can read a handbook at irs fixed rate can protect himself against variability of the floating leg. This feature is a specific of insurance.If we consider in theory the case when floating leg tends to a linear function with known coefficients then in limit we get that fixed rate and floating leg payments are equal. Hence the spread between floating US Treasury rate and corresponding IRS is subject to Treasury rate variability over specified period. As we mentioned above the use IRS fixed is an insurance against variability. It is a first order approximation as we ignored external world effect.Of course, it takes a couple lines to present alternative point.Yet again, you spout utter gibberish, which not only demonstrates a total lack of understanding of the market instruments being discussed, but, also more generally, makes no sense whatsoever. This was the case in the older thread dedicated to the same topic and you have clearly not gotten any saner and/or wiser. Therefore, I refuse to engage in any discussions with you here and I ask you, please, to stop, just like I did previously. As an aside, where is the "Ignore" feature when you need one?Of course I respect the right to 'refuse to engage in any discussions ' and it is possible that my initial remark about swap insurance was dimmed and it should be specified more accurately.About "an aside, where is the "Ignore" feature when you need one?". From broader point of view we deal with the values of US Treasury demand for world trading and the world cost of variability of the US bonds. It is rather incomplete observation but probably the essence of the deviation between interest rate and swap spread.

OK guys, I have look over some of the posts earlier.I have a different perspective on this problem given that I work in ALM at a bank.What would be the correct discount curve to use?Everywhere I see, the swap curve is the discounted curve with some bootstrap.Let's say to price a 10 year treasury bond, we use the swap curve to discount. We shock the curve, get a DV01 number.Or how about we construct a curve with on-the-run treasury, then we get a different number when we shock the curve.How do I know which DV01 number to use?A 20bps difference on the 10 year is significant.Let's further assume that the bank is running a duration immunization strategy, how do we hedge?

"What would be the correct discount curve to use? " It is quite simple answer and it would depend on what is the definition of the discount notion we are going to use. Initially discount was defined by zero coupon bond. We pay $q at t to get $1 at T defines risk free discount on [ t , T ] interval. Whether it is unique discount definition or there exists other formal definition?

I agree, but as far as I know, all the models are base on swap curve.That means everyone is overestimating the price of securities and underestimating risk?This sounds a bit scary.

I am really really confused... Why, in general, would you use the swap curve to discount bond cashflows (other than analytics)? As to the DV01 calcs, as long as you keep your spread constant and your calcs are self-consistent, I don't really see the issue... Same with your hedges.

Well, I am not sure what traders are using.I know insurance software and ALM software are using swap curve to calculate.Even if I keep my spread constant, how can I be self-consistent?The DV01 number is going to be different when I choose different curve to discount.I don't think I can even use swap curve for analytics purpose.The 10 year swap spread is like 20bps now, which is huge.As for hedges1.no one can hedge 100 percent2.we swap fix for float to match funding, which means we use swap curve for swap and swap curve for bonds... not correct3.how can you price with treasury curve, there are so many missing points along the curve?

To calculate what, that's the question?For DV01, it's reasonable to use your LIBOR swap or OIS curve as a base. To calculate DV01, you can perturb the base curve, while keeping your swap spread constant for every pillar. This will allow your methodology to be consistently applied to bonds and derivs both. Separately, you can perturb spreads to obtain that sensitivity. In general, it may not be a flawless methodology, but, in my experience, it's perfectly adequate and will produce reasonable DV01s.As to the analytics, I was referring to OAS/z-spread calculations. Nothing wrong about using the swap curve in such a context.As to building treasury curves, well, where there is a will, there is a way. People still do it, even nowadays and, believe it or not, it mostly works fine.

I got your point.For DV01 calculation,swap curve isn't so bad.Of course, for pricing, it is way off.Thanks,Fish

QuoteOriginally posted by: kfcnhl Hi Guys, I am having trouble understanding why US swap/GOV spread could be negative for 5 year onward. How can the credit quality of banks be more than US government?My take on it: Treasury / Swap spreads are negative. And what?