Recovery Rate

[ShootaMcGavin]

Hi all,

Chapter 18 introduces what's called the recovery rate d. This is used in credit risk models to allow for some reduced payment in the event of a default.

I was wondering about this "recovery rate" more generally, and how it's used in real-world applications.

To me it seems a proxy for the severity of a default given a default occurs -ie the bigger the failure of the company the smaller the recovery rate is likely to be. This is obviously impacted by other external factors such as ranking of the debt etc.

But what seems strange to me is that these credit risk models put most of the effort into calculating a probability of default and then just assume a broad say 60% recovery rate. It seems that without any meaningful analysis behind this recovery assumption the usefulness of the rest of the model is put in jeopardy?

Thanks,

 

[Mike Lewry]

Agreed. The material on credit risk in CT8 is not designed to be sufficient for us to take a serious stab at valuing a credit derivative.

 

[Oxymoron]

Hi all,

Chapter 18 introduces what's called the recovery rate d. This is used in credit risk models to allow for some reduced payment in the event of a default.

I was wondering about this "recovery rate" more generally, and how it's used in real-world applications.

To me it seems a proxy for the severity of a default given a default occurs -ie the bigger the failure of the company the smaller the recovery rate is likely to be. This is obviously impacted by other external factors such as ranking of the debt etc.

But what seems strange to me is that these credit risk models put most of the effort into calculating a probability of default and then just assume a broad say 60% recovery rate. It seems that without any meaningful analysis behind this recovery assumption the usefulness of the rest of the model is put in jeopardy?

Thanks,

Merton model has an analytical way of calculating recovery rates. You might want to check that out.

 

[Whippet1]

In the intensity-based models, given values for any three of the ZCB price, the risk-free rate, the recovery rate and the risk-neutral transition intensity function, you can find the implied value of the other one that you don't know.

In practice, I guess that recovery rates would be estimated by looking at actual recovery rates in past defaults.

 

[Mike Lewry]

You also have the complication that it's unrealistic to assume a recovery rate that is constant for all durations. This raises the question of what would make a suitable time-dependent function for the recovery rate.