quick and dirty way to do PnL

[RussianMike]

Does anyone know of a quick and dirty way to do PnL for a CDS position? (I know if you use spread DV01 in dollar terms it will work for small changes in spread). What about for larger 100bp+ changes?

 

[naeron]

use in the convexity of the underlying bond as well with the duration .

 

[RussianMike]

I'm not sure how this would work for a CDS? I must be missing something

 

[tim]

interesting question... i also want to know the answer....cause i have five pricing sources for one same deal and they give me five different numbers.......oh~~my god...

 

[Andy Nguyen]

interesting question... i also want to know the answer....cause i have five pricing sources for one same deal and they give me five different numbers.......oh~~my god...

Welcome to mark-to-market hell !
My advice: always pick the number that increases your P and lower your L
This part of the job is when you realize you hate your controller A LOT. Depends on your firm, your controller may live in another continent and can care less about your position. He will take the quote available to him without checking whether the quote is off. Like a few dozen bps or more.

It's your job to look after your group best interest and that includes marking the book the most correct way you can. Your controller most likely get the marks from the flow desk that trades that product if you are at IB. And if you are on a prop desk, you get the marks from dealers and Bloomberg. There is a huge difference between those marks specially if the product you trade is illiquid and OTC.

If you are lucky enough that your controller's office is nearby, you sit him down and show him what the ask/bid everyone is trading at. And hope that he will do it right tomorrow.

Since most books are MTM nowadays, it's always a daily fight. Your controller don't benefit anything from your book so there is no incentive for him to mark it your way.

That's been my experience with the whole MTM so far.

And Mike, what do you want to know ? If I remember correctly, besides DV01, there are bunches of other parameters we have on the book. If i understand exactly what you are saying, I may be able to say something useful.

 

[RussianMike]

If I remember correctly, besides DV01, there are bunches of other parameters we have on the book

Like what? What we are doing is for small changes we multiply DV01 by change in spread to get the new value.

 

[tim]

I dont know what else on the everyone's book.

but, Bloomberg CDSW provides the following things:

Repl Sprd: the spread to price the deal to par
Sprd DV01: by BB's definition, that's the value when the WHOLE spread curve move up 1 bps
IR DV01: the value changes when the swap curve move up 1 bps
Sprd KRR: it's like the value cahnges for each Key Rates

Besides, in CDSW, if go to "VIEW" dropdown menu, there is more detail measurements.

hope this could help a little bit....


Back to Andy's advice "always pick the number that increases your P and lower your L "

i'd like to do that too...but the auditors dont think so, especially that's something related to "credit" nowadays......and, of course dont even bother to mention that how it is a pain in the ass process to comply with the "FAS157 Fair Value" stuff..... after all....that's all about a battle between Accounting Values and real Economic Values

 

[RussianMike]

"Sprd DV01: by BB's definition, that's the value when the WHOLE spread curve move up 1 bps "

Since this is non-linear its only works for small changes in BP i.e. to multiply Spread DV01 by change in BP's. For larger changes it doesnt work as well.