5 ms·

Out of curiosity, how did you calculate this?

a4a4a4a4 4y ago
Not the person you're replying to, but you can calculate the "implied volatility" using the current option pricing vs the current stock price. As the consensus of price movement (up or down) increases, the option prices go up. https://www.optionsplaybook.com/options-introduction/what-is-volatility/ https://www.optionsplaybook.com/options-introduction/what-is...
FormerBandmate 4y ago
Credit default swaps pay the full value of a bond when a bond defaults, and pay nothing when it doesn’t. If a 1-year CDS is 10% of face value there’s a 10% chance of default in a year
- mason55 4y ago
  I'm not a quant or banker but I don't think that's true. When a credit event is triggered, there's an auction for recovery of the defaulted bonds/loans, and then recovery is what's left from par. In addition, you have to take net present value of the settlement into account. Money compounds, it doesn't grow linearly. Let's say there's a 10% chance of a credit event in a year, a 0% chance today, and the chance grows linearly (27 bps/day). Even if the chance of a credit event grows linearly, and you hold the recovery rate steady, the net present value of the recovery amount grows as a function of e. All that to say, I don't think what you are saying is correct. See here for the correct formula: https://news.ycombinator.com/item?id=35154072 https://news.ycombinator.com/item?id=35154072
trillic 4y ago
Implied probability of default over a given time-span can be approximated with the equation P = 1 - ( e ^ ( ( -S * t ) / ( 1 - R ) ) where S is the CDS spread and R is the recovery rate. The spread can be solved using the inverse S = ln ( 1 - P ) * ( ( R - 1 ) / t ) Probabilities and rates are both expressed as percentages not basis points. S is the spread. t is years. R is the recovery rate. Source is my notes from undergrad. Options, Futures, and Other Derivatives (9th Edition) by John C. Hull. Take all this with a grain of salt as I am not a quant. (but I am looking for a job!) Doing some additional reading, there are some more precise approximations but they are less general.[1] Last number I was able to pull up the $CS CDS was trading at 551 BP. Up from 446 yesterday (an all-time high for $CS) Lehman Bros hit 640 BP just days before collapse. [1] https://quant.stackexchange.com/questions/15986/how-to-compute-the-implied-probability-of-default-from-a-cds-spread https://quant.stackexchange.com/questions/15986/how-to-compu...
- lordnacho 4y ago
  What did you use as the recovery rate? Weird thing about the recovery rate is that everyone I've asked says the same thing, and I've traded CDS: the recovery rate is 40%. It's a bit of a free variable in that equation, and it matters. Trouble is how on earth do you estimate it? But random people I've met in the business will just say 40%, for every issuer, somehow.
  - oakwhiz 4y ago
    Sounds like the financial version of dark matter
    - lordnacho 4y ago
      A fudge factor that makes your models agree with observation? Yeah, pretty much. I was never an expert in CDS so I always wondered if other people had better ways of dealing with it. But nobody I've come across has ever offered anything other than 40%.
    - WastingMyTime89 4y ago
      It’s not necessarily. A peculiarity of finance as a field of study is that a lot of the people studying finance only care about direct applications and a lot of the people teaching finance had this mindset when they learnt. It’s easy to end up with some poorly taught material. If you carefully looked at the model which gave rise to the equation you are considering, there probably is a very tangible meaning to the figure everyone is ball parking.
      - itronitron 4y ago
        >> there probably is a very tangible meaning to the figure everyone is ball parking 40 is awfully close to 42
    - curiousgal 4y ago
      Not really. Quants modelled stochastic recovery rates and they saw that it doesn't really add much. Fixed recovery rates work just fine.

Not the person you're replying to, but you can calculate the "implied volatility" using the current option pricing vs the current stock price. As the consensus of price movement (up or down) increases, the option prices go up. https://www.optionsplaybook.com/options-introduction/what-is-volatility/ https://www.optionsplaybook.com/options-introduction/what-is...

FormerBandmate 4y ago

Credit default swaps pay the full value of a bond when a bond defaults, and pay nothing when it doesn’t. If a 1-year CDS is 10% of face value there’s a 10% chance of default in a year

mason55 4y ago
I'm not a quant or banker but I don't think that's true. When a credit event is triggered, there's an auction for recovery of the defaulted bonds/loans, and then recovery is what's left from par. In addition, you have to take net present value of the settlement into account. Money compounds, it doesn't grow linearly. Let's say there's a 10% chance of a credit event in a year, a 0% chance today, and the chance grows linearly (27 bps/day). Even if the chance of a credit event grows linearly, and you hold the recovery rate steady, the net present value of the recovery amount grows as a function of e. All that to say, I don't think what you are saying is correct. See here for the correct formula: https://news.ycombinator.com/item?id=35154072 https://news.ycombinator.com/item?id=35154072

trillic 4y ago

Implied probability of default over a given time-span can be approximated with the equation P = 1 - ( e ^ ( ( -S * t ) / ( 1 - R ) ) where S is the CDS spread and R is the recovery rate. The spread can be solved using the inverse S = ln ( 1 - P ) * ( ( R - 1 ) / t ) Probabilities and rates are both expressed as percentages not basis points. S is the spread. t is years. R is the recovery rate. Source is my notes from undergrad. Options, Futures, and Other Derivatives (9th Edition) by John C. Hull. Take all this with a grain of salt as I am not a quant. (but I am looking for a job!) Doing some additional reading, there are some more precise approximations but they are less general.[1] Last number I was able to pull up the $CS CDS was trading at 551 BP. Up from 446 yesterday (an all-time high for $CS) Lehman Bros hit 640 BP just days before collapse. [1] https://quant.stackexchange.com/questions/15986/how-to-compute-the-implied-probability-of-default-from-a-cds-spread https://quant.stackexchange.com/questions/15986/how-to-compu...

lordnacho 4y ago
What did you use as the recovery rate? Weird thing about the recovery rate is that everyone I've asked says the same thing, and I've traded CDS: the recovery rate is 40%. It's a bit of a free variable in that equation, and it matters. Trouble is how on earth do you estimate it? But random people I've met in the business will just say 40%, for every issuer, somehow.
- oakwhiz 4y ago
  Sounds like the financial version of dark matter
  - lordnacho 4y ago
    A fudge factor that makes your models agree with observation? Yeah, pretty much. I was never an expert in CDS so I always wondered if other people had better ways of dealing with it. But nobody I've come across has ever offered anything other than 40%.
  - WastingMyTime89 4y ago
    It’s not necessarily. A peculiarity of finance as a field of study is that a lot of the people studying finance only care about direct applications and a lot of the people teaching finance had this mindset when they learnt. It’s easy to end up with some poorly taught material. If you carefully looked at the model which gave rise to the equation you are considering, there probably is a very tangible meaning to the figure everyone is ball parking.
    - itronitron 4y ago
      >> there probably is a very tangible meaning to the figure everyone is ball parking 40 is awfully close to 42
  - curiousgal 4y ago
    Not really. Quants modelled stochastic recovery rates and they saw that it doesn't really add much. Fixed recovery rates work just fine.