wilmott.com

Hi, I have picked up some questions as I read through some CDS literature, your opinions and insights into these comments please.(1) In Asia, the CDS market is now more liquid than the cash bond market, in a number of sectors or individual credits, this implies that pricing in the CDS market provides a more reliable indicator of credit quality than the cash market.(2) In Asia. the universe of players that have access to the CDA market is smaller than it is in the cash market, hence volatility can be more acute in the CDS market.(3) How is the CDS/ credit derivatives market developing in Asia?(4) What are the concerns in using CDS spreads of similar credit ratings and industry to derive implied default probability and applying this probability to other CDSs? Is this the normal procedure?(5) Does liquidity in the CDS market reduces liquidity in the cash markets?(6) Can anyone provide an insight into how banks price CDSs (just a peek will do)?Thanks in advance,Gerard Tan

1. It is always better to have directly credit related price than a price that first has to be run through a model to arrive at an implied number, such a probability of default or recovery rate.2. Well, alot of the stuff in Asia gets done on an Asset Swap basis, but I wouldn't really say that its more volatile.3. Still growing, but obviously the heydays are past. Still will see growth well into the double digit area. Derivatives are always popular in Asia 4. Yes, this is normal, although one has to be careful, as with all proxy solutions. In general the method works, but you have to be careful about the specifics of the case. But this goes for all proxy solutions.5. No.6. Well, without intending to type my fingers bloody: There are generally two types of models: Structural models and intensity based model. Intensity based models try to get the dynamics of the default process under control by basically assuming a risk-neutrality between risk-free bonds and risky bonds.Structural models try to estimate when a point of default is reached by trying to model the volatility of the firms value and determine when it will be too low. They are calibrated with the use of bonds and equity and equity volatility. This is how they allow model-based hedges using equity rather than only the bond.Hope this helps.

Hi FDAXHunter,Thanks alot for your response.Some further questions:(1) To get the default probability, I can either use bond prices or equity prices (Merton's model) of say issuer ABC, or use a proxy issuer DEF to reverse derive the default probability (if DEF has the same ratings and is in the same industry). Correct? What I know about the default probability I can generate so far is that ABC will default within a year, is it important to derive the default probability at any time of the year?(2) Don't quite understand what you meant by Structural model or Intensity-based model (sorry!), is it what I mentioned above? As in Structural model means I can use corporate and treasury bond prices to derive the PV of cost of defaults etc, and eventually lead to default rates. Does Intensity-based model mean using Merton's model whereby a company's assets and equities are compared etc, then eventually derive the default rates?(3) I am stumbled by the Integral I saw in the CDS spread equation in Hull and White, do I have to get the inputs (be it the default rates, discount factors, PV of payments, PV of accruals) on a daily basis or on an annual basis? In other words, do we look at default in an annual time frame or on a dailt time frame?Thanks a lot. Greatly appreciated.Gerard Tan

Hi FDAXHunter,Thanks alot for your response.Some further questions:(1) To get the default probability, I can either use bond prices or equity prices (Merton's model) of say issuer ABC, or use a proxy issuer DEF to reverse derive the default probability (if DEF has the same ratings and is in the same industry). Correct? What I know about the default probability I can generate so far is that ABC will default within a year, is it important to derive the default probability at any time of the year?(2) Don't quite understand what you meant by Structural model or Intensity-based model (sorry!), is it what I mentioned above? As in Structural model means I can use corporate and treasury bond prices to derive the PV of cost of defaults etc, and eventually lead to default rates. Does Intensity-based model mean using Merton's model whereby a company's assets and equities are compared etc, then eventually derive the default rates?(3) I am stumbled by the Integral I saw in the CDS spread equation in Hull and White, do I have to get the inputs (be it the default rates, discount factors, PV of payments, PV of accruals) on a daily basis or on an annual basis? In other words, do we look at default in an annual time frame or on a daily time frame?Thanks a lot. Greatly appreciated.Gerard Tan

(1) Short-term timing is generally not very important. A small change in default or recovery assumption will have much greater impact than whether the default occurs at the beginning or the end of a coupon period. Whether a coupon payment is actually made can be important, also (of course) near the expiry of the swap.(2) A simple structural model is that the assets of the firm follow a Gaussian random walk, the common stock is a call option on the assets with exercise price equal to the face value of the debt, the debt value is assets minus call value. A simple non-structural model is the annual probability of default is equal to the difference between the yield on the issuer's debt and the risk-free rate.(3) The equations are usually done on a continuous time basis.